TEXAS AGRICULTURAL EXPERIMENT STATION C. H. McDOWELL, ACTING DIRECTOR i College Station, Texas 1 BULLETIN NO. 677 DECEMBER 1945 FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT AND OTHER LIMITING FACTORS A. A. DUNLAP Division of Plant Pathology-and Physiology AGRICULTURAL AND MECHANICAL COLLEGE OF TEXAS GIBB GILCHRIST, President D26-1245-6M-L180 Bulletin] ‘l. ' Origina 1n Cotton plants become well-fruited under conditions that favor the production of a large number of blooms, little shedding and moderate vegetative growth. The studies reported in this bulletin show that adequate sunlight is an important factor in these c.on- ditions. Variations in the amount of light, such as accompany periods of cloudy weather, short days, and close spacing of plants, were found to result in impairment of fruiting and excessive vegetative growth. The heavy shedding of young bolls often noted by growers fol- lowing periods of rainy weather is probably due to the interruption in high sunlight intensity rather than to the direct effect of rain on the flower. Modified light conditions and other factors that induce shedding have been found to hinder photosynthesis and to reduce the carbo- hydrate content in cotton leaves. Differences in the sensitiveness of cotton varieties to unfavor- able light conditions indicate the importance of careful selection of varieties for a given region, according to prevailing weather conditions during the fruiting season. and the need for attention to these inherent varietal differences by the cotton breeder. CONTENTS ~ Page Introduction and Historical Backgxfiund . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5 Methods and Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10 - Definition of Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10 Experimentation . . . . . . . . . . . . . . . . . . . . . . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 11 Varieties ....' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 11 Source of seed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 12 Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 12 Temperature and humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 13 Variations in light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 14 Variations in soil moisture and temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 15 Taking of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 15 Analysis of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17 Light Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17 Effects of Cloudy Weather . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ._ . . . . . . . . . . . . . . . . . . . . 17 , Varietal differences in shedding responses to cloudy conditions . . . . . . . . . . . . . . .. 23 Seasonal Differences in Fruiting and Shedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 24 Varietal differences in response to seasonal variations . . . . . . . . . . . . . . . . . . . . . . . . .. 28 Effects of Reducing the Number of Hours of Sunshine per Day . . . . . . . . . . . . . . . . . . .. 29 Effects of Supplemental Lighting on Cloudy Days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34 Experiments with Low Light Intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 36 Continuous Low Light Intensity During the Fruiting Period . . . . . . . . . . . . . . . . . . .. 36 Periods of Low Light Intensities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Varietal differences in response to shading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 45 Effects of shading plants in the field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 48 Age of Plants and Fruiting Forms in Relation to Shedding . . . . . . . . . . . . . . . . . . . . . . .. 48 Close Spacing of Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 52 Effects of Variations in Soil Moisture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . o . . . . . . . . . .. 55 Relatively brief periods of wilting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 55 Variations in soil moisture and nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 60 Excessive soil moisture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 62 Effects of Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 65 High Temperature Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 65 Extended periods of high daily temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 65 Brief periods of high temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 69 Evidence of varietal differences in response to high temperature . . . . . . . . . . . . . . .. 69 Effect of low temperatures on shedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 69 Variations in Cultural Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 70 Varying Composition and Amount of Nutrient Solution . . . . . . . . . . . . . . . . . . . . . . . . . . .. 70 Comparison of One, Two, and Three Plants in Each Container . . . . . . . . . . . . . . . . . . .. 74 Other Experiments in Relation to Shedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 74 Effects of “Chemical Hormones” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 74 Wetting of Open Flowers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 76 Pruning and Related Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77 Topping of plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77 Removal of leaves from stem and fruiting branches . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77 Removal of squares from certain fruiting branches . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 78 Removal of bracts from fruiting forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .._.. 78 Covering of Fruiting Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 78 Comparison of black cloth and cellophane covers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 79 Mild Shade Treatments Followed by Clear Weather . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 80 Excessive Shedding by Plants with no Boll-Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 81 Quantitative Carbohydrate Analyses of Cotton Plants Under Various Conditions and Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 82 Effects of Shade on Carbohydrate Content of Cotton Leaves and Stem Bark . . . . . . .. 84 Effects of Wilting on Carbohydrate Content of Leaves and Stem Bark . . . . . . . . . . . . .. 86 Effects of High Temperatures on Carbohydrate Content of Leaves and Stem Bark. 86 Increase During the Day in Carbohydrate Content of Leaves of Different Ages . . . . .. 90' Analyses of Leaves from Plants Grown in the Greenhouse During the Summer . . . . .. 91 Analyses of Other Parts of the Plant in Relation to Shedding . . . . . . . . . . . . . . . . . . . . .. 92 Stem bark vs. root bark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 92 Woody portion of stem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 92 Young bolls and bracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 92 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 93 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 Literature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 ETIN NO. 677 DECEMBER 1945 ITING AND SHEDDING OF COTTON IN RELATION TO LIGHT AND OTHER LIMITING FACTORS A. A. Dunlap, Chief A Division of Plant Pathology and Physiology tton is considered a sun-loving plant, although the importance of 'ate sunlight in the production of cotton is n clearly understood. _dition to the water and temperature requirements, Canney (11) has “ abundant sunshine as a third dominant factor essential for high- of this crop and he has shown that the well established cotton fields “" world lie within the belts of “minimum cloud zones”, namely, n 10° and 35° N and 12° and 35° S latitudes. Even Within the can cotton belt, Canney has pointed out certain areas, including tal plain from New Orleans to Galveston, that have too little sun- for profitable cotton culture. Regions of “half cloudiness” annually lgnated as having too little sunshine, while “three-fifths cloudiness” ’ femely unsuitable for cotton. Although specific responses of the plant to poor light conditions are not described by Canney, it is ble to presume that stimulated vegetative growth, decreased flower- rmation, and increases in rates of shedding might result as coinci- ysiological disturbances. ‘p-many years, probably ever since the crop has been cultivated for 1;} ercial purposes, the shedding of young cotton bolls and immature ’ buds has been a matter of concern to cotton growers. In 18-99, f1(20) mentioned shedding of bolls along with the diseases of cotton in ,v a, and stated that it was a physiological trouble that often caused ‘erable loss. Although heavily-fruited cotton plants with but few ‘where bolls have been shed are often found in the field, it is con- pi» more or less normal for the cotton plant at time of maturity to shed from 20 to 50 percent of its immature fruits. Excessive shed- frequently amounting to 80 percent or more of the fruiting forms Qften commencing soon after the first bolls start to develop, may inoticeable losses in cotton yields. High shedding rates are fre- Agy associated with favorable moisture and fertility conditions which in close stands of large plants. Lower shedding rates are usually ted with cotton plants that have attained only a small or medium _e to low soil moisture or fertility. iew of certain studies involving the yields of cotton in the same _, in- different years brings out the fact that high rainfall during the 5- season is frequently accompanied by low yields of cotton. For , data taken by McDowell (40, 41) at the Iowa Park substation, irrigated conditions, from 1932 to 1942 show that those years during I conditions, may result from the shedding of non-injured, young bolls dur-_ 6 BULLETIN‘ NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION which the highest rainfalls were recorded were not the ones in which the _ best yields of cotton were made. In the Iowa Park area, boll-weevil damage was not animportant variable factor in the different seasons. Likewise Reinhard (44) reporting 14 years’ work at College Station, where boll-weevil damage is common, presented data which show that the poorest yields of cotton were obtained during the years which had the highest rainfall during May, June, and July (1929, 1936, and 1940). After _ making correction for boll-weevil damage, Johnson and Wadleigh (32) l showed that the yields of cotton in Arkansas in 1919, 1923, and 1930, when there was high rainfall in May, June, and July, were on the average one- - third) less than those of 1926, 1931, and 1934, when the rainfall for the~ same months was only two-thirds that of the low-yield years. These. writers state that rainfall of from 3 to 6 inches during July appeared most favorable for high yields of cotton and that monthly precipitationsi “above 6 inches in the summer months correspond to markedly decreased‘ yields.” Carver (12), working in Florida, states that there is “a general. tendency for low yields in rainy years and, high yields in dry years.” Al- though variations in many factors as well as light are associated withf variations in rainfall, it is possible that the increase in number of cloudy days during months with high rainfall may be, at least in part, responsible for these unfavorable effects of rainy seasons on yields of cotton. Under such conditions, a part of the reduction that may occur in the amount of fruit set on the plants may be due to increased rates of normal shedding. < Fruiting tendencies in cotton that are commonly. observed under field conditions indicate the importance of suflicient sunlight and suitable: moisture conditions in maintaining high- yields. Cotton growers often note ' excessive shedding of young bolls following periods of rainy weather or‘ showers. When the first bolls are nearing full size, they often make the; remark that a good yield would be obtained with the amount of moisture; already present in the soil, if clear weather could be had for the remainder. of the fruiting period. Unfavorable cloudy weather during critical period i may often be encountered in such regions as the lower Gulf Coast, in the‘ Corpus Christi area for example, where the cotton is planted relatively‘ early in the season and where cloudy or rainy weather may be expected ; each year about the time the first bolls are developing in late May and June. On June 19, 1945, examination of shedded forms gathered from‘ beneath cotton plants in an irrigated field selected at random near Mission,‘ Texas, showed 72 percent of the shedding to be “normal” and consisted of small bolls a few days after flowering. In this field the plants were in an actively growing condition about 30 inches tall with from 6 to 1. good sized bolls per plant and each plant had shed from 5 to 12 forms mostly from the lower fruiting branches. The plants were closely spaced and shaded one another in the usual manner. Large plants with relatively few bolls, frequently occurring under Lower Rio Grande Valley irrigated’ ing the early fruiting stages. If this early shedding could be prevented, smaller plants with much larger numbers of bolls might be obtained. w IIFRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 7 College Station on July 28, 1945, counts of shed forms gathered be- otton planted on April 17 showed 55 percent of the shedding to I’ field had previously received three applications of calcium arsenate 6n July 5, 11, and 20*. The plants in this case were about three tall, in a close stand and only the oldest bolls had‘ reached full another occasion in northwest Texas (near Aspermont, Stonewall ), a field in which the cotton plants were about 2 feet tall was l‘ ed on August 10, 1943, after a period of dry weather. The p-lants j: time showed wilting symptoms early in the day. Only a few of the q}; in this field had. bolls that were past the shedding stage, yet “half of the fruiting forms, including the squares of various ages as §as the young bolls, had been shed. No boll-weevil damage was found s field. These observations indicate that similar relatively heavy remature shedding of cotton may occur frequently in certain areas nder certain growing conditions. obtained... It is also difiicult to estimate the depreciation in yield generally poor fruiting activities as a result of unfavorable en- gental conditions. However, the general awareness of the cotton g_ to losses from shedding and the large number of scientific in- ‘tions that have been made of the problem indicate that the phe- ‘lg-v is of prime importance in cotton production. Prescott (43) that “the low yield in 1916-17 was associated with excessive I g, boll disease and insect attack,” in connection with cotton ex- “nts in Egypt. After working for several years in Mississippi on liedding problem, Ewing (23) has made the following statement: these figures it is evident that shedding must influence the yield .1 tton and is to be taken into account in any analysis of the fruiting ” The necessity in some areas of producing an early crop of may also make low shedding rates important in the attainment l,» yields. For example, before the invasion of Egypt by the pink orm, young cotton plants were kept dry so as to force root de- tent in preparation for a heavy late crop. This practice, accord- Templeton (51), caused considerable shedding of squares, a re- t was overlooked. With the danger from the boll worm, however, wers now water the plants to prevent early shedding and to obtain _ gest early crop possible. 1y reasons have been given for “normal” shedding ofcotton forms, which the most frequently stated causes are drought, high tem- ' es, load of fruit on the plant, fluctuations in environmental con- and incomplete fertilization of the flowers. ‘a for shedding as given most frequently in the literature include rmal and the remainder to be boll weevil and bollworm damage. ' ‘ ate estimates of crop losses from shedding in cotton are not. i cient soil moisture (1, 7, 23, 30, 38), and high temperature (7, 27, 38, 8 BULLETIN NO. 6'77, TEXAS AGRICULTURAL EXPERIMENT STATION 50). In addition, the presence of too much water in the soil or a high water table (2, 7, 34, 43) and daytime rain (38) have been mentioned as pre- disposing factors to shedding in cotton. In many cases sudden changes in certain, often unspecified, environmental conditions have been em- phasized as important considerations in this problem. For example, Berkley (9) found that cotton pl-ants grown under a long day length (12 to 24 hours) shed all of their squares when placed under 8 hours of light per day. These causes are distinct from the several different types of injury due to insect or fungous attack that are recognized as common factors in shedding and which do not constitute “normal shedding.” Speaking of conditions within the cotton plant, Mason (39) summarizes as follows: “The general conclusion was drawn that the proportion of shedding over any given period was the resultant of two opposing factors, the rate at which food was synthesized by the plant and the rate at which it was utilized in the maturation of the fruit; and that any check in the former augmented the rate of shedding.” In a few cases, research workers have noted possible correlation be- tween periods of cloudy or rainy weather and high rates of shedding oc- curring a few days later. In writing of his work with Sea Island cotton in the West Indies, Mason (39) states that, during the later stages of the plant’s growth, dark rainy days were “the invariable precursors of augmented rates of boll-shedding.” In Arizona, King and Loomis (35) noted heavy shedding of Acala cotton during rainy seasons. According to Thornton (52), the “small amount of sunshine” and high humidity ap- peared responsible for shedding during the wet season in Nigeria. It was also noted by Thornton that shedding during the wet period was not uni- form but was higher on certain days following cloudy or rainy weather with high humidity. The following statement by Ewing (23) is of further interest: “In some seasons it is quite possible that lack of sunshine during protracted cloudy weather may be detrimental to the cotton crop, quite "apart from the other conditions which accompany such weather.” Lloyd (38) made a detailed study of boll-shedding in Alabama and noted a tendency for higher rates 'of shedding to follow a few days after day- time rains. The partial explanation was advanced by Lloyd that rain pre- vented normal fertilization in the cotton flower, but he also states that “the shedding of bolls older than 8 or 9 days and which are therefore prob- ably past the danger from imperfect fertilization, can be caused ex- perimentally to shed in as short a period as younger bolls and large squares.” Mason (39) has stated that “destruction of pollen by rain was responsible for a small proportion of the shedding.” Apparently no at- tempt was made by Lloyd or other workers to associate increased rates of shedding following rain with impaired light conditions during the rainy weather. However, in discussing the effects of heavy rainfall on shedding, Earle (20) has stated that “the continued dark, cloudy weather would interfere with the normal action of the leaves.” Differences among types and varieties of cotton in regard to rates FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 9 p; shedding have been noted many times by different investigators. ' ey and Peebles (33) have shown that differences in shedding rates inheritable in crosses of Pima and Acala strains. Beckett and Hub- (8) found that 4-lock bolls tend to shed less readily than 5-lock w , in the case of Lone Star and Acala varieties, although environmental - 1,. 'tions may affect the relative susceptibility to shedding of these two l‘ _s of bolls. ew comparisons of cotton varieties in regard to their shedding ‘dencies in response to unfavorable environmental conditions have been Y»- in the literature. In the words of Balls, “the cotton plant in t is the slave of environmental circumstances which permit to it but T" display of individuality in the matter of its yield, and the conjecture i,» by the author in 1910, that the causes of yield deterioration should bought in the environment and not in varietal deterioration, seems to V lly justified.” It is well known that some varieties tend to shed a percentage of their forms than other strains and Ewing (23) has ed differences in shedding rates among certain varieties of cotton ississippi. Egyptian varieties were observed by Kearney and Peterson i, to show drought effects sooner than upland varieties. In a comparison Ticertain varieties under the dry-soil conditions in East Texas, Young j noted apparent drought resistance and relatively little shedding of age by the Dixie Triumph W. R. Str. 21 cotton. In Arizona, Hawkins found that Acala suffered more than Pima from, drought, as shown iheavier shedding of young bolls from the Acala plants. Also, under na conditions, King and Loomis (35) noted that the Acala variety ed a higher shedding rate than Pima during the last three weeks , uly, 1926 when there was heavy rainfall and plant growth was rapid. lation between the daily shedding of young bolls of Acala cotton and (tic pressures of expressed leaf fluids was found bylHawkins et al. _), although this correlation did not hold for the Pima variety. There somewhat better correlation between holl shedding and the concen- “on of boll fluid than with the osmotic pressure of the leaves. A in the case of shedding, relatively few reports have been found in vliterature pertaining to the effects of light on vegetative growth and ‘A al fruiting processes in the cotton plant. In regard to the influence y length, Berkley (9) found fruiting to occur in cotton under condi- n; of an 8-hour day, or longer, provided suitable temperature and other itions were maintained. In regard to the effects of varying light in- gity, Shantz (47) noted that cotton made the best growth at 1/5 to normal light and no great reduction in growth was noted at 1/ 15‘ of '- al light. No data on fruiting were reported by Shantz. In discussing ' effects of day length on different species of cotton, Konstantinov (37 )‘ j? s: “The conclusion to be drawn from what has been stated here is Qall cottons are not characterized by the same sensitivity to light.” ever, few of Konstantinov’s findings are concerned with differences _ iting capacities of upland American cotton varieties. As a result ‘ ading experiments during ‘two years in Northern Sudan with Delrect » to such unfavorable changes in light intensity as might be caused by 10 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION II, an American variety of cotton, Knight (36) concluded: “After making‘ due allowance for differences in diseases and pests, the results obtained in both seasons show clearly that partial shading (and hence, presumably, any prolonged heavy cloud-effect) “greatly reduces the yield of cotton, largely by a direct effect on the growth of the plants, though partly through an increased pest incidence,” and that “continued cloudiness may be a“ major factor in lowering the yield of cotton.” In this work, Knight found that continuous shade with coarse white cloth reduced the yield by nearly two-thirds, although the amount of shedding was less on the shaded plants,- due to production of fewer blooms under the shade. Novikov (42), work- ing with potted cotton plants in Russia, found that lowering the light in- tensity by shading to 21 percent of full sunlight resulted in greatly in-l creased vegetative growth, although the best yield of cotton was obtained under full-sunlight conditions. A study of Stoneville 2B cotton, includ- ing data on fruiting processes, has been reported by Eaton and Rigler (22). with plants grown under white-cloth shade in the greenhouse during the‘ winter in comparison with outdoor plants in full sunlight in the summer. Temperature studies in relation to the growth of cotton have been made by Balls (7) and byIBerkIey and Berkley (10), although little work has. been reported concerning the effects of high temperatures on fruiting processes in the plant. Many reports of the effects of drought conditions on growth and fruiting in cotton have been published (1, 23, 27, 28, 30, 40, 41, 43, 51, and others). It has been the primary purpose of this research to ascertain the im- portance of light efiects, especially periods of low light intensity, among the factors that influence fruiting in the cotton plant. In addition, an attempt has been made to evaluate these light effects in comparison with other factors such as variations in water supply and high temperatures? A few studies‘ on cotton fruiting involving variations in nitrogen and certain mineral elements have been included in this work. Varietal dif- ferences among upland cottons have been studied in relation to sensitivity cloudy weather. All of these studies have involved the taking of data under a wide variety of environmental conditions at College Station, over a G-yearfperiod from 1939 to 1945. A few preliminary reports of this work have been published previously (16, 17,18). I METHODS AND PROCEDURE The conditions and methods applicable in general to broad aspects of the work are stated in this section. In many instances, however, details of experimental conditions, materials, and procedure are given along with the presentation of data under subsequent headings. Definition of Terms Shedding is used to designate the dropping of bolls or squares by formation of an abscission layer near the base of the stalk. Unless other- FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 11 ildesignated, the termis used here to indicatei normal shedding which rsto abscission due to purely physiological causes and does not include “ding due to insect, fungus, or mechanical damage. e flower-buds up to the time of flowering are designated as squares; ‘flowering the developing cotton fruit is called a boll. Bolls are desig- t» as young or old depending upon the time that has elapsed after ering. The term form, or fruiting form, is used to designate a fruit- lstructure in either the square, flower, or boll stage (see fig. 1). this work, fruiting period is used to designate that period in the ~. iliof the plant beginning with the opening of the first flower and ex- ing up to the opening of the first bolls (time of harvest in this work). rcent set is the ratio obtained by dividing the number of bolls main- a past the shedding stage by the number of blooms. iting index is obtained as a ratio of a the weight of green bolls to fresh weight of leaves and stem, at time of harvest (opening of first _); this index could be used .to show the amount of plant growth neces- g to produce a boll. ‘ The. Experimentation rieties. During the first two. years of this Work the Startex variety tton was used in the preliminary experiments. Later, the Rogers 1. Various ages of cotton fruiting forms: A. Small squares; B. Large squares: Km; D. Small bolls a few days after blossoming; E. Bolls about 1/4 to 1/2 grown; -i approaching full size and too old to be shed readily. The bloom opens about 20 r the square becomes large enough to be een; young bolls reach full size in about : but from 6 to 10 weeks are require-d after blossoming before the boll is ripe enough Under ordinary conditions, most of the shedding takes place just after blossoming 12 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Acala No. 111 variety was used in several experiments and this was fol lowed with Stoneville 2B as the main variety. Since considerable variatioi was found among the above varieties in their reactions to differences i environmental conditions, a group of varieties was suggested by Mr. D. T Killough, agronomist, for comparative study. These were: Stoneville 2B Rogers Acala No. 111, A. D. Mebane Estate, Deltapine 14, Roldo Rowde Qualla, Coker 4-in-1, Washington (Delfos), Lone Star, and Half and Half In addition, a few plants of Dixie Triumph, Miller 610, Sunshine Rowde ' and Watson were used in a few instances. v Source of seed. Certified seed of the various varieties was in nearl every case obtained directly from the seed grower through Mr. Killoug and in most cases the grower was the originator of the variety. Culture. Two types of culture were used——soil and sand. The soil con sisted of a mixture ofiapproximately 3 parts sandy loam, 1 part rotte manure and 1 part granulated peat moss. In most cases, additional chemi cal fertilizer treatments were added after the plants had begun to se, fruit, consisting of measured amounts of dry nutrient chemicals (for 5 gal. of soil—usually about 10 g. (NH4)2SO-l, 5 g. of KH2PO4, and 2 1 MgSO. were added to the soil surface). Unless otherwise stated, th soil was watered frequently enough to prevent wilting for more than brief period. Usually the soil was allowed to become fairly dry, almos to the wilting point before a suflicient quantity of water was added ~10 moisten the soil to near the bottom of the container. With sand culture, the plants were grown in washed river or cree sand with nutrients supplied in solution. The river sand was reddish bro w’ in color and usually gave an alkaline reaction around pH 8. The graj creek sand was slightly acid to neutral (pH 6-7). Tests with acid show the river sandto contain many carbonate particles. The cotton seeds wer planted directly in the moist sand and given nutrient solution after emerg‘, ence. As soon as the plants showed several small squares they were give each day about 1 liter of a nutrient solution of the following approxima =1 compositionz‘ 90 p.p.m. N, supplied as NH4NO3; 40 p.p.m. K and 32 p.p.m P supplied as KHgPOi; 5 p.p.m. Mg as MgSOi; 3 p.p.m. Ca as Ca(NOg) 4 p.p.m. B as H3303; and a trace of Mn as MnSOl. The acidity of the fin solution was adjusted to about pH 5.0-5.5 with sulphuric acid. Commercia grades of nutrient salts, when available, and tap water were used in mak. ing the nutrient solution. In addition to the daily applications of th nutrient solution the plants were given tap water whenever necessary t prevent wilting. Tests of the pH of the sand after the plants had grown u. maturity indicated that the substrate in theicase of the brown river san remained near 7.0 pH while the gray creek sand varied from 4.8 to 6.0 pH Good growth of the plants was obtained with both types of sand, and r consistent differences in any of the results of the experiments appeare. that were attributable to differences in the substrates. In general, the -' sults obtained with soil cultures closely paralleled those with sand. As - FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 13 iv e, the sand-culture plants continued to grow for a longer period than mparable plants in soil and they were somewhat later in reaching a rmant condition following a heavy set of fruit. i, The containers in which the cotton plants were grown consisted for most part of glazed jars with a drainage hole in the bottom. For soil ’ture, 4-gal. jars were used most frequently, with certain experiments i olving 1-, 2-, 3-, and 5-gal. 'jars. With sand, 2-gal. jars were commonly although certain phases of the work were done with plants in 7-inch rnished flower pots, as well as 1-gal., 3-gal., and 4-gal. jars. Except l. certain special cases, a single plant was grown in each container. The , were normally spaced so that the stems of the plants at the soil sur- '_ e were from 20 to 24 inches apart. general, four series of plants were raised during a 12-months period. first series (spring) was planted in January and harvested in May. "s was followed by an early-summer series planted in April and harvested July. A late-summer series was planted in June or July and harvested * September or October. The fourth series (usually referred to as the , or fall-winter, series) was planted in September or October and har- in December or January. The spring and fall series were raised ' ely under greenhouse conditions. The early summer series was started flthe greenhouse and the plants placed outdoors during May, while the I summer series was raised entirely outdoors. l sect control, when needed, was obtained outdoors by dusting the nts at regular intervals with calcium arsenate for boll weevils and tying with nicotine sulphate for cotton aphids. In the greenhouse, ids were controlled by fumigating with “Nicofume Pressure Fumi- t”; tartar emetic and brown sugar solution was used against thrips; ii the houses were fumigated with “Cyanogas” for control of white ‘-~ and leafhoppers. f emperature and humidity. Records were made with a Friez Hyther- vph of the temperature and humidity under the conditions of the experi- vfj’ at different times of the year. Fig. 2 shows representative tempera- 1- records for greenhouse conditions in the winter and spring and for *- conditions in the summer. Under greenhouse conditions, with thermostat set at 85° F., the average maximum temperature during Jate fall and winter months (November to January, incl.) varied from nd 88°-93° on clear days, and the minimum night temperatures varied . Y» 65° to 85°. The relative humidity in the greenhouse for this period ined between 40 to 60 percent most of the time, with frequent periods f=l to 90" percent and infrequently as low as 20 to 30 percent. During pring series (March to May, incl.) the average maximum temperature greenhouse was somewhat higher (around 92°-95°) on clear days, l‘ in the fall. The relative humidity records show little difference be- Y» the fall and spring months" in the greenhouse. During the summer rs, the usual maximum. temperatures on clear days varied around 14 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION 90°-96° F., and at night the temperature varied from 74°-80° F. In- some years, certain periods had considerably higher maximum daily temperatures than in other seasons. Extremes in weather conditions caused occasional, ‘I brief periods of temperature conditions considerably beyond the limits of ; the averages given above. The relative- humidity outdoors in the summer varied from 25-85 percent on clear days to around 790-90 percent at a y night. A,‘ H $ Fig. 2. Weekly temperature records that represent conditions in connection with these experiments at different seasons of the year. A. Outdoors in shade of shrubbery, August 13 to 21, 1945. B. On greenhouse bench, March 28 to April 4, 1944. C. On greenhouse bench, November 16 to 25, 1940. D. Under greenhouse bench, May 20 to 26, 1941. Note the taller and much broader peaks in the summer curve (A) than in the fall or springs curves (B and C). Varia-t/ions in light. The cotton plants grown at different seasons of the year were subjected to distinct differences in light intensity and length of d-ay. For example, the light intensity measured with a Weston Model 60-3 Illumination Meter on clear days at noon in midsummer was around 14,000 foot candles while similar outdoor readings in December showed only about 8500 f.c. Likewise, the length of day (period between sunrise and sunset) varied from about 14 hours in June to about 10‘ hours in December. In addition to these seasonal variations,‘ the average per- centage of cloudy days as recorded at the Main Station Farm, College Station, over a. five-year period (1939-1943) was 30 percent for April and May, 26 percent for July and August, and 43 percent for November and December. l Light intensity on both‘ clear and cloudy days was measured on several occasions at different seasons of the year. Since much of the experimental work involved varying the light intensity by shading the plants, several FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 15 E‘ ‘c. élofthe readings taken under different weather conditions are listed in Etable 1. These readings were taken near the greenhouse with the target yqof the Weston meter directed towards the zenith. Within the greenhouse the light intensity was usually 15 to 30 percent less than the intensity out- i doors. a , Experimentally, differences in light intensity were obtained by placing .1 he plants indoors, by shading with cloth or slat shades, or by varying the spacing between potted plants. Black cloth which reduced the light from 90 to 98 percent was most frequently used. Some white cloth shades were also used, which reduced the light intensity about 70 to 80 percent. In every case, the shades were placed over the plants in such a manner p: to allow free circulation of air from the outside. Even under conditions bright summer sunshine, the differences in temperature between the fputside and inside of the shaded enclosures were not more than a few de- ogrees F. The shading effects of cotton foliage were also studied by spacing "plants close together (about 12 inches between stems) in comparison with fplants spaced farther apart (about 22 inches between plants). In certain pther experiments involving length of day or number of hours per day of 7direct sunshine, a light-impervious airplane fabric* was used for shading ‘during definite periods each day. I i Variations in soil moisture and temperature. In the. experiments with 2' ting, low soil moisture conditions were obtained by one of two methods: 1(1) Maintaining the plants at or near the wilting point for definite periods 7w frequent light waterings and (2) Daily weighing of the plant in con- tainer and adjusting to a definite weight each day by the addition of water. lptermittent periods of water-stress were obtained by allowing the plants remain wilted for a certain period before each watering. Certain differences in temperature occurred in the growing conditions at different seasons of the year, as shown in fig. 2. For studying the ef- g ts of higher-than-normal temperatures, the plants were kept in an shaded greenhouse during the summer or in a closed or partially venti- .= greenhouse at other times of the year. In one case studies of the Iects of low temperature were made in a non-heated greenhouse in the 'nter. QTaking of data. In many of the experiments, records were taken at ptervals of one, two, or three days of the number of bolls (and squares) 1-» from each plant. Occasional bolls or squares that showed injury, d_ue ‘(insects or other causes, were not included in the data. Final records ire taken at time of harvest of the plants, when the first bolls started open. By the time the first bolls were opening the outdoor plants of the mer series (soil or sand) had reached a determinate stage of growth, no new squares were being formed, and the growing points on the g in stem and lateral branches were inactive. Under greenhouse condi- when the first bolls opened, only the most mature soil plants showed lilGrade B airplane fabric “seconds” purchased from Landers Corporation, Toledo, Ohio. 16 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION this condition. Sand-culture plants in the greenhouse and soil-plants which had been subjected to different experimental treatments usually had not completely ceased growing when their oldest bolls were open. Measure- ments in taking of final data included: height of stem from the soil to the main growing point (tip‘ of stem); number of scars where fruiting forms had been shed, number of bolls around 2 cm. in diameter or larger (past the usual size for shedding, fig. 1, F.), fresh weight of these bolls, and fresh weight of above-ground parts of the plant exclusive of the recorded bolls. Table 1. Light intensity measurements, in foot candles, under various weather and seasonal conditions, at College Station, Texas. 1941-1944 Time of Time of Light year a Weather conditions intensity C. S. T. f. c. Jan. 27 . 1:30 P. M. Clear 8,600 Feb. 11 1:00P. M. Clear 10,000 April 11 12:45 P. M. Clear _ 12,400 April 24 '1 :00 P. M. Clear 12.800 April 24 12:30 P. M. Dark, rainy 1 .100 April 27 f 1 :00 P. M. Clear 13,200 May 7 1 :15 P. M. Heavy overcast, light areas in sky 1 .900 June 26 1 :00 P M. Few drifting clouds ' 15,000 June 26 2:00 P. M. Few drifting clouds ' 12,000 June 27 1:00 P. M. Large white clouds ' 16,000 July 6 2:30 P. M. Just after rain, clearing 2,000 July 10 12:30 P. M. Few clouds 14,000 July 16 1:30 P. M. Heavy, broken clouds 14,000 July 29 12:00 N. Raining 400 Aug. 16 A 12:30 P. M. Mostly clear, cloud over sun 2,500 Aug. 17 12:00 N. Hazy atmosphere; shadow bgely visible _ 6,000 Aug. 22 12:00 N. Cloudy, just before shower 1 .400 Aug. 27 12:30 P. M. Raining 820 Aug. 28 1:00 P. M. Shower 220 Aug. 29 12:30 P. M. Light overcast, scattered showers 2,500 Sept. 7 12:30 P. M. Cloudy, after shower 500 Sept.15 11 =45 A. M. Overcast 1.400 Sept. 20 12:30 P. M. Overcast 3.000 Oct. 25 12:45 P. M. Clear 10,000 Dec. 7 1:15 P. M. Heavy cloud, just before shower 30 Dec. 14 12=00 N. Cloudy i - A cs0 Dec. 17 1:00 P. M. Clear 8,400 _ .....u FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 17 f’ .1 ysis of data. Statistical analyses of variance have been made of ‘f1 of the data presented, as shown by the frequent references to the Tlcance of the differences between the mean values obtained. Insome (especially in the study of seasonal effects) analyses have been made }- combined sand- and soil-culture data. In these cases, chi-square have shown the data to be homogeneous. LIGHT RELATIONS e effects of variations in light intensity, day length, and amount of I ht available to the cotton plant in a given period were studied in p: different ways. In this work, the fruiting and vegetative ten- Les of different varieties of cotton were compared at different seasons year to show the effects of environmental differences in light and irature. Experimentally, low light conditions were provided by g the plants with cloth or by placing them in a shaded greenhouse ioratory room. Close spacing of containers with cotton plants was : without changing other environmental conditions appreciably. When ‘by spaced the leaves of neighboring plants intermingled and shaded fanother. The effects of brief periodsof shade and of longer periods light intensity were both studied in relation to shedding and fruit- ndencies. In addition to variations in day length encountered under f8‘ erent seasonal conditions mentioned above, the effect of the length _e daily period during which cotton plants were exposed to the direct ght was studied by shading the plants for definite periods each day. iny cases, shedding records were taken and examined in connection p‘ the preceding treatments or weather conditions. Wherever varietal jxresented following those showing the effect of such conditions on the j- cotton plants in general. Effects of Cloudy Weather A. many of the experiments conducted from 1939 to 1944, the young l_ that abscised were gathered and recorded separately for each plant, fly at 2- or 3-day intervals, from the time of the first shedding until fpldest bolls had begun to open. From these records graphs have been l. that show these variations. In addition, the weather records for l' same periods have been secured from the meteorological data. taken he Main Station Farm about one mile distant from the site of the ‘riments. The graphs showing the rates of shedding and the condition e weather (clear, partly cloudy, or cloudy days) during the same _- are combined in figs. 3-6, inclusive. i examining the graphs showing the effects of cloudiness upon shed- Q1, it should, be kept in mind that a certain amount of shedding occurs L. 11y during the fruiting period of the cotton plant. Under uniformly Q weather conditions, a greater portion of the shedding has been found means of providing low-light-intensity conditions for the experimental _ ~7~ nces in response to these variable conditions were found these data _ 18 ‘BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION I 9O - fl-cwuov - 3 3Q _ Q-PARTLY CLOUDY A - if, 7° _ EI-FAm .. g _, 6O - O D u so — 0 PLANTED was o: 4O - 8 2 3O — D Z g 2O - IO ,- 1 1 1 1 1 1 1 1 1 1 1 1V1 1 1 1 1 1 1 l I l l I l 4 6 8 IO l2 l4 I6 I8 2O 22 24 26 28 2 4 6 8 uovsmsea oeoeuasn 1s as Ei§:?i?i?i'-IIIitizfliiIIISQFS:kit?I§3II%%%§§1EI%Et¥St?$:FiEi?III%II%IE1El 400- _ 3 42° [:1 c1011 B _ f, - uawr CLOUDY - ,, aso- - _, _ Q cuouov _ 5' aoo~ - Q -- .. a 240- - m '- —f' l: l80- - 2 l- u- g 120- - so - - o _ 1 1 I 1 1 1 1 1 I 1 I I I I I I I I 1 I 1 I I 1 1 I I I I I I 1 I I ‘- 2 4 6 8 IO -I2 l4 l6 l8 2O 22 24 26 28 3O 2 ApmL MAY I940 FiEIZ1EIEIE-II%%?E¥E1_E?E§E 640 a seo- C - x Inn " 4ao- ' m I. .- . J _ 3 40o- o __ .. 3 320 — t: 1r ’ __ g 24o - _ a l i g ueo- __ ao- - I "|,,..4. 111. 16111111111111 2O 22 24 26 28 3O 2 4 6 8 IO I2 l4 I6 I8 2O 22 2494‘ JUNE l JULY ' Fig. 3. Efiects of cloudy weather on shedding. A. Two series of greenhouse plants in sand culture, one consisting of 43 Rogers Acala variety planted August 8, and the other, 39 plants of Startex variety planted August 24, 1939. Note similarity in shedding curves for. the two ages of plants. B. Curve showing shedding record of 106 Startex plants in sand, planted January 13, 1940. C. Outdoor summer series of 112 Rogers Acala plants in sand culture, planted April 10, 1941. Note peak in shedding curve on July 15, associated with overcast sky due to a Gulf storm. The daily weather conditions, as to cloudiness, are shown at the top, in each case. ‘V. _ 1 .,‘._..._».. . .1;.A2MIQj-;A-nn|@»1L1_OAuJAb . 1.. 1.. _.-___11..1:.... ...»..\.,.,11..1...\.e_.1_;~. 1.. “Malia... _~n..1,...m.r..1.. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 19 occur late in the fruiting period. At times, indeed, no sharp peaks i}: in the shedding curve, as may be seen, for example, in the sand J during the late summer of 1944 (fig. 5, C) and in the behavior of i April 8th planting in 1944. (fig. 19), when light and temperature con- ins were relatively favorable and uniform. In’ most of the graphs (figs. one or more distinct peaks are apparent in the shedding curves. In i " tically every case, the peaks of shedding are preceded by one or more i-vdy days that usually occurred two to six days in advance of the shed- i‘ g response. By mechanically injuring the young bolls, Lloyd (38) " d that the time necessary for abscission- was six days or less, depend- " upon the age of the boll and also upon the intensity of the stimulus. fTWlll be indicated in experiments described later, shedding may be ex- i ed to occur as early as two days and as late as six or eight days fol- ,‘ g the initiation of a low-light-intensity period, depending probably n the condition (amount of sugar and starch reserves) of the plants as designations of cloudy and partly cloudy days as shown in these graphs j more or less arbitrary in that certain days shown as partly cloudy, . example, may have been almost clear while others may have been w heavily overcast. Likewise, some of the days designated as cloudy have been only lightly overcast, with a fairly high light intensity most of the day. xamination of the shedding curves for the 1939 fall-winter series of n- plants in the greenhouse (fig. S-A) shows a high rate of shedding a a prolonged period in November coincident with about eight consecu- cloudy days. Following somewhat less cloudy weather, the curves "w to a low point at the end of November, only to rise sharply again f_December 4 during clear weather but following several cloudy days in November. The shedding curve for the spring series in 1940 (fig. ) shows a trend towards an increase in shedding during May as the its become older. The four peaks in this curve also correspond more or "7: regularly with groups of cloudy or partly cloudy days that precede by short intervals. The single abrupt peak in the curve for early y... plants in 1941 (fig. 3-C) is especially interesting since it occurred iithe middle of a week of overcast weather caused by a tropical Gulf i‘ - in July. The other parts of the fruiting period for these plants was "mpanied by rather uniformly clear weather and no other shedding - occur on the graph. in, in the fall series, 1940, three definite peaks occur in the shed- curve (fig. 4-A) following groups of cloudy days as shown by the dither data. Similarly, in fig. 4-B for the fall-winter 1940-41 series l? * in fig. 4-C for the fall-winter 1943 series, there are three distinct ding peaks, in each case, which can be associated with cloudiness g." the few previous days. e graph for the 88 Rogers Acala plants in late August and in Sep- ber 1941 (fig. 5-A) shows a high, sustained rate of shedding from Sep- as upon the intensity of the shading. It should also be stated that 20 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION o To‘ 0'0‘ 0'0‘- 1 o 1 o c 0 o Q 01.5111: PARTLY CLOUDY E cuouov 240 > 2IO llll I80 I50 I20 9O NUMBER OF BOLLS SHED 60 llllllllllllllllf lllllllll 3O 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1 1 1 1 1 1 1 1 6 8 l0 I2 I4 I6 I8 2O 22 24 26 28 3O 2 4 6 8 I0 I2 l4 16 l8 N0“ 0E0 1940 . ~ . - - < - - . . , . . . . . . . . . . 00-100000000000‘ waoo-wofff -v\o'.~y- r .-~vnooq-- ~-~ -... .. -0o0010¢OOrQ40o 00100301...’ be... 01 .0_0Q;.,.,¢4 '\Q.‘\ . M."w‘.'.%'.%?»'.'1%?-t.rstt-I» 32.28.91 A ‘fir: 329:1 [1521-131-335 '15:‘ 3'3 9:9: 8O N O O O I [j CLEAR - PARTLY CLOUDY B : Q ccouov - (I O b O 0| O NUMBER-OF DOLLS SHED a s; O [IIIIIIIIIIIIIIIIIIII I 1 a s 1 s 11 1a 15 11 19 21 2s 25 21 2s a1 2 4 s 956- am. _ 1041 . . _ a.’ ,.,._,...__..._,,.,_... w, ,fl-._...__...‘.- , , . . - n.0,. . _. . ' _ II-t.s.dIa.t-e.z-z.z-r.ae.-i$101.2. 309130?! IIQIG . . . . _ ‘ 270 _ 24o — n: O I I80- LLS SHED 3150- or F5 O II 90- 60- NUMBER 1 30- 111111111111111111111 22 24 2s 2s so 2 4 e a 1o 12 14 1s 1a 2o 22 24 2s novemsea ozcemsen I943 Fig. 4. Elfects of cloudy weather on shedding. A. Shedding curve of a fall-winter greenhouse series of 73 Rogers Acala plants in sand culture, planted September 4, 1940. B. Record for 35 ‘Rogers Acala plants in sand culture, planted September 25, 1940. C. Two seriesof plants-SO in sand and 60 in soil—each series made up of 10 varieties. There was a large percentage of cloudy days in each of the three fruiting period shown here and shedding was relatively heavy over an extended period in each case. E FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 21 i-tember 12 to 18. This broad peak coincides with a period of 10 cloudy or partly cloudy days during the 12-day period from September 7 to 18, in- clusive. The shedding record of plants during April and May 1944 shows ~two distinct peaks (fig. 5-B), one about April 26 following five cloudy days between April 16 and 23 and the second peak about May 3 following another period of several cloudy days. Fig. 5-C shows a marked peak of > shedding about September 2, 1944 after a period of twelve consecutive days of cloudy weather. The sand plants in this figure, that matured somewhat ‘later than those in soil, did not show a high peak of shedding in the clear weather during the latter part of their fruiting period. It is of interest to note the shedding curves of cotton plants of different ages in each of the two graphs, figs. 3-A, 4-C, 5-B, and 5-C. In the first ,of these graphs (November-December 1939), there is a difference of 16 Ldays in the age of the plants represented in the two curves, and in the i third case (fig. 5-B, April-May 1944) one series of plants (broken line) iwas nearly two weeks older than the other. In fig. 5-C (Aug.-Sept. 1944), the earlier (soil) plants showed an extremely high shedding rate following cloudy weather late in August, and at the same time the later maturing plants (in sand) showed a rate almost as high as at any time during their ientire period of fruiting. In spite of differences in age, shedding behavior in the younger plants closely paralleled that of the older plantsin each jcase. Since all of these plants, with the exception of those represented in ifig. 5-C, were grown in the greenhouse where uniform, favorable tempera- itures and soil moisture conditions were maintained, variable sunlight in- itensity appears as the most likely factor responsible for the close co- ‘l-incidence in peaks of shedding in the young plants in comparison with the .older plants. If cotton plants tended to show excessive shedding rates iwhen they reach a certain stage of maturity or fruiting regardless of en- l-‘vironmental conditions, it would be expected that the peaks of shedding =in the younger plants would occur at a somewhat later date than in the case of the older plants. Such a condition has not been found in these experiments . u , After consideration of these shedding graphs (figs. 3, 4, and 5), it seems "reasonable to conclude that excessive shedding in cotton is likely to follow junfavorable weather conditions, such as a few to several cloudy day-s oc- icurring consecutively or close together. These results are based on data ftaken on nearly 700 individual plants grown at different seasons of the year over a 4-year period. In only one case, under ordinary cultural con- tlitions, has a distinct shedding peak occurred following a period of several ys of clear weather and this occurred during a period of unusually high '3’This conclusion that cloudy weather encourages shedding in cotton is 35%| agreement with observations by both Mason (39) and Lloyd (38). oyd’s graphs (on pages 89 and 96, l.c.) show that the peaks in the shed- g'curves for 1912 are due to the probable effects of rain. Furthermore, 22 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION I%&%fiI%%§fiIII£fi%fi%fiI%&@I% ‘TX 360 — E] CLEAR p I 32o : PARTLY cuouov A - 3 - Q CLQUDY I I 28o ~ _ w h. i 3 24o - _ 6' 20o I ' m I * u. H I: o I60 — _ m F. G I‘; 12o - _ E h‘ i g so - - 4o - I o h I I I l I I I I I I I I I I I I I (I I I I I I I I I I I I I I I I I I I I.‘ 2s 27 29 3| 2 4 s s no |2 a4 I6 1e 2o 22 244 2s 2s so AUG. SEPT. |94| I I l I I I I] I I fill I I l friri§iitiliiéiééiéiiIE:E1EIR§I&§iEiFiI§§III§§§ 21o - CLOUDY Q PARTLY CLOUDY D Z40~ ZN)- CLEAR I80~ numacn or sorts snco a O 1 PLAN-TED FEB.3 90" 60- 30- . —~-// I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I LE1 no I4 I6 l8 20 22 24 2s 2s 3o 2 4 s s no APRIL MAY I944 36 - /\ - - I I -1 O a2 - I \\ - gé" " I p-sosr-ransv sLoou, Aurlw I m 23 " I \ ._ | \ =1 '3 24 - I I - -l _ I \ w O | d m 2O - | I A u - | I I SAND-FIRST sLoou, sums "‘ o .. I I - I 6 ‘ \ 2 ‘ I ‘ m '2 “' I s - I " z 8 — " 4 _ - _I I II I I II I II I II 2| 2325212931 2 4 s a IO n2 l4 as I8 202224 AUGUST SEPTEMBER I944 Fig. 5. Eifects of cloudy weather on shedding. A. Shedding curve during late summer for 88 Rogers Acala plants in sand culture, planted June 14, 1941. B. Two series of green- house plants in the spring—114 plants (broken line) made up of equal numbers of six varie- ties, planted in 3-gal. jars of soil on January 21 and 64 plants (solid line) of two varieties, planted in 4-gal. jars of soil on February 3, 1944. Note similarity in shedding curves for the two difierent ages of plants. C. Two series of late summer plants; the 8 plants in S-gal. jars of soil (broken line) were somewhat earlier than the 11 sand-culture plants (solid line) although both were planted on the same day, June 24, 1944. Note the increased shedding in both groups following a period of cloudy weather late in August. ‘aw ¢umv ~ - Y"Y"_I:'__IVW“- - g l I u 4o- O _ E so- D g _ z 20f- no- , _| l I I TTT‘ 68IOl2l4l6l82022242628302468l0 ApmL ' MAY I944 eo 7o_ [:1 CLEAR B szB—-a--4>- qgL.--m_m _ ' _ z-gé- --------------------- on .- 5°" E CLOUDY ' ' ' NUMBER OF DOLLS SHED 0| O I 1 4Q" 3Q- 2Q- l0" -| | | 1| ., ‘ | I " l0 l2 l4 l6 l8 2.8 3Q 2 APRIL MAY. i944 Fig. 6. Varietal dilferences in shedding responses of cotton to periods of cloudy weather. A. Shedding curves for 16 plants each of Rogers Acala and Stoneville 2B varieties, planted in soil January 21, 1944. B. Similar curves for Stoneville 2B, Rogers Acala, A. D. Mebane Estate, Deltapine 14, Qualla, and Half and Half varieties, planted in soil February 3, 1944. Note the tendency for certain varieties to start shedding more quickly than others following a few cloudy days from April 17-22. ‘ Lloyd shows (graphically on p. 94, l.c.) that the shedding of bolls Which were in open flower on rainy days was much more severe than that of bolls from flowers which opened on clear days. The heaviest shedding was also found to take place around the fourth to sixth day after flower- ing (time of known rain occurrence). These periods of excessive shedding following rainy weather as reported by Lloyd correspond in all major respects with those obtained in this work, even under our greenhouse con- ditions where rain did not come in contact with the pollen in the open flower. Ewing (23) also noted increased rates of shedding following rainy days as compared with rates for young bolls whose flowers had opened during rainless weather. Varietal differences in shedding responses to cloudy conditions. It may be seen in fig. 6-A that the Rogers Acala variety of cotton began to shed 24 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMIQNT STATION its young bolls somewhat earlier than the Stoneville 2B plants of the same age, apparently in response to the cloudy-weather stimulus occurring a few days previously. The higher shedding rate of the Acala variety was first apparent on April 16 and this difference was noticeable until April 23. Stoneville 2B did not show this first marked response in shedding until about 2 or 3 days later, and thq shedding peak also occurred 2 days later (April 28 instead of April 26). In view of the fact that an early set of bolls is characteristic of the Stoneville 2B variety, this early shed- ding might indicate a considerably greater sensitivity to the shedding stimulus in the Acala variety. Later, during May, the shedding graphs are quite similarfor the two varieties. On the average, both varieties, up to the time of harvest, had shed 16 young bolls per plant although the average Stoneville 2B plant in this test had 13.6 bolls and showed a final set of 46 percent as compared with 10.1 bolls and 39 percent set for the Rogers Acala. In a greenhouse test of six varieties during the early spring of 1944, three of the varieties (A. D. Mebane Estate, Qualla, and Rogers Acala) showed (fig. 6-B) a marked early shedding of young bolls, beginning on April 18 and reaching the highest peak on April 22. This early shedding was preceded by three days of cloudy weather from April 17 to 19. The other three varieties (Stoneville 2B, Deltapine 14, and Half and Half) showed practically no shedding during this period, although all varieties commenced blooming at practically the same time. These results are com- parable with data reported later showing a higher degree of sensitivity in certain varieties to unfavorable light conditions. Seasonal Dilferences in Fruiting and Shedding During the summer of 1941, the check plants in an experiment with the Rogers Acala variety produced an average of 24 bolls per plant. Under greenhouse conditions in the late fall, plants of this same variety, grown in the same kind of sand and with a nutrient solution of similar compo- sition, produced only 6 bolls per plant. It was also noted that the fall- grown plants were nearly equal in fresh weight of leaves and stems and also in height to those grown during the summer. These results led to the comparison in 1943 of three series of plants; spring, summer, and fall, each containing 140 plants; these were comprised of 14 plants each of 10 varieties, 8 of which were grown in sand culture and 6 in soil. The results of this test, with the data for all varieties averaged together but with the sand and soil plants listed separately, are shown in table 2. In this table, attention might be called first to the close similarity between the sand-grown plants and those grown in soil, in regard to size and fruiting capacities. Also, the vegetative growth and fruiting be- havior of the cotton plants in the greenhouse during the spring are much more closely comparable with those raised outdoors in the summer than with the greenhouse fall plants; the greatest difference between the spring and summer series was a somewhat smaller boll weight for the former. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 25 _ In comparing the fall series of plants with the summer series, a tendency towards a marked reduction in fruiting capacity was found in the fall i plants and a less striking difference in vegetative growth. The fall plants in the greenhouse, as may be seen in the fall : summer ratio in the last column of table 2, were nearly equal. in height to the summer plants, al- though the difference in height shown between the two seasons was found significant. The fall plants were one-third smaller in regard to fresh weight of leaves and stem, and the differences were again found to be sig- onificantp Likewise, a highly significant difference was found between iisummer and fall plants in regard to number of blooms per plant. At the same time, the fruiting of the fall plants was greatly reduced- iamounting to only 35 percent of the summer plants in number of bolls produced, and to only 29 percent in the fresh weight of bolls (both dif- iferences are significant at the 1 percent level). The percent-set and fruit- Eng-index figures are likewise low for the fall plants and are equal to only one-half or less of the comparable figures for the summer series. An ex- planation of at least a large part of these difierences seems to lie in the relatively larger number of bolls shed in the fall series. There was no significant difference between summer and fall in regard to total number ‘7 of bolls shed per plant. This large number of bolls shed in the fall, together ¢with a decrease in number of blooms, accounts for the much lower percent- set in the fall-winter series. Even though a. smaller yield might be ex- ipected in the fall, a correspondingly low percent-set figure might also be expected provided the plants had similar fruiting tendencies during both seasons. The fruiting index was less than. half as large in the fall as in the summer. These results indicate that some factor (or factors) aside from vegetative growth and aside from square formation is contributing to a marked degree towardsithe impaired yield of bolls in the fall-grown Table 2. Vegetative growth and fruiting behavior of cotton plants in sand and soil culture at different seasons of the year. Average of l0 varieties—80 plants in sand, 60 in soil, in each of the three seasonal series, 1943. Ratio Data per average plant (grgsnrliiroise) (ggidliiioig) (greelhililmiuse) lsglrlidsgzmsrgilg Sand Soil Sand Soil Sand soiT Percent Height—cm. 96 so 92 9o so s7 95 ightl stem and lcaves—g. 281 296 288 263 189 183 68 ———— i umber of blooms 32 31 3o_ 33 23 22 71 lumber of bolls shed 17 1o 14 15 17 1o 114 umber of bolls set 15 _ 15 1o 1s s o 35 cent of bolls set p 47 4s 53 55 2o 27 49 eightl of bolls—g. 357 304 435 437 126 13s 29 “itingindex? 1.3 1.0 1.5 1.9 0.7 0.3 44 26 BULLETIN NO.‘ 677, TEXAS AGRICULTURAL EXPERIMENT STATION a Fig. 7. Two cotton plants (leaves removed in lower picture) showing relatively poorer set of bolls by Lone Star variety on left as compared with Deltapine 14 variety on right,‘ when grown under winter greenhouse conditions. In this experiment, the former variety averaged 3.2 bolls per plant while the latter averaged 9.4 bolls (6 plants in each case). In the upper photograph, it may‘ be seen that the two varieties made a similar amount of vegetative growth. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 27 nts. Heavy shedding of young bolls appears to be a factor worthy of "nsideration in this connection. The following facts concerning the weather conditions during the period ;_ growth of the plants following opening of the first blooms, in each of three series at different seasons, may be significant. During the 45 ys preceding harvest of the summer series there were 11 cloudy days compared with 19 cloudy days during this same relative period in the p, l series. There were therefore, over 70 percent more cloudy days dur- 3- the blossoming period in case of the fall plants as compared with Jose in the summer series. There were only 7 cloudy days in the same lative period for the spring series of plants. étThe length of day for these same periods was on the average: 13 hours r the spring series, 14 hours for the summer, and 10 1/3 hours for the y lseries. In connection with a consideration of the effects of day length cotton it may be recalled that Berkley (9) found this factor to be of pile importance ordinarily, as far as formation of flower buds was con- _ ed. Berkley did report, however, an important interaction between "th of day and temperature, in relation to the production of squares. e following quotation from correspondence with Dr. John W. Shive (49) x‘ of considerable interest at this point, since he grew cotton plants to lturity at different seasons of the year in a greenhouse at New Bruns- "pk, New Jersey, where seasonal differences in length of day are much _.ater than those for Central Texas: “We have carried out considerable prk With the cotton plant here, but I have never noticed any sensitiveness this plant .to length of day. We have grown it both in summer and in fter. It grows more slowly in the winter time than in the summer time in does not get as large in winter as in summer. It seems to bloom larly and without much change in the cycle at any time of the year. imay be that some varieties are more sensitive than others, but I do not .1 anything about this. None of the work that we have done would in- fate whether the plant is sensitive to len.gth of day, but I do know that nges in temperature make a marked difference in the manner of its h.” There is no doubt but that the lack of sensitiveness on the part ‘cotton mentioned by Shive refers to the ability of the plant to grow " bloom well under widely different day-length conditions. Indications t length of. day may be important in the accumulation of a sufficient I nt of light, however, ‘to provide optimum conditions for cotton fruiting presented in the following section of this bulletin, he importance of temperature, also noted by Shive, cannot be over- "ked especially in regard to growth differences of the cotton plants at erent seasons of the year. As shown in fig. 2, the average daily perature is much higher in the summer than under greenhouse condi- ns in the winter or in the spring. Since the growth of the plants in the g was found to equal or to exceed the growth in winter, it could ‘i-assumed that the improved light conditions (longer days and less cloudy ther) were more important than the temperature differences which .. m.\._fl 28 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION were somewhat similar in these two seasons (fig. 2, B and C). Data to be presented herewith indicate further that a few days of shade or reduced - day length may also lessen cotton fruiting processes, under approxi- mately identical temperature conditions. Varietal differences in response to seasonal variations. Part of the data in table 2 has been reassembled according to varieties in table 3 to show varietal differences in growth and fruiting between culture during the-late » fall in the greenhouse and comparable culture outdoors during summer. The ten varieties of cotton used here include: Deltapine 14, Roldo Rowden, Stoneville 2B,. Coker 4-in-1, Half and Half, Washington, A. D. Mebane Estate, Qualla, Rogers Acala, and Lone Star. The data recorded in table _ 3 represent the average of 14 plants—8 in sand and 6 in soil-for each variety. The measurements in the summer are given just above those for the fall in each case and the ratio, given just below these two figures, shows the relative value of the fall data as a percentage of the summer value. Therefore, the varieties with the largest percent figures for a ratio represent the ones least seriously affected by the fall conditions. The varieties have been arranged in this table so that those (Stoneville 2B, Half and Half, Washington, Roldo Rowden, Coker 4-in-1, and Deltapine 14) showing the smallest effect on fruiting capacities from the adverse environ- A mental conditions in the fall appear on the left side. The varieties more seriously affected (Rogers Acala, Lone Star, Qualla, and Mebane) in. respect to fruiting processes, as indicated by number of bolls set, weight of bolls, percent-set and fruiting index, are listed on the right side of table 3. It may also be seen that the difference between varieties, in re- gard to their responses to unfavorable fall conditions, is much less in the case of the vegetative processes, such as height and weight of plants, and in the number of flowers produced or number of bolls shed, than in case of the above fruiting process (see fig. 7). Statistical analysis of the data, which have been summarized in table 3, pertaining to the number of bolls set, number of bolls shed, and the fresh weight of the plant showed significant interactions (11 percent level) be- tween varieties and seasons. The interaction between varieties and seasons for boll weight was significant at the 5 percent level. There were no sig- nificant interactions in the data on plant height. These facts would in- dicate that certain varieties reacted differently than others, in regard to some characteristics, to the fall conditions as compared with summer per- formance. Further confirmation of varietal differences in response to. seasonal effects is shown in the ratio figures computed for each variety and for each type of data. The difference required for significance be- tween varieties in these ratios is shown in the last column of the table.~ Examination of the ratios shows that the four varieties on the right (Lone Star, Mebane, Qualla, and Rogers Acala) were adversely affected in their fruiting activities in the fall to a greater extent than the other varieties I tested. In regard to impairment of growth processes in the fall, however, ; the difference between varieties is small and in most cases insignificant. 1 FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 29 : J differences between the ten varieties in regard to growth and fruiting ' dencies in summer and fall are shown graphically in fig. 8. Elfects of Reducing the Number of Hours of Sunshine per Day ~ In order to obtains some information as to the factors causing the dif- erences in fruiting capacity of cotton at different seasons of the year, fcertain experiments were conducted in which the number of hours per day ofgsunshine on the plants was made a variable factor. In this manner, was hoped to learn whether the decrease in fruiting capacity already iitoted in the fall-and-winter series, as compared with spring and summer -. ies, might be due mainly to lower light intensities during the fall and winter or to some other factors. ' 3' Table 3. Vegetative growth and fruiting of ten cotton varieties in the summer (OUIdOOIa) and in the late ' fall (greenhouse). Average of l4 plants of eazh variety. 1943 St.2B H&H Coker Wash. Roldo DPL Rog. Meb. Qual- Lone Diff 4-1 14 Est la __Type of data Season Rowd. Acala Star mi lfleight cm. Summer ...... . . 99 90 92 ss 91 9s 99 s9 92 99 9 Fall .......... . . s1 s9 ss s1 s9 9o 9o s2 9o s1 Ratio F/S, 17;, 90 99 99 92 9s 91 94 92 9s 91 ‘ ss ‘eightofstem Summer ...... .. 29s 241 25s 219 21s 25s 2ss s19 2s4 s20- ffmdleaves—— Fall .......... .. 1s2 119 1so 1s9 191 155 209 1so 1s2 214 i ' Ratio F/S, % 99 1s 10 9s 11 91 12 5s 94 91 1s.1 Summer ...... . . s4 s9 s9 s1 2s s9 s4 Fall .......... . . 2s so 2s 2s 2s 21 25 Ratio F/S, 9;, 9s ss 12 14 s2 99 14 91 1o 11 19.0 udlx’) %\I bllx’) \|O> CO KI Hi U! Q9 .. F‘ 0O O0 DP © CO 0O 5.6 30 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION E. loo _ suuuzn nu. A g eo- Z F“ f, so- E 4o- ; zo- g 375 - sumusa nu. B fisoo- 7 5225- w é o . g. 15o - a m- 15 - é 25 - 7 suuman FALL C f2‘ / Z g IO — Z Z 2 5 ‘ ‘ E 25- summer: FALL D i 20- v . g 1s- FT J F "7 j 1; 2 10- / 4 / - o |_ 2 ‘ , i ’ / / 6O_ SUMMER FALL E Q’ so— 7'7 g7 V) 7 é 7 3 40- ,4 a; Z LA 7' 7 Z g ZZj 4 1 A .4 A ‘L; |0~ 530m 7 z wages égu. F Ii4°<»//v é/e‘ y? 4 y a / g 200- Z 2 1A Z é A E IO\O—- ‘A I A Z x DPL. COK. l-LOH. 512B ROG. ROLD. WASH. MEB. QUALLA LONE l4 4'lN.-l ACALA ROWD. STAR Fig. 8. Graphical representation of difierences (shaded blocks, except in D) in growth and fruiting of ten varieties of cotton in the summer (outdoors) and in the fall (green- house). A. Height of plants; B. Fresh weight of stem and leaves; C. Number of bolls set per plant; D. Number of bolls shed per plant; E. Percentage of bolls set; and F. Fresh weight of bolls per plant. There were 14 plants (8 in sand and 6 in soil) of each variety per season. Note the relatively large differences between the summer and fall values in the case of bolls set (C), percent set (E), and weight of bolls (F), while smaller difierences exist in the height, (A) and weight of plants (B), and number of bolls shed (D). . , W"; FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 31 In these experiments, the plants were shaded with light-impervious airplane fabric. This was supported over the plants in the greenhouse by a pole hung over the center of the bench and the cloth was draped over the outside of the jars on both sides (fig. 9). Outdoors an inverted U-shaped structure was constructed over a row of plants, high enough to, allow one to walk alongside of the plants beneath the shade cloth.‘ The‘ ends of the shades were left open in every case to allow free circulation of air and t0 prevent. temperature increases under the shade. A light in- tensity of less than 50 f.c. was recorded on guard plants placed a few feet inside the ends of the cloth. In the center of these shades, with the: light-meter target registering light from any direction the intensity was less than 10 f.c., with bright sunshine on the outside. In the greenhouse experiment in the spring of 1944 six varieties of _cotton were used: Stoneville 2B, Rogers Acala, A. D. Mebane Estate, ;Deltapine 14, Qualla, and Half and Half. In this c-ase the treated plants iwere covered each afternoon at 3:00 o’clock (C.S.T.) and the shade was: iallowed to remain over the plants until 8:00 o’clock the following morn-- ling. This same treatment was repeated in a series of four varieties of ; fig. 9. Airplane fabric shade placed over cottoh plants during late afternoon and early mforenoon to study effect of variation m number of hours of sunshine per day. Close-spaced mlants are on the right. g - 32 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION cotton (two plants each of: Stoneville 2B, Rogers Acala, A. D. Mebane Estate, and Half and Half per treatment) during the summer. At the same- time, another treatment with a comparable‘ set of summer plants consisted of shading of the plants from 10:00 0’c1ock in the morning until 1:00 o’clock in the afternoon. All of the plants in these experiments were grown in 4-gal. jars of soil. ' It was desired to ascertain by these experiments whether decreased fruiting of cotton in the fall might be due to the lower light intensity or to the short-day effect at that time of the year. Consequently, these experiments might be classified as dealing with length of» day but since no attempt was made to exclude light totally by the manner in which the shading was applied, it seems preferable to consider the treatment merely as aldecrease in the number of hours of direct sunlight per day. This work is distinctly different from so-called photoperiodism experiments in which the length of day is usually studied in relation to flower bud forma- tion or to the period of blooming. As mentioned earlier, the cotton plant is not particularly sensitive to variations in length of day in regard to growth and to formation of flower buds. The results of these two experiments on variations in the number of hours of sunlight are given in table 4. In the greenhouse test, there was no significant difference between the check and shaded plants in regard to height an.d weight. The reduction in number and weight of bolls due to the treatment was highly significant. In the two outdoor experiments, there was a significant increase in height and weight of plants, while the number of bolls set and weight of bolls showed a significant decrease in each case. The reduction in number of hours of sunshine in every case resulted in decreases in the data associated with fruiting capacity. In this respect the results resemble those presented later for shaded and close- spaced plants. It may be noted that the shade applied for three hours near the middle of the day was apparently more inhibitive to the fruit- ing processes and induced more shedding than the shade treatment from 3 P. M. to 8 A. M., although these differences were not significant in most cases. Analysis of variance showed no significant interaction. between varieties and the short-day treatments. However, on a percentage basis (ratio be- tween short-day treated plants and the checks), the fruiting activities of the Lone Star, Qualla, Mebane, and Rogers Acala varieties were re- v duced to a greater extent than those of the Half and Half, Stoneville 2B, and Deltapine 14 varieties. This same trend in ratios for fruiting processes was found following each of the three experiments in table 4. In vegetative growth, these ratio differences were less marked, and the differences between the two above groups of varieties were not consistent. The shedding curves in fig. 10 also indicate a greater sensitivity to short-day effects on the part of the Qualla and D. Mebane Estate varieties than in the other four varieties under comparison. 0 Table 4. Eflect o_f reducing the total number’ of hours of sunshine per day on cotton plants. Data per average plant. A ‘ a l a Number Wei Pht Number Number Number Welght Percent Fruiting / ' Treatment " of plants Height of p ant of blooms bolls set bolls shed ‘of bolls set index - - . cm. g. _ g. ' Planted February 3, 1944—Shade, March.30 to May 10——6 varieties——-greenhouse No shade as 97 ' 294 l I s2 f 14 I 1s 29s 44 1.0 Shade—3 P. M. to 8 A. M. ll 36 94 253" I 21 I 8 t 13 162 38 .6 _ Planted April 6, l_94_4—Shade, June 26 to July 25—-4 varieties——0utdoors No shade I ' 8 79 231 I 37 i .16 i 21 361 43 1.6 Shade—3 P‘. M. to 8 A. M. 8 88 285 29_ 11 18 240 38 .8 Shade——10 A. M. to 1 P. M. 8 99 278 32 . 9 23 185 28 .7 [Riel l-l a l-i Z c: n» z, c: m "l: :11 c: u l-i Z ca o u: Q o a a o t , Z l-l Z w :4 a :> a i-l . o z I. a o . r 5 t: "3 34 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION The effects obtained from these shade treatments suggest that high light intensity is not as important as the length of the period of adequate illumination insofar as fruiting processes in cotton are concerned. In other words, the same "detrimental effects on fruiting in cotton as seen in the low-light-intensity days of fall and winter can be duplicated during seasons of high light intensity by merely shortening the daily period of illumination. This conclusion appears justifiable since the treated plants, in the summer series at least, were exposed to high light intensities for 7 hours in one case (early morning and afternoon shade) and for about 10 hours in case of the noonday shade. Furthermore, these periods of high intensity correspond rather closely with the 7 to 10 hours of sunshine available to the plants during the fall-winter series. As shown in table 1, the light intensity on the plants in the summer was con- siderably higher than that during the fall and winter months. The tempera- ture factor again seems to be much. less important than the light effects, since here both the spring and summer treatments, which produced effects similar to those obtained in the fall, were applied under temperature» con- ditions even more favorable for cotton fruiting than those prevalent in the greenhouse during the shorter fall-winter days. Elfects of Supple-mental Lighting on Cloudy Days During the late fall and winter, 1941-1942, thirty-two Rogers Acala plants (planted September 4) were grown in soil on a north greenhouse bench along the center of which 300-watt Mazda Reflector Flood electric’ lamps» in 18-inch porcelain reflectors were suspended at 8-foot intervals (fig. 11). The plants were arranged alofig each side of the bench in two rows so that four plants received illumination from each lamp and a group of four plants between these received practically no light from the lamps. ..llllllllllllllllllllllllllllllllllIlla U) O | s2a¢--o- R.A.-<=——o-m1-:s.-<——-+o.1>.1..-——-o—<>—<>-11.a 11.-<1—-<;- - o1 O ' I l Numezn 01-‘ BOLLS snso 0| -b O O l l l 2o - - 1o — - O 1 1 1 1 1 1 l- ' 1 1 1 1 1 1 1 1 1 1 1 1 \r _-_._ IO l2 l4 l6 8 2O 22 24 26 2B 3O 2 l6 APRIL MAY 1944 Fig. 10. Shedding curves for six varieties of cotton under short-day conditions (heavy shade from 3 P. M. to 8 A. M.). Note relatively heavy shedding by the Qualla and A. D. Mebane Estate varieties early in the fruiting period. Compare this figure with fig. 6-B (check plants for this same series) and note relatively greater amount of shedding by all varieties under short-day conditions, in the first half . of this period. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 35 The lights were used on all or parts of 23 different days, whenever the sky became completely overcast, beginning about the middle of November. On a clear afternoon (2:30 P. M.) in December the average of several light-intensity readings over this bench was 5200' f.c. At the same hour on a cloudy day, this average was about 250 f.c. between the lamps, while readings taken seventeen inches below the lighted lamps showed about 1000 f.c. intensity. " The plants were harvested on January 2.2 and considerable difference » was found between the groups (illuminated and nonilluminated) in vegeta- tive growth. In fruiting behavior, however, there was but little difference between the two groups. The plants that received the extra light on cloudy days were taller and averaged 102 cm. in height as compared with 84 cm. for the nonilluminated plants. This difference in height was due Fig. 11. Effects of supplemental light on cloudy days on growth of cotton plants under winter conditions in the greenhouse. The plants directly beneath the lamps made much more growth in height. to increases in length of internodal spaces, however, since both groups of plants had 20 nodes to the stem on the average. The average illuminated plant had 53 expanded leaves while the check plants had 49 leaves. The average number of bolls per plant was practically the same for the" illumi- nated plants (7.7 bolls) as for the check plants (6.9 bolls). In both groups, 40 percent of the flowers formed mature bolls. A significant difference between these two groups of plants was found in the fresh weight of bolls at harvest; each illuminated plant had a fresh boll weight of 220 g.’ per plant. whereas the weight for the average nonilluminated plant was 187 g. This increase in weight of bolls has also» been found associated with the better light conditions in other experiments reported herein. It seems, however, that the additional light provided by the electric lamp was not 36 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION ‘sufficient in this case to increase materially the number of bolls set or to change the rate of shedding, but the height of the plants and the weight of bolls were markedly increased by the supplemental illumi- nation on cloudy days. ' EXPERIMENTS WITH LOW LIGHT INTENSITIES Most of the studies reported here on the effects of artificially-produced low light intensities on shedding and fruiting in the cotton plant have involved brief periods of treatment lasting usually from two to eight days. Such treatment-periods have been used in an effort to reproduce approxi- mately the time intervals involved in ordinary periods of cloudy Weather. The materials used for shading have, in most cases, reduced the light in- tensity from around 10,000 to 12,000 f.c. at midday to 50 to 3000 f.c. on the shaded plants at the same time. These low light intensities are within the ranges of intensity measurements made during overcast Weather as shown in table 1. Other experiments have involved the growing of cotton plants for a period of several weeks, during the fruit- ing period, under constant_ low-light-intensity conditions (1000 to 4000 f.c. noonday sunlight), that might simulate continuously cloudy weather. ' Continuous Low Light Intensity During the Fruiting Period Four separate tests were made of the effects of continuous periods of shade on the fruiting process in cotton. The first of ,these tests was conducted during the summer of 1941 with sixteen Rogers Acala plants which were kept in a section of the greenhouse that was covered with slat shades from the time the first blooms appeared until the oldest bolls started to “open. On the average, these plants each produced 5 bolls and each shed 8 immature bolls, making a boll-set of 38 percent. The plants were exceedingly tall with long internodes. Check plants outdoors produced 24 bolls with 19 bolls shed, or 56 percent-set. This experiment was re- peated in the summer of 1944 using 18 plants (3 each of 6 varieties: Stone- ville 2B, Rogers Acala, A. D. Mebane Estate, Deltapine 14, Qualla, and Half and Half) in 4-gal. jars of soil, with the results shown in the first part of table 5. The shaded plants were significantly taller and heavier than the outdoor checks, while there was a significant reduction in number of bolls set and in weight of bolls as a result of the shade. Photographs of one of these plants may be seen in fig. 18. In order to avoid any possible heat effects from greenhouse tempera- tures during the summer, a third test was set up outdoors with 14 plants (8 of Stoneville 2B and 6 of Rogers Acala variety), also in 4-gal. jars of soil, during the late summer, 1944. In this case the plants were shaded by placing them at time of first blooming, under a slatted greenhouse roof shade supplemented with black cloth (fig. 12). The light intensity under this shade varied from 1000 to 2000 f.c. at noon on clear days, de- pending upon the direction in which the light-meter target was held. Temperature records showed no increase over those taken in the shade FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 37 among the control plants. The results given in the second part of table 5 were similar to those obtained earlier in the shaded greenhouse. The shaded plants were significantly taller although not quite as heavy as the checks. There was a significant reduction in the number of bolls set and in weight of bolls and the percent-set Was much less for the shaded plants. Another experiment, conducted in the greenhouse in the spring of 1944, involved the use of white cloth shade over 16 plants (8 of Stone- ville 2B and 8 of Rogers Acala) in sand culture. The plants were shaded continuously beginning with the opening of the first blooms. On clear days at noon, the light intensity on the upper leaves of the plants under this white cloth shade was about 2000 f.c.~The results of this ex- periment are given in the last section of table 5. ' These experiments with cotton plants under conditions of continuously low light intensity indicate that the fruiting processes are considerably inhibited by the reduced amount of light. These results might be in- terpreted as indicative of the fruiting behavior of cotton during an ex- tended period of continuously cloudy weather. Even With the low light intensities under the shade in these experiments, it should be noted that the height and fresh weight data for the continuously shaded plants were >- Fig. 12. Greenhouse slat shades, supplemented with black cloth for shading plants out- doors in the summer. The temperature readings beneath this shade did not exceed those taken at the same time among the lower leaves of nearby plants exposed to the sun. Table 5. Efiects of continuous shade throughout the fruiting ‘period on growth and fruiting of cotton. Average per plant l 2,000 101 224 ‘ 19 l 8 Approx. v _ noon Height Wei ht Number Number Number Weight Percent mtenslty cm. of p ant of blooms of bolls of bolls of bolls set f. c. g. set shed g. Under greenhouse shades, 18 plants per treatment, early summer 1944 I Check | 14,000 II 81 ll 228 ll 35 i 15 20 344 43 Shade I 2 , 000 I 128 320 I 36 i 9 27 128 25 Outdoor shade, 7 plants per treatment, late summer 1944 I I Check l 13,000 | 86 ll 263 || 35 a 17 18 374 49 - Shade ’ 1,500 I 112 i 228 I , 21 I 3 18 31 14 White cloth shade, greenhouse, 16 plants per treatment, spring 1944 I I I l Check | 1O , 500 I 81 l 208 | 24 H 1 2 12 248 50 Shade 11 163 42 88 NOLLVLS LNEINIHHJXH TVHILIIIIIOIHDV SVXHL ‘L219 'ONI NLLEFYIIIH FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 39 greater than those for the controls, with the single exception of the fresh weight of the outdoor series (heaviest shade). In one series (early summer, 1944) it may be noted that the number of open flowers was the same for the shaded plants as for the check plants. There was a differ- ence of 5 blooms per plant in the greenhouse series when white cloth shade was used. The greatest difference in number of blooms (14 per plant) occured in the outdoor series where the heaviest shade of all of the treatments was used. The difference in shedding between the shaded plants and the controls is quite marked as shown by the percent-set figures. The greatest dif- ference (49 vs. 14 percent) occurred with the heaviest shade (1500 f.c.) in the outdoor series. The reduction in fresh weight of bolls per plant following the shade treatments was severe in each case. The greatest dif- frence (343 g. per plant) was associated with the smallest number of bolls and occurred under the heaviest shade. Increased shedding in all three series was associated with the experimental shading and the excessive shedding that was induced could account for the small number of bolls and, at least in part, for the inferior weight of bolls on the shaded plants. Periods of Low Light Intensities Evidence of severe shedding following a period of low light intensity was obtained early in this work when two cotton plants, demonstrating the effects of high and low nitrogen applications in sand culture, were ex- hibited for two days in a hotel lobby at Waco, during the Second Cotton Research Congress, June, 1941. When the plants were returned to their former places outdoors at College Station, all young bolls, flowers, and squares of all sizes had been shed. In order to ascertain experimentally the effects of such periods of a few days of low light intensity, which also might accompany spells of cloudy or rainy weather, cotton plants were shaded at definite intervals and the differences in subsequent growth, shed- ding, and fruiting were noted. The first planned experiment of this sort (16) was conducted in the summer of 1941 when eight Rogers Acala plants (started April 23) were shaded outdoors with black cloth for four days (July 8-12) after several blossoms had opened on each plant. On August 3, these plants had on the average 12 bolls per plant and they had shed 24 fruiting forms. The nonshaded control plants at the same time had, on the average, 23 bolls and only 18 scars where forms had been abscised. The percent-set for the shaded plants was only 33 while the corresponding figure for the check plants was 56. The shedding curves for these two sets of plants are shown in fig. 13. Here it may be seen that a peak in shedding for the shaded plants occurred on July 12—the day the shades were removed. This was followed by high points in» the shedding curve for the shaded plants on July 15 and 17. The control plants also showed an increase in the shedding rate on July 15 and 17 apparently in response to cloudy weather starting on July 10 and lasting until July 19. This peak in the check-plant curve, however, is not as high as the curve for the 40 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION shaded plants at this point, although the latter plants had already shed many more forms than the checks. The next high shedding rate for the check plants started on July 25 and lasted until the time of harvest on August 3. These final peaks may have been associated with several days of cloudy Weather from July 13 to. 19 followed by another cloudy period starting on July 24. The exceptionally high peak of shedding on July 25 for the plants that were shaded earlier is of considerable interest since it shows increased shedding activity in plants that had already shed heavily. Moreover, this second shedding peak occurred somewhat earlier than the corresponding peak for the control plants. This indicates that conditions which cause a cotton plant to show an abnormally high shedding rate im- mediately might also predispose the plant to more serious shedding later and make ‘it more sensitive to subsequent shedding-inducing stimuli. ZQEEEIEVAIIIIV/AZI§3EE§EEE%%IIZ ~ -¢ 48- - o _ _ lil42- - (D '-— TIME OF - SHADE __ s : " m30_ 2' 324.. _ é us: : gr '-' 0-1 s- J/ / q 111111 /|||||||||1|||||\/||1|1|111 27 29 "l 3 5 7 9 I l l3 l5 I7 l9 2| 23 25 27 29 3| 2 JUNE JULY |94| sue Fig. 13. Shedding curves showing effects of 4-days shade, applied July 8-12, to 8 Rogers Acala plants. Note immediate eifect of shading shown by increase in shedding on July 12 as compared with the curve for 8 nonshaded plants (check). Another sharp peak of shedding for the shaded plants on July 25, apparently caused by the cloudy weather July 10-19, is also of interest. This experiment was followed by one in the greenhouse in the spring of 1942 in which three sets of 8 Stoneville 2B cotton plants each, planted January 9, were used. One set served as checks while the second was shaded with white cloth on Monday, Wednesday, and Friday of each week from March 23 to May 5 and the third set was shaded with black cloth on these same days. The plants were harvested on May 8 and the following numbers of bolls per average plant were obtained: checks, 21; white cloth shade, 18; black cloth shade, 15. The average weights of bolls per plant were 411, 312, and 249 g., respectively. There was but little difference in the percent-set figures for the three groups. A Following these preliminary tests, several experiments on shading of cotton were conducted under both greenhouse and outdoor conditions. The results of five experiments, including the most significant of these studies ;rr"n 1 s-vvrcp =Y 7fwv _,,. ~@V‘.-'I .1 a“, , per treatment. FRUITING AND SHEDYDING OF COTTON IN RELATION TO LIGHT 41 I are given in table 6. With the exception of the test in the spring of 1944, l in which 4-gal. jars of soil were used, the plants were all grown by the sand culture method. Two of the experiments were in the greenhouse (spring series) and the three summer tests were made outdoors. Black cloth was used for shading in all cases although in the late summer 1944 experiment its use was supplemented by the slat shades, fig. 12. In the second summer series (planted June 25, 1942) in table 6, half of the plants were subjected to a second period of shade by placing them in a laboratory room for six days, following an earlier period under black cloth. The - first column of figures in this table shows the light intensities on the I shaded and nonshaded plants near noon on a clear day. In the first experiment (1943) shown in table 6, 16 plants each of Stone- ville 2B and Rogers Acala varieties were used—4 plants of each variety The first shade treatment consisted of black cloth shade placed over the plants on Monday, Wednesday, and Friday of each of the first three weeks in April; the second treatment involved two 4-day periods of shade beginning on April 3 and 24, respectively; and the third treat- sment was one of 6 consecutive days of shade beginning April 13. Sta- tistical analysis of the results of this greenhouse test shows that the differences in height were not significant although each of the treatments showed a taller average plant than the checks. This same condition was found in the data on fresh weight of plants. As compared with the checks, » all three shade treatments produced smaller numbers of bolls per plant , (significant at the 1 percent level). There was no significant difference between the shade treatments, however, in regard to number of bolls. The number of forms shed in all three shade treatments was significantly Iidiiferent (at the 1 percent level) fromthe checks, and a significant dif- » ference between the 3-days per week and other treatments was also found . in this case. Again, all three shade treatments produced a significantly (1 percent level) smaller weight of bolls than the checks, but here the diiferences between the shade treatments were not significant. The per- cent-set figures show a sharp reduction in case of the treated plants. The second part of table 6 is comprised of data taken under greenhouse conditions from 6 plants of each of 6 varieties in the case of the checks l, and from 5 plants of the same varieties (Stoneville 2B, Rogers Acala, A. D. ‘ Mebane Estate, Deltapine 14, Qualla, and Half and Half) in case of the 11-day shade treatment. The increases in height and fresh weight of the I shaded plants were highly significant in this experiment as were the de- creases in number and weight of bolls. Inasmuch as the period of treat- ment was relatively brief, it may be assumed that these vegetative dif- ‘pferences were due to increased growth following the loss of fruiting ;fstructures by shedding rather than to an etiolation effect during the period of low light intensity. The data on shedding, 22 forms from the checks and 26 from the treated plants, are not comparable since nearly all the shedding "recorded for the shaded plants occurred during or just after the treatment and many squares of various sizes were involved. The shedding curve (not f shown) for this experiment had a sharp pealc for all varieties on April Table 6. Effects of few-day periods of shade-on grolwth and fruiting of cotton Approx. l N b Average data per plant noon um er * _ light plants l _ ll Weight I Number | Number | _ | s Treatment 1ntens1ty,l per | Helght stem and | of bolls | of forms | Werght | Percent f. c. treatment cm. leaves set shed of bolls set 8- 8- w C! . - . E.‘ Spring senes_ln Greenhouse Planted January 9 1943——Harvested May 7 E . i ,—‘ Z None l 9 , 800 8 l 97 l 308 l 18 l 13 l 384 58 z Shade-Mon“ Wed, Fri., 4 /5-23 350 8 I 105 | 324 l 9 | 20 l 179 31 o Shade 2 periods (4 da.) 4/3-7 & 4/24—28 350 l 8 I 116 | 365 I 7 | 25 l 156 22 ' Shade 1 ‘period (6 da.) 4/13-19 . 350 l 8 l 113 l 323 7 l 27 l 142 21 g ~ a Planted February 3, 1944——Harvested May 18 a ' > None l 10,800 l 36 l 97 294 14 l 22 l 293 39 w Shade——4 days, 4/10-14 l 300 l 30 l 115 413 4 l 26 l 8O 13 I> C) E Summer Series—-—Outdoors Q ‘ Planted May 12, 1942\—Harvested August 24 a . | '4 None 1 3 , 400 l 6 | 96 l 248 16 l 26 l 305 l 38 g Shade, 7/21-25 and 7/31-8/3 450 l 6 l 118 307 7 25 l 193 l 22 f> F‘ B1 Planted June 25, 1942—Harvested October 19 f; H None l 14,000 l 8 l 80 235 l 18 13 l 386 l 58 a Shade, 9/11-13, in lab. room, 9/18-24 l 500-50 l 8 l 95 290 l 6 20 l 135 l 23 g v Z Planted June 24, 1944—Harvested September 15 a m 1 .500 86 263 49 g 30s g Z t k E' F; i. Li. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 43 17 following removal of the shade on April 14. About one-half as many shed forms were recorded on April 15 as on the 17th and many of these had dropped before removal of the shade. Following this peak the shedding fell off abruptly to a low rate on April 19 and 23, after which there was practically no further shedding up to the time of harvest of the plants (May 15). The shedding graph for the check plants in this same experi- ment is shown in fig. 6-B, in which peaks occur about April 26-28 and May 3. In this experiment, again, the 13 percent-set for the shaded plants is relatively low and amounts to only one-third of the corresponding figure for the check plants. The first of the three outdoor summer series shown in table 6 was made up of 12 Rogers Acala plants. The treatment in this case consisted of black cloth shade for 4 days beginning on April 10. The differences obtained be- tween the checks and shaded plants in regard to height and number of forms shed are not significant. However, the increase in weight of plants and decrease in weight of bolls were found to be significantat the 5 per- cent level. The reduction in number of bolls is highly significant. There was practically no increase in the number of forms shed but the smaller percent-set in the case of the shaded plants is in keeping with the other experiments in this table. The results of the second outdoor shading experiment, conducted with a late summer series of plants in 1942, are shown in the fourth section of table 6. In this case 8 plants of the Stoneville 2B variety were used as checks and 8 comparable plants were shaded outdoors for 2 days and later placed in a laboratory room for 6 days. The difference shown in height measurements is significant at the 5 percent level while the differ- l ences in weight of plants, number of bolls set, number of forms shed, and weight of bolls are all significant at the 1 percent level. The usual reduc- tion in percent of bolls set is evident in this test. The last experiment listed in table 6 was conducted outdoors in the late summer of 1944 with 4 plants of Stoneville 2B and 3 of Rogers Acala variety per treatment. The treatments in this case consisted of placing the plants under the slat and black-cloth shade shown in fig. 12. The first three treatments in this experiment, as listed in the last section of table 6, consisted of 2, 4, and 6 days of shade, respectively, beginning on August 22. The fourth shade treatment consisted of two periods of 3 days each, beginning on' August 22 and September 8, respectively. The last two treatments consisted of 4 days under shade, the earlier treat- I ment beginning on August 28 and the later on September 4. The differences in height of plants resulting from the shade treatments were not significant when analyzed. In the case of weight of plants, number of bolls set, num- ber of forms shed, and weight of bolls, however, a significant difference at the 1 percent level was found between the treatments and the Qhecks. ‘ The differences required for significance (5 percent level) in each of these cases are given below the respective figures. It will be seen that the treat- 44 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION ments consisting of 6 days under shade and of two shade periods of 3 days each produced values significantly different from the checks in each of the four measurements and counts just noted above. These two treat- ments also produced the greatest reductions in percentage of bolls set, in this last experiment. The percent-set figures for the shaded plants, in all experiments in this table, are lower than the percentages shown for the corresponding control plants. The greatest reduction (66 2/3%) in percentage of bolls set oc- curred in the spring- series in 1944. This is in keeping with many obser- vations in this work that indicate a greater sensitivity to shade in the fall or spring series than in the summer series. The fall and spring plants have also shown a greater tendency towards vegetative growth (see fig. 23-D), probably on account of lower reserves of carbohydrate food ma- terials. As pointed out by Shirley (48), many kinds of plants attain maxi- Fig. 14. Increased vegetative growth and the forcing of adventitious flower buds as a result of shading. A. and B. Photographs of two Stoneville 2B cotton plants, planted April 3, 1943; the plant on the left was shaded for 4 days (June 3-7). In B, all foliage except the youngest leaves have been removed. Note the larger number of new leaves, blooms, and nodes, where shedding has taken place, on the shaded plant, which is also larger than the unshaded check (on right). C. Adventitious bolls formed in the axils of leaves on the fruit- ing arms of the shaded plant. The old scars may be seen where the original bolls were shed. Photographs taken July 16. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 45 um vegetative growth at about 25 to 50 percent of normal summer sun- t in temperate regions, while maximum fruitfulness and dry weight ur at higher light intensities. The increase in vegetative growth and her effects of shading may be seen in fig. 14. It may be noted in the last section of table 6 that the shortest shade l tments (2 days) produced a slight increase in number and weight of __lls. Although these brief periods of shade were sufficient to cause a i‘ k in the shedding rate, like the one shown for 3 days of shade in fig. B, the effects were not serious enough to cause a final decrease in num- or weight of bolls at time of harvest. The data on the 1944 summer series of plants are of special interest ause of the comparison of periods of shade of varying duration and at erent times in the fruiting period. There is a definite gradation of ects from the check plants through the 2-day, 4-day, and 6-day periods shade. It is noteworthy, also, that the ,6 days of’ shade provided at different times (3 days each, beginning August 22 and September 8) uced a more serious effect on the fruiting behavior of the plants than single peroid of 6 days (beginning August 22). A similar effect can be n in the second shade treatment applied to the spring plants in 1943, shown in the first section of the table. These data add further evidence the marked effect of a second period of low light intensity following a vious period by several days. Such an effect has been shown earlier "fig. 13. It is also of interest to compare the effects of the three 4-day- de treatments in the last section of table 6. Here, the most apparent action in fruiting took place in the group of plants that was shaded liest (August 22). When shade was applied on August 28 and on Sep- ber 4 less decrease in the fruiting activity of the plants resulted. a results are in accord with those reported later where different groups during the fruiting period. arietal diflerences in response to shading. The results of two sand- i re experiments, in which comparisons were made of the responses of i‘? ent varieties of cotton to a few days of low light intensity during ' ruiting period, are given in table 7. Both of these experiments were er sensitiveness to shading than those growing outdoors in the sum- time. The data in the first part of this table show the Rogers Acala ty of cotton to be more adversely affected in its fruiting activity ‘by treatment applied than the Stoneville 2B cotton. At the same time, L, Acala variety slightly exceeded the Stoneville 2B in vegetative h under both the control and shade-treatment conditions. In this ex- ‘ent the shade treatment produced, in both varieties, a highly sig- nt effect in the case of height and weight of plants, number of bolls " number of forms shed, and in weight of bolls. In the case of the ht of bolls, a highly significant difference was found in the interaction rieties and treatment. This indicates a varietal difference in the fruit- responses obtained in reaction to the shade. plants from the same planting were placed under shade at different p cted in the greenhouse, where plants in the fall or spring have shown . 46 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Table 7. Varietal difierences in response of cotton to periods of shade. Average per plant. _ _ Weight Number Number Wei ht Per Variety Treatment Height 0 bolls bolls o? cent cm. plan set shed bolls set 8- 8- Planted Feb. 4., 1944—-shaded 3 days, 4/22-25——8 plants of each var. per treatment Stoneville 2B ‘Check . . . . . . . . . 75 195 14 12 281 54 Shade . . . . . . . . . . 97 271 7 7 185 50 Ratio, S/C1.... . 129 139 50 58 66 ‘ 93 Rogers Acala Check. . s’. . . . . . 87 220 10 11 215 48 . Shade. .. . . . .. 105 287 2 7 54 22 Ratio, S/C. . . . . 121 130 » 20 64 25 46 Planted Feb. 3, 1944—shaded 4 days 4/l0—14—check, 6 plants; shaded, 5 plants i Stoneville 2B Check. . . . . . . . . . 97 310 19 21 410 48 Shade . . . . . . . . . . 107 406 7 29 142 19 Ratio, S/C. . . . . 110 131 37 138 35 40 Half and Half Check . . . . . . . . .. 94 301 19 23 358 45 Shade . . . . . . . . . . 117 430 6 27 _ 100 18 Ratio, S/C. . . .. 124 143 32 117 28 40 Deltapine‘ 14 Check . . . . . . . . . . 100 273 15 24 300 38 Shade . . . . . . . . . . 127 414 5 26 98 16 Ratio, S/C. . . . . 127 152 33 108 33 42 Qualla Check . . . . . . . . . . 90 260 10 23 241 30 Shade, . . . . . . . . . 110 336 4 24 88 14 Ratio, S/C. '. . . . 122 129 40 104 37 47 Mebane Estate Check . . . . . . . . . . 93 275 8 23 223 26 Shade . . . . . . . . . . 111 384 2 26 42 7 Ratio, S/C. . . . . 119 140 25 113 19 27 Rogers Acala Check . . . . . . . . . . 106 343 12 19 223 39 Shade . . . . . . . . . . 123 508 1 22 12 4 Ratio, S/C. . . . . 116 148 8 116 5 10 lAverage for shaded plants divided by corresponding figure for the checks and expressed as percentages. The other experiment was conducted in the greenhouse during the spring, V 1944. Here, six-varieties of cotton were compared in their response to a single 4-day period of shade and the results are shown in the last part of table 7. At that time of the year, the check plants were subjected to somewhat less favorable conditions for fruiting than during the summer, as shown in table 2. The six varieties are arranged according to the num- . ber of bolls produced by the shaded plants. It may also be noted, ac- , cording to the ratios shown between the checks and shaded plants, that , FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT‘ 47 . the varieties listed first, in general, show a better maintenance of favor- able fruiting following the shading, than the varieties near the bottom of the table. At the same time, the vegetative growth of all varieties was increased to a fairly uniform extent by the shade treatment. Photo- graphs of the check and shaded plants in this experiment may be seen in fig. 23-A and D. In addition to the significance already mentioned for these data, in connection with the second section of table 6, a significant difference was found in the interaction of varieties and shading treatment in regard to the weight of bolls produced. It is of interest to note that the varieties listed first in the 'last part of this table, namely, Stoneville 2B, Half and Half, and Deltapine 14, are the same varieties that have been shown to be less seriously afiected by fall-winter growing conditions (table 3) and by reduced day length. It should be stated that differences between varieties in experiments shown in other sections of table 6 (be- sides the second section) have also been found on a percentage-ratio basis. In these cases, the Stoneville 2B variety has shown higher percentages in regard to fruiting performance after shading‘ treatment than the Rogers Acala variety. Owing to the relatively small number of plants used of each variety, however, and the tendency.for all varieties to show a relative- ly wide variation between individual plants following shading treatment, no significance was foundamong the variety-treatment interactions. The differences obtained are not shown in tabular form. 4Q l ' Q Fig. 15. The center row of Coker 4-in-1 cotton was shaded with black cloth for 4 days when the first bolls were small. The shade caused considerable shedding and a renewal of vegetative growth. The photograph was taken 5 weeks after shading, following a long period of drought. The foliage of the shaded plants was still green and much heavier than that on the nonshaded rows on either side. 48 BULLETIN NO‘. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Efiects of shading plants in the field. On July 15, 1943, black cloth shades were placed over five 25-foot single row plots of Coker 4-in-1 cotton at the Main Station Farm. The average light intensity beneath the shade was about 600 f.c. when the outside measurement was 16,000 foot candles. The plants on this date were flowering and some of the oldest bolls were about one-fourth grown. The plants were shaded for 4 days. Two days after the removal of the shades, an average of 243 young bolls and squares were recorded from beneath each of the five shaded plots, while the average number for the check plots was only 60 bolls and squares per plot. This period of shading was followed by a severe drought which caused excessive shedding of bolls on all plots .and no significant differences between the shaded and check plots were found in the yield of seed cotton at time of harvest. One of the outstanding results of the brief shading was the much more abundant foliage on these plants throughout the period of drought (see fig. 15). This would indicate a marked effect of the shading in addition to the dropping offorms. Possibly new vegetative processes were initiated by the shading that was maintained during the drought period. Such a response suggests the possible serious effects (even under favorable growing conditions) that might follow a period of cloudy weather of sufficiently low light intensity to cause excessive shedding. Age of Plants and Fruiting Forms in Relation to Shedding In the summer of 1943, a group of sixty cotton plants comprised of six plants of each of ten varieties growing in 2-gal. jars of soil was divided into six groups of ten plants (one plant of each variety: Stone- ville 2B, Rogers Acala No. 111, A. D. Mebane Estate, Deltapine 14, Roldo Rowden, Qualla, Coker 4-in-1, Washington, Lone Star,. and Half and Half). One of these groups was kept outdoors continuously as a check. The other groups were placed in a laboratory room (noonday light intensity 50-75'f.c.) for four days, in a consecutive order as shown in table 8. The first group had been in flower about ten days when placed indoors (on July 10). The different types of forms (small squares, large squares, flowers, small bolls, and large bolls) that were shed were recorded separately each day. At the end of this 4-day treatment most of the fruit- ing forms had been shed from the plants, which were then returned to their _ former outdoor positions. The plants that received no indoor shade produced the best yield of bolls and the highest percent-set and they also made the least vegetative growth (weight of stem and leaves) of all the groups. In general, the groups that received the shade treatment early in the fruiting period were adversely affected to a greater extent, in regard to final crop of bolls and percent-set figures, than the groups that were shaded later. An exception to this conclusion may be noted in the second group of plants (shaded July 10-14) which showed a better fruiting record than the third or fourth groups. This was due for the most part to the fact that new squares - FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 49 Table 8. Effect of age of plant on the response of cotton to a 4-day period of shade. Average of l0 plants; planted May 14, 1943 Date of shade 0n Height olfvegi allrlt ljfurbdlollllsr ldfulilotlllsl- (ill, e1i>§iil Percent different groups em. g. set shed g. set N0 shade 69 148 9 19 158 32 July 10-14 87 258 6 29 96 i 17 July 14-18 103 319 1 24 14 4 July 18-22 92 213 3 22 4O 12 July 25-29 76 163 5 22 81 19 July 29-Aug. 2 81 177 7 25 133 22 Diff. req. at5% 12.8 99.4 2.7 5.2 14.5 . . . . . . . . .. developed on the plants following the shading of this early group and these had suflicient time to produce bolls large enough to be harvested on Au- gust 20. The most serious effects from shade were obtained in the second shaded group (July 14-18), as shown by the decrease in fruiting and by the increase in vegetative growth. Later treatments on other plants of the same age produced effects correspondingly less damaging to the fruiting processes. ' The data in table 9 show the relative susceptibility of squares of dif- ferent ages, flowers, and bolls of different ages to shedding caused by un- favorable light conditions. In keeping with many recorded observations of shedding under field conditions, the small bolls (just after flowering, fig. 1-D) were the first forms to be shed and the heaviest shedding of those young bolls took place on the second day following placement of the plants indoors. With the older plants (groups shaded later) it may be noted that a number of large squares and also of small squares were shed on the second day indoors. On the third day, the small bolls were again shed most severely by the first two groups. In case of the last three groups, the shedding of large squares (fig. 1-B) exceeded that of the small bolls and a few small squares were also shed (fig. 1-A) at this time. By the fourth day only a few small bolls and large squares remained on the plants and the large squares were more numerous than any other single type of fruiting form. At this time the small squares were shed in much greater abundance than on the two previous days. A few of the fair-sized bolls (about one-fourth grown, fig. 1-E) were shed on the third and fourth days under shade. In keeping with the frequent statement that shedding of upland cotton flowers rarely occurs, there were only a few forms with recently opened eorollas shed and most of these were dropped on the fourth day. The last column in table 9 shows a greater amount of shedding of all ages of forms by the older plants (last three groups). The smaller number of forms shed by the first two shaded groups was as- sociated with the smaller number of forms on the plants at that time. Practically all of the shedding in the check plants outdoors took place \ Table 9. Relative susceptibility of fruiting forms of difierent ages (squares, flowers, and bolls) to shedding due to 4-day shade treatment. Same plants as in table 8. Figures given are totals for l0 plants. Shed 2nd day Shed 3rd day Shed 4th day Total Date of shade for Small Large Flow- Small Lar e Small Large Flow- Small Lar e Small Large Flow- Small Lar e 3 days, sqs. sqs. ers bolls bol s sqs. sqs. ers bolls bpl s sqs. sqs. ers bolls bol s all forms 7/10-14 16 9 2 12 12 3 2 2 60 7/14-18 27 4 10 5 15 19 11 4 95 7/18-22 2 1 35 37 14 1 17 19 4 ‘ 7 4 154 7 /25-29 6 50 1 32 1 4 2 6 1 126 7/29-8/2 10 13 54 12 38 26 1 3 5 175 Total forms shed: Der day 16 21 1 182 1 28 113 1 63 10 56 63 8 29 16 3 days 100 197 1O 274 27 09 NOLLVLS LNEINIHEIJXH TVHILIIIHOIHDV SVXELIL ‘L119 "ON NLLEITIIIH FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 51 ~ between August 2 and 20 which was after the treatment was applied to _ the last group and the forms shed were mostly small bolls. This experiment indicates that young bolls, just after the flowers have dried up, are more sensitive to shedding following low-light-intensity _- treatment than fruiting forms of other ages. When light conditions become @ more unfavorable, however, other forms (squares, and larger bolls) may » be shed abundantly. These results are in keeping with observations made if during the different seasons in this work. During the summer, for ex- ample, few forms other than young bolls have been shed from the check a plants in the experiments. Under greenhouse conditions during the fall, winter, or spring, however, there has appeared at times a tendency to shed ._ a larger number of squares, especially very small squares five millimeters 4 or less in diameter (see smallest square in fig. 1). This shedding of small I squares was particularly marked in the soil plants grown during the winter, _ 1942-43. In this case, the average number of bolls (nine) per plant for ten varieties was rather low although the average percent-set was unusually 3' high (53%) for that time of the year, with shedding data including only 7 the young bolls. In many cases the entire growing point was shed along I with the small square which resulted in a determinate type of growth on j the fruiting branch. Sand-culture plants grown in the same greenhouse i at the same time gave a comparable low yield (seven bolls per plant, j average of eight plants each of ten varieties) but the 32 percent-set was *_ correspondingly low, showing that more shedding of small bolls and large squares took place. It is believed that the nutritional differences between 3" the soil and sand cultures, coupled with peculiar weather effects for that i season, may have resulted in this unusual shedding of small forms by the a * soil-grown plants. Careful records were kept of the shedding of these a small squares in the fall-winter and spring series 1943-44 for both soil and sand plants. Although considerable shedding of small squares took jlplace (up to 10 or 15 per plant in some cases) it did not appear to affect ‘ the yields appreciably. All of the six varieties in the test shed similar ~ numbers of these small forms, with the Half and Half variety showing a 1 slightly higher rate of shedding than the others. This shedding mostly ~ took place after a fair number of bolls had been set and, in actively-grow- ing plants, new flower buds promptly appeared on the next node. How- 1 ever, the fact that the summer-series plants shed few small squares would indicate that some factor associated with the fall-spring conditions in the s greenhouse was responsible for this loss of the smallest forms. Very few l of these forms (not more than 3 to 5 per plant), however, were shed from ’ the plants even under continuously low-light-intensity conditions in the summer series. Also, in the 1945 spring series, a few small and medium- l sized squares were shed from all plants in an experiment, following a con- ‘ siderable amount of shedding of small bolls in response to cloudy weather from February 16 to 20. The plants were in an early fruiting stage at . this time. The results of an early shading “test with twenty Stoneville 2B (cotton plants during the late fall and early winter of 1943 is also of interest in 52 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION ' connection with the shedding 0f small squares. The seed were planted on September 21 in 2-gal. jars of soil and the young plants were placed in the greenhouse on October 8. On November 20, five days before the first flower opened, ten of the plants were placed in a laboratory room for four days‘ under low-light-intensity conditions. The first small squares (largest 15 mm.) abscised on November 22 and by the time the plants were replaced in the greenhouse they had each shed on the average 4 small squares and 2 large squares. On this date each of the check plants had 14 squares (2 mm. or more in diameter) and two of the check plants had a single flower. At the same time the shaded group averaged six attached squares p_er plant. During the two days following removal from the shade, these (shaded) plants each shed 2 squares, most of which were small. No further shedding was recorded until December 11 when shed- ding, of a few (from all ten plants) small squares and small bolls took place on both the checks and the shaded plants. This initiation of shedding followed a period of eight cloudy days (December 3 to 10, inclusive). Following this period, all of the forms shed were small bolls and nearly all of them were from the check plants (6 per plant), with a total of only 3 small bolls shed from all ten of the shaded plants. At time of harvest on December 4, the shaded plants averaged 12 cm. taller than the checks (77 cm. vs. 65 cm.). The check plants each had 5 bolls that were past shedding stage and no other fruiting forms. There were no bolls on any of the shaded plants but the plants had an average of 7 squares of various sizes. This experiment shows the detrimental effect that the shedding of small squares due to low light intensity might have on the final set of cotton bolls, even though most of the shedding took place in advance of the so- called fruiting period. Also of interest is the fact that the shaded plants, even after the earlier heavy shedding, again started to shed at the same time as the check plants when unfavorable weather conditions occurred. Close Spacing of Plants As mentioned earlier, one of the methods used in this work for pro- viding variations in the intensity of light reaching the experimental plants consisted in different spacings between the plants growing in jars of sand or soil. By this means, the foliage of the plants spaced closely together, so that the jars touched one another, was subjected to much more shading from adjacent plants than in cases where the jars were spaced about two feet apart. At the same time, other conditions such as temperature, at- mospheric humidity, and the amount of available soil moisture and nutrients were practically identical for the close-spaced plants and the checks. In these experiments, the shading effect first became noticeable on the plants about the time of the appearance of the first blooms, when the leaves and branches of the close-spaced plants commenced to inter- mingle. As an indication of the density of shade provided by leaves of a cotton plant, it may be of interest to state that light-intensity readings taken among the lower leaves of plants on a greenhouse bench averaged I FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 53 around 600 f.c. while the direct sunlight intensity beside the bench Was‘ about 11,000 f.c. at noon on a clear day in April. Data obtained in six separate experiments involving close-spacing of cotton plants are given in table 10. In the first three of these experiments an equal number of Stoneville 2B and Rogers Acala plants were used, While in the last three tests six varieties, Stoneville 2B, Rogers Acala, A. D. Mebane Estate, Deltapine 14, Qualla, and Half and Half were in- cluded in each experiment. The first, third, and fourth tests were con- ducted in the greenhouse during spring months (fig. 9), While the second, fifth, and sixth experiments were conducted outdoors in the summer. The data show a significant increase in height of plants in all of the close- Table 10. Eflects of close spacing of plants (in jars) on growth and fruiting of cotton. Average per plant. I I I I I I I Wt. of I . I I Number leaves Number NumberI NumherI Weight Per Treatment ' of I Height and of o of of cent , plants cm. I stem blooms bolls bolls bolls set I I g. I set shed g. I I I I Planted January 27, l942—2-gal. sand—2 varieties I Check I 16 I 137 475 41 20 21 . . . . . . . . 49 Close spacedI 16 I 150 445 I 31 14 17 . . . . . . . . 45 Planted July 17, 1943——2-gal. sand—2 varieties I I ' Check I 16 92 242 3O I I 14 16 312 46 Close spaced I 16 102 133 24 I 8 16 193 33 Planted January 21, 1944—3-gal. soil——2 varieties I I Check I 32 I 93 I 269 I 28 I 12 I ' 16 I 253 I 43 Close spaced I 32 I 105 I 250 I 22 I 8 I 14 I 187 I 36 Planted February 3, 1944——4-gal. soil—6 varieties I Check 36 97 I 294 I 32 14 18 293 44 Close spaced 36 102 I 239 I 21 8 13 172 38 Planted April 5, l944——2-gal. sand—6 varieties I Check I l2 69 208 26 14 I 12 I 318 I 54 Close spaced I 12 79 209 27 l 12 I 15 I 265 I 44 Planted April 6, 1944—4-gal. soil—6 varieties Check 18 81 I 228 I 35 15 20 366 43 Close spaced 24 94 I 246 I 35 14 21 308 4O 54 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION spacing treatments. Unlike the data previously shown for shaded plants, the results here do not show a greater fresh weight of leaves and stems due to the close spacing. In some cases, somewhat fewer blooms per plant were produced under the close-spaced conditions. The differences shown in number of bolls set between the check and close-spaced plants are highly significant (1 percent level) in the first four experiments shown in table 10, but they are not significant in- the last two experiments. The difference in the number of bolls shed was found to be statistically significant at the 5 percent level only in case of the first and fourth experiments re- ported in table 10. The fresh weights of bolls on the check plants differed to a highly significant degree from those of the close-spaced plants in the second, third, and fourth experiments. The boll-weight differences in case of the last two experiments in table 10 were not great enough to show significance upon analysis. Percent-set figures for the close-spaced plants were consistently lower than those for the checks in all of the experi- ments listed in this table, although the amount of difference was not great in any case. It may be noted that the effects of the close spacing in the last two experiments (table 10) are less striking than thelresults obtained in the first four tests. This difference was probably due to the fact that the last two experiments were conducted outdoors in early summer while the others were carried on either in the spring or late summer, when the general light conditions were somewhat less favorable than in early sum- mer. The response of the July 17 planting resembled the spring series more closely than the early summer plants since blooming did not start un- til the middle. of September and the fruiting period occurred at a less favorable time of the year as far as light conditions were concerned. It is evident that‘the effect of the steady shading from close spacing was less detrimental to the fruiting processes and also gave less impetus to vegetative growth than the single brief periods of heavier artificial shade in the experiments just described. In connection with these close-spacing experiments, it is of interest to note that Cook (13) records the dropping or poor development of young bolls from cotton plants that were crowded in the field. On a percentage basis, the amount of reduction in fruiting associated with the close spacing was considerably greater in some varieties (Rogers Acala, A. D. Mebane Estate, and Qualla) than in the other varieties in the comparison (Stoneville 2B, Half and Half, and Deltapine 14). The fruit- ing index, for example, was reduced 63, 46, and 73 percent, respectively, for the first group of 3 varieties just mentioned; while the corresponding reductions for the second group of varieties, due to close spacing, were only 80, 89, and 84 percent, respectively. Upon analysis of the data, no positive evidence of interaction between varieties and the close-spacing treatment was found. FRUITING AND SHEDI-DING OF COTTON IN RELATION TO LIGHT 55 EFFECTS OF VARIATIONS IN SOIL MOISTURE Inadequate water supply to the cotton plant has been recognized ‘for some time as a contributing factor to excessive shedding and impairment of yield. However, it was considered advisable to conduct several experi- ments on wilting of plants to varying degrees for the purposeof com- paring wilting effects with those of low light intensity, in regard to the severity of shedding. Other experiments have been made in which the effects of wilting upon the uptake or utilization of available nitrogen were studied in relation to fruiting and shedding in cotton. Relatively Brief Periods 0f Wilting In the fall and early winter, 1939, an experiment was completed in which forty cotton plants (var. Startex) were grown in sand in four dif- ferent sizes of containers, six-inch flower pots, and 1-, 2-, and 3-gal. jars. Approximately 2 quarts of nutrient solution were supplied daily to each plant by a slow drip-culture method at such a rate that the solution was continuously supplied throughout the forenoon. During the afternoon the plants were forced to draw upon the moisture reserve in the container which in the two smaller sizes was usually insufficient to prevent wilting on a bright day. The wilted plants were allowed to remain in that con- dition for about one hour before the sand was flushed with plain water. Late in the afternoon the sand in all of the containers was flushed with water. In this manner an attempt was made to provide all of the plants with approximately similar amounts of nutrients, regardless of the size of the container. At the conclusion of the experiment, the average heights of the plants in the six-inch pots, 1-, 2-, and 3-gal. jars, were 90, 110, 121, and 124 cm., respectively. The fresh weights of the plants without the bolls were 151, 215, 225, and 257 g. inthe same order. In spite of these marked differences in size of plants there was but little difference in number’ of bolls per plant (4, 6, 5, and 6, respectively) or in the percent-set (21, 28, 33 and 28 percent, respectively). Although the lowest percent-set occurred with the smallest containers (six-inch pots), the difference between the 1-gal. jars and the 2- or 3-gal. jars was not significant even though the plants in the 1-gal. containers often became wilted in the afternoon. It seemed, there- fore, that the rather frequent wilting of the plants in the small con- tainers for short periods of time was not sufficient to cause marked shed- ding of fruiting forms or serious decreases in the number of bolls set per plant. The above experiment was repeated in the spring of 1940 using 105 plants and three sizes of containers, 1-, 2-, and S-gal. jars. Similar re- sults were obtained except that the plants showed better fruiting than in the fall. For each of the three sizes of containers the results were: number of bolls, 12, 13, and 12; and percent-set, 44, 40, and 35 percent, respectively. In this test, the plants in the smallest containers, although 56 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION wilted for a period in the afternoon practically every day during the fruiting period, produced as many bolls as the plants in the larger con- tainers and these same plants showed a slightly better boll-retention percentage. In the summer of 1941, two groups, each containing eight Rogers Acala plants, were wilted on two occasions for two days at a time, during the fruiting period. The two periods of wilting for one group were imposed soon after the first blooms had appeared. The other group was wilted after the second and third week following flowering. There was a re- duction in size of the early-wilted group of plants and both of these groups gave a somewhat smaller number of bolls per plant than the checks. The data for this experiment were as follows: for the checks, fresh weight per plant—390 g., number of bolls-—24, and percent-set—56; for the plants wilted early, fresh weight—323 g., 21 bolls, and 59 percent-set; and for the late-wilted plants, fresh weight—380 g., number of bolls set-—22, and percent-set—54. ‘ The results of three additional experiments on the effects of frequent periods of wilting, for a few hours at a time, on growth and fruiting of cotton plants are shown in the first three sections of table 11. In the first two of these tests, Stoneville 2B plants were grown in sand culture and the treated plants were wilted for about two hours in the afternoon beginning soon after blossoming had commenced. Longer periods of wilt- ing in sand culture are difficult to maintain owing to injury to the plant that may occur from excessive drying of the sand. As may be seen in these data, these short-period wilting treatments again caused apparent reductions in vegetative growth, height, and weight (difference highly sig- nificant in the second experiment) of plants, and reductions (significant at 5 percent in the first and second experiments) in fresh weight of the bolls. The differences in number of bolls set were relatively small, al- though significant in the second test, and the percent-set figures were similar for the checks and wilted plants. In the third experiment, Stoneville 2B and Rogers Acala plants were grown in soil during the late summer, 1944, and part of the plants were wilted for a period of three or four hours just before each watering. During clear weather, this treatment usually resulted in wilting of the plants at least once every other day. The results of this test are given in the third section of table 11 and the differences obtained from the treatments are similar to those discussed above. There was a highly significant reduction in vegetative growth, with no corresponding de- crease in number of bolls, weight of bolls, or in the percentage of bolls set. In this case, the increase in the total number of forms shed was significantly less (5 percent level) in the wilted plants than in the checks. During the late summer of 1943, twenty Stoneville 2B plants were divided into four groups of five plants each and three of these groups were subjected to wilting for periods of different lengths when the first few bolls had set. The plants were kept outdoors except for the four- Table 11. Elfects of wilting on growth and fruiting processes in cotton. Average per plant. Number _ Weight of Number Number Number _ ‘ v _ Treatment of Height stem and 0 of bolls of bolls Weight Percent plants cm. leaves blooms set shed of bolls set 8- 8- Planted May 12, 1942—2-gal. sand-l variety Check a i s i 9s i ' 25s i 35 i 20 , 15 425 55 Wilted daily (after 7/18) I 8 I 88 i 250 i 33 l 17 16 335 52 Planted June 24, 1944—2-gal. sand——~1 variety Check i 8 i 84 268 4O i 19 21 410 48 Wilted daily (after 8/22) Ii 8 [I 80 218 34 I 15 19 321 44 Planted June 26, 1944—4-gal. soil—2 varieties . | _ Check i 8 i 92 i 313 i 42 i 15 27 384 36 Wilted before each watering I 8 ll 84 243 I 38 i 16 22 334 42 Planted July l7, 1943—3-gal. soil—1 variety Check 5 89 194 3O 1O 20 226 33 Wilted 2 days, 9/21-23 5 87 194 27 1O 17 242 37 Wilted 3 days, 9/21-24 5 79 i 168 | 26 9 17 206 35 Wilted 4 days, 9/21-25 5 76 || 148 ‘ 23 4 19 88 17 Planted December 5, 1944—4-gal. soil——1O varieties Check 70 93 231 32 13 - 19 a 272 41 Wilted 5 days, 3 /1-6 7O 90 253 31 1O 21 222 32 L9 LHDIT 0L NOLLVTEIH NI NOLLOO JO DNIGGEIHS (INIV ONILIHHéI 58 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION day period (September 23 to 27) when they were all placed in a green- house in order to prevent interruptions in wilting due to rain. During the wilting treatment, the plants often became extremely wilted. They Were maintained in a wilted condition for the desired length of time by light waterings at certain intervals. Shedding curves for these plants, fig. 16-A, show that the wilting treatments caused an early loss of fruiting forms. The increase in shed- ding was noticeable on the third day following the start of the treat- ment and the highest peaks in the curves are associated with the three- and four-day treatments. With the exception of the two-day wilting treatment, there was but little shedding following these early effects, as compared with the later high rate of shedding for the check plants. Data on this experiment are shown in the fourth section of table 11. The »O LU 5, 5° - 2 DAYS---- a DAYS ----- --4 oAYs-—--- - m ' PLAN-rs CHECK ~ . i E 4°. " "wluzo F" - 2 q ‘i: "55 A q- IL 3o '- - o "' 5'.’ ‘\.' c1 5 2o _ - g _ '-._ _ D " l’, ._ - = 1’ l’ \ " _.- ____ ____J "} \ ll 1r: ilnnnnnFrfifi~o--n_pl4ffglg||ll\ I6 l8 2022242628 302 4 6 8 l0l2l4l6 I8 2022 SEPTEMBER OCTOBER I943 lillllllllllll][IIILIIIIIIIIIIIIIIIIIIIIIIIIIIII 275 PEmOD - 250 - I‘WILTING>I _ u 5 200- ‘(D I75 - 2 g 15o - t :25 - 0 . o: I00" E _ 2 75 Z 5o- \ i z Oplll||TYfiTfrllfll' l||| 11111|11||| 2I232527I3579lll3|5|7l92l232527293|246810 FEB MARCH APRIL I945 Fig. 16. Eifects of wilting on shedding of cotton. A. Each curve represents five Stoneville 2B plants in soil, planted July 17, 1943, showing the efiects of 2-, 3-, and 4-day periods of wilting started on September 23, early in the fruiting period. B. Each of the two curves shows the shedding from 70 plants (7 each of 10 varieties, planted December 6, 1944) fol- lowing 5 days of wilting from February 28 to March 5, 1945. Note the marked increase in shedding of forms immediately following the wilting treatment. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 59 reduction in vegetative growth, noted also in the previous wilting tests, is again apparent in the averages for the height and fresh weight of the wilted plants, although no significant difference between the plant weights was found by statistical analysis. In addition, the 4-day wilting caused a highly significant reduction in both the number and weight of bolls. The percent-set figure for this longest wilting treatment was exceptionally low. A series of 140 cotton plants, including 14 plants of each of the ten varieties used in the seasonal studies (table 3), were raised in the green- house during the spring of 1945 for the purpose of testing the relative effects of a period of wilting on fruiting and vegetative behavior. Be- ginning on March 1, water was withheld for 5 days from seven plants of each variety, about two weeks after the first blooms appeared. The weather on these days was for the most part clear and, with the exception of March 3 when some of the wilted plants recovered from visible wilting, the plants were kept in a wilted condition by light Waterings until March 6. After this single 5-day period of wilting both the check and wilted plants were given the usual, adequate waterings up to time of harvest. In order to ascertain any possible skrinkage in stem tissues as a re- sult of the wilting of the plants, the diameter of the stem of‘ each plant was measured on March 1 with a micrometer caliper. The same measure- ments were taken again on March 6 as close as possible to the same points on the stems. As a result, it was found that 53 of the 70 check plants had increased measurably, 14 showed no change, and two of the measure- ments showed a decrease in diameter during the five-day period. Of the '70 wilted plants, however, 43 showed a definite decrease, 8 showed an in- crease, and 19 had not changed in diameter during the same period. This would indicate a shrinkage of the stem tissues had taken place during the period of water stress in the plants. Lloyd (38) found a similar skrink- ing of both stems and leaves of cotton on days of ordinary bright sunshine. The shedding curves for this experiment are shown in fig. 16-B where a sharp peak of shedding occurred during the last day or two of the wilting treatment. This was followed for some time by a lower-than- normal shedding rate until a small peak in the curve for the wilted plants occurred just before harvest. The data for the two groups of plants are shown in the last section of table 11, with the ten varieties averaged together for each treatment. A highly significant reduction in number and weight of bolls was found to result from the single wilt- ing treatment. Also a slight increase (significant at 5 percent) in the average fresh weight of the above-ground parts of the plants (without bolls) was found in the wilted plants. A difference of 9 percent was found in the percent-set figures. ‘ ' As in the case of the close-spacing, the varieties used in this experiment showed considerable variation to the wilting treatment on a percent-re- duction basis in regard to the data on fruiting. Likewise, however, no significant interaction was found between varieties and the wilting treat- 60 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION ment and the data showing the varietal differences found are not pre- sented. On the percentage-of-check performance basis, the Deltapine 14, Half and Half, Roldo Rowden, Lone Star, and Rogers Acala varie- ties were less seriously affected, in fruiting activities, by the wilting, than the other varieties tested in this experiment (Coker 4-in-1, Stone- ville 2B, Qualla, A. D. Mebane Estate, and Washington). The ratios (wilted plant : check plant) for the fruiting index were 100, 91, 69, 79, 73, 77, 71, 64, 67, and 58 percent, respectively, for the ten varieties in the order named above for the two groups. Variations in Soil-Moisture and Nitrogen The first of a series of experiments to study the effects of variations in levels of soil-moisture and available nitrogen on cotton fruiting and shedding was set up in the summer of 1941. Twenty-four 4-gal. jars of soil were planted on June 14 with the Rogers Acala variety and the first blooms appeared on these plants on July 22. On July 31, half of the plants were given the first of a series of wilting treatmentsby allowing the plants to remain in a wilted condition for three to four hours before each watering. Each of these two lots of plants were then divided into two additional groups (making four groups of six plants each) and each plant in two of these groups (one wilted and one nonwilted) was given extra nitrogen by adding ten grams of ammonium sulphate to the surface of the soil on August 8 and again on August 29. The plants were harvested on September 23. At this time, the average nonwilted plants with the extra nitrogen had the greatest fresh weight (475 g.), the largest number of bolls (30 per plant), and the highest percent-set (52 percent). There was but little difference between the other three groups either in vegetative growth or in fruiting. The fresh weight of the wilted plants with extra nitrogen showed the closest approach to the nonwilted-plus- nitrogen plants, in the data taken. For these three groups the data were as follows: Nonwilted, without extra nitrogen, fresh weight of plants- 194 g., number of bolls—10, percent-set—42; Wilted plants plus nitrogen, weight of plants—254 g., number of bolls—11, percent-set—44; Wilted plants without extra nitrogen, weight of plants—145 g., number of bolls—7, percent-set—39. It may be seen from these figures that the adequate water supply was as important as the extra nitrogen in obtaining a good yield of bolls and only where ample water was present to prevent wilting at all times did ‘the high nitrogen applications result in an increase in number of bolls. The high-nitrogen plants that were wilted, however, had a darker green color than those similarly wilted and lacking the extra nitrogen and it was in this group (extra nitrogen) that the above-noted increase in fresh weight from the nitrogen occurred in spite of the frequent wiltings. I ‘vww 'y—'_y"lrw'~y- -,_ ‘7 Table 12. Elfects of experimental variations in soil moisture, with and without extra applications of nitrogen. on growth and fruiting of cotton plants. Average per plant. I I I Weight 0f Number Number Weight of Number Height stem and of bolls of bolls I bolls per Percent I I11 I of plants cm. leaves set I shed I plant set w Treatment g. g. q I I I I I :1 I —N +N I ~—N I +N I -—N I +N I -—N +N —N +N —N I +N E C) Planted October 1, 1941—4-gal. soil-Stoneville 2B i; I I I I . I I I U Check I 5 I 75 I 79 I 208 I 292 I 11 I 14 I 13 I 13 . . . . . . . . . . .. 46 I 52 m 1-1/2 lb. above wiltingl 5 I 62 I 57 I 154 I 161 I 6 7 I 7 6 . . . . . . . . . . .. 46 I 54 g Wilted? - 5 I 62 I 55 I 119 I 114 I 4 4 I 7 6 . . . . . . . . . . .. 36 I 40 U ' U 9-4 Planted February 12, 1942—-4-gal. soil-Deltapine 14 g I I I I I o Check 3 I 98 I 105 I 192 I 308 I 14 I 18 15 I 6 . . . . ..I . . . . ..I 48 I 75 "7 4lb. above wiltingl I 3 I 78 I 87 I 148 192 I 10 I 14 10 I 11 . . . . . . . . . . .. 50 56 8 2lb. above wiltingl 3 I 82 I 77 I 155 145 I 12 I 11 10 I 8 . . . . . . . . . . .. 55 58 I.) Wilted? 3 I 70 I 63 I 100 95 I 8 I 8 I 6 I 7 I . . . . . . . . . . .. 57 53 g I I I I I I z Planted April 29, 1942——4-gal. soil-—3 varieties E Check I 6 I 70 I 86 I 147 I 288tI 8 I 17 I 12 I 23 I 147 I 403 I 40 I 43 p: I, 61b. above wiltingl I 6 I 66 I 73 I 140 I 300 I 8 I 15 I 11 I 19 I 148 I 327 I 42 I 44 F] 31h. above wiltingl I 6 I 63 I 73 I 104 I 236 I 7 13 I 12 13 132 I 296 37 I 50 11> Wilted? 6 I 66 I 63 I 102 I 155 I 8 10 '13 9 145 230 38 53 g Z Planted November 18, 1942—2-gal. soil—2 varieties H _ O I 63 72 71 123 3 5 E CD 50 1 te) 3 " 4 64 68 3 1 1 lThe plants were weighed daily and enough water added to make the total weight this amount above the wilting-point weight. _ iPlants were wilted for 3 or 4 hours before each watering, beginnifi about time of first bloom. 3Wilting treatment started February 19, 1943. Plants harvested arch 11. I9 62 BULLETIN N0. 677, TEXAS AGRICULTURAL EXPERIMENT STATION A somewhat similar experiment was conducted in the fall of 1941 with Stoneville 2B plants in 4-gal. jars of soil (planted October 1). In addi- tion to the check (abundantly-watered) and a wilted group (5 plants each), another group of 5 plants was given sufficient water each day to bring the weight of the plant and container to a certain weight (1% lb. above wilting point at time of first bloom). The treatments were com- menced on November 15 and the extra nitrogen (10 g. ammonium sulphate per plant) was added on December 10. The data were taken January 22 and they are shown in the first part of table 12. Additional experiments dealing with the availability of nitrogen in soil under varying moisture conditions and the coincident effects on fruiting and growth are also listed in table 12. In most of the cases the check plants (watered frequently enough to prevent wilting at all times) were compared with plants that were wilted before each watering and also with intermediate treatments consisting of the addition of suf- ficient water to bring the container and plant to a given weight daily. It may be seen in this table that the extra nitrogen was effective in in- creasing the number of bolls per plant only under the favorable soil moisture conditions. In two of the four experiments where the data are given, an increase was also found in weight of bolls due to the nitrogen and this effect was noticeable even in the drier soil. Vegetative growth in every case was inhibited by all of the drought treatments that were started early in the fruiting period, as shown by the height measurements and fresh weights of stems and leaves at time of harvest. In table 12, the percent-set figures within a given experiment are con- siderably alike regardless of the treatments involved. These relatively similar percentages are apparent even in cases where the treatment produced marked differences in vegetative growth or in yield of bolls. Exceptions to this uniformity in percent-set figures occur in two cases: In the check high nitrogen plants in the second experiment, the percent of bolls set is high (75 percent). This may be attributed to the active vegetative growth of these plants up to the time of harvest and to a slight delay in blooming. In the other case, the wilting treatment was not ap- plied until a large plant had developed with the extra nitrogen, and the treatment was severe enough to cause considerable shedding which re- sulted in the low percent-set (21 percent) for the high-nitrogen plants. A condition similar to this is shown in fig. 17. In general, the results in table 12 compare favorably with those in table 11 where somewhat similar wilting treatments were applied. Excessive Soil Moisture In the late summer, 1943, a study was made of the effects of water- logging of soil on the shedding of cotton forms. Thirty-two plants of the Stoneville 2B variety, planted July 17, in 3-gal. jars of soil, were separated into four groups of 8 plants each on September 16 and the fol- lowing treatments were started on that date: FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 63 Fig. 17. Effects of several-hours’ wilting every other day on Deltapine 14 cotton plants in soil with high nitrogen content. In this case the wilting was started (plant on right) when the plants were over 2 months old, as the first bolls were approaching maturity. The heavier shedding that has occurred on the wilted plant is apparent in the lower right picture (leaves removed). This represents one of the few intermittent wilting treatments that resulted in increased shedding. 64 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION (a) Checks, usual watering at first signs of wilting. (b) Two waterings of soil each day, to maintain a saturated condition of the soil. (c) Drainage hole in bottom of jar closed and moisture con- tent of soil maintained at a high level by daily watering. (d) Jars (with drainage hole open) placed in a saucer which was kept filled with water. ~ During the first eight days after the treatments were applied, the plants in the jars with no drainage (c) wilted conspicuously and five of these eight plants died between the fifth and eighth day. In this same period many of the (d) plants standing in saucers of water also wilted but none of these were dead on September 24. A few of the plants that were watered twice daily (b) wilted for two or three days towards the end of this 8-day period. In spite of the continued wilting to which many of the plants had been subjected, there was no shedding of any fruiting forms up to Sep- tember 24. At this time a period of cloudy weather with occasional rain began and lasted until October 2, during which the living plants in all treatments apparently recovered from any previous wilting and the foliage remained turgid. On September 28, a large number of forms were recorded as shed from all of the living plants, with a significantly larger proportion of these from the d plants standing in water than from either the checks or the plants watered twice daily. Following clearing of the weather on October 2, many of the treated plants showed severe wilting symptoms. During the following two weeks, shedding of forms, mostly in the young boll stage, proceeded at a fairly high rate in all of the living plants. In this period, also, the three remaining plants in the c treatment died, together with four of the d plants standing in water. Only one plant died in the b-treatment group and this was recorded as dead on October 5. The final totals of shedding counts for the eight plants in each of the four groups up to the time of harvest (October 18) were as follows: Check (a) 31 small squares, 20 large squares and 67 bolls; watered twice daily (b) 34 small squares, 41 large squares, and 64 bolls; no drainage (c) 9 small squares, 11 large squares, and 55 bolls; and jars standing in water (d) 57 small squares, 48 large squares, and 82 bolls. The total number of large bolls on the plants at time of harvest; per group of eight plants, were 50, 51, 12, and 24 for the checks and each of the three treatments (a-d), respectively. In these data, all large bolls (past the shedding stage) were recorded for all plants including those that had died during the course of the experiment. It is of interest to note, in the data just given, that the d plants, standing in water, showed an increased shedding rate and a smaller number of bolls retained,‘ as compared with the checks. This increase in shedding was not apparent for the plants that received excessive water- ing (twice daily) nor for the c plants in jars without drainage, most of which died early. It seems therefore that some of the high soil FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 65 moisture treatments may have predisposed. the plants to shed a few more forms than normal. In this experiment, however, many of the forms on the plants that died were not shed but remained attached to the branches even after the leaves were shed. In such cases the plants may have been killed by drowning of the roots before formation of the abscission layer had become complete. The heaviest shedding in the early part of this test was also associated with the earliest wilting symptoms following the water-logging treatments. Albert and Armstrong (2) have reported in- creased shedding of small squares following flooding of the soil. In this connection, it may be of interest that certain large plants in the 1942-1943 fall-winter series growing in a fine sand that did not allow proper drainage showed severe wilting on clear days following periods of cloudy weather in December. In no case, however, was any excessive shedding recorded following this wilting which in some cases extended over a period of two days or more. The percent-set figures for these plants were similar to those of the same varieties in the 1943 fall series when a coarser sand was used and the wilting effects following cloudy days were not as marked. EFFECTS OF TEMPERATURE Since high air temperatures are considered as contributory to exces- sive shedding in cotton, a few experiments were made to study the ef- fects of temperature as a factor in shedding and to compare tempera- ture effects with those of light, under similar cultural conditions. Some of the studies of the growth and fruiting of cotton under high tempera- ture conditions were made by growing the plants in an unshaded green- house during the summer. On clear days in the summer, the temperature in the greenhouse without shades was about eight to twelve degrees higher than outdoors, as shown in fig. 2-D. Maximum temperatures of from 105° F. to 115° F. were recorded under these conditions by a thermometer shaded by the foliage of the plants around noon on cloudless days. Also, one test was made in which cotton plants were removed from a heated to a nonheated greenhouse for several days in the winter to obtain a low- temperature effect. High Temperature Conditions Extended periods of high daily temperature. The first evidence in this work of the unfavorable effects of temperatures above 100° F. on cotton fruiting was obtained from a group of sixteen Rogers Acala plants in 2-gal. jars of sand that were started in an unshaded greenhouse in April and kept there until time of harvest, July 25, 1941. These plants made ex- cessive vegetative growth and were much taller than check plants of the same age that were kept outdoors after May 30. On the average, the greenhouse plants produced 64 blooms and eleven mature bolls, giving 17 percent-set. The average outdoor check plant produced 41 blooms and 66 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Fig 18. The center The cotton plant on the left was grown outdoors in the summer. plant was placed in a section of the greenhouse covered with roof shades, as soon as the first squares were full grown. At the same time (June 6), the plant on the right was placed in an unshaded part of the greenhouse. The photographs, taken August 2, show the increased vegetative growth of the greenhouse plants. The nonshaded greenhouse (high temperature) plant produced more bolls than either the comparable shaded plant or the check. The out- door check plant produced the largest weight of bolls, however, and showed the highest percentage of bolls set. -..~»L.-. FRUTTING AND SHEDDING OF COTTON IN RELATION TO LIGHT 67 24 bolls, or 56 percent set. A high shedding rate accompanied by low yield of bolls were outstanding effects of the high temperature prevalent in the greenhouse during the summer. Another test similar to the one above was made during the summer of 1944 with cotton plants in 4-gal. jars of soil. Two groups totaling 36 plants, three plants each of six varieties (Stoneville 2B, A. D. Mebane Estate, Rogers Acala, Deltapine 14, Qualla, and Half and Half), were. used. One group was kept in the greenhouse from June 6 until August 2 (time of harvest) and the other served as outdoor checks. The results are shown in the first part of table 13. As in the previous test, a higher shedding rate was found for the high-temperature plants. In this case, the difference in set of bolls, although only three per plant, was significant at the 5 percent level while the decrease in weight of bolls was highly significant. At the time of harvest, the average boll on the check plants weighed 23 grams while the average boll on the high- temperature plants was only 14 grams in weight. Analyses of the data show that the greater number of bolls shed under the high temperature conditions is highly significant. The increased vegetative growth (fig. 18) of the greenhouse plants is shown by the average weight of stem and leaves which is 20 percent greater (significant at 1 percent level) for the greenhouse group. This more active, growth is associated with a signifi- cantly (at 1 percent level) larger number of blooms and a lower fruit- ing index figure for the plants under the high temperature conditions. In the shedding curve for the greenhouse plants (not shown) the young bolls were shed at a steady, high rate throughout the fruiting period while the outdoor plants showed a single peak coinciding with unusually high maximum daily temperatures outdoors (see fig. 19). I I I I I I I I yéfliiiéliifiiélIl r I I I T I I I I I I I80 - E CLOUDY PARTLY CLOUDY l] FAIR __ __ '\ p‘ --— I05‘ 3 I60 - i -Ioo x MAXIMUM DAILY TEMP. w I40 - - 95 w g I20 - V/y- - so In _ 1 ~ _ g I00 . l; 85 5 °° ' i /l '- g 6O - PLANTED APRIL a l/ - i’ 4o - / PLANTED APRIL zz — ,1‘ . I I I I I I I I I I - 4 L - I I--I--I"T'T I I I I I I I I I I 1m I I I I 2| 23 25 27 29 I 3 5 7 9 ll l3 I5 l7 I9 2| 23 25 ' 27 JUNE JULY I944 Fig. 19. The solid-line curve represents shedding from 48 cotton plants, Stoneville 2B variety, planted in 2-gal. jars of soil on April 8, 1944. The broken line shows the shedding from 42 plants (six varieties) planted in 4-gal. jars two weeks later. The weather during the period was mostly clear except for a few -cloudy days early in July. Note the relatively steady rate of shedding for the older plants and the sharp peak for the later planting which occurred late in the fruiting period and apparently was associated with the high daily temperatures. , Eflects of high and 10w temperatures on fruiting of cotton plants. Average per plant. I ht Check (outdoors) High temperature greenhouse 6/6-8 /2 ; 89 I Weig Nunf1ber| Height of stgm Nunfmber Nungber| Nungber o cm. an o o plants leaves I blooms bolls I g. set Planted April 6, 1944——4-gal. soil-(ivarieties I I I I I I 18 I 81 I 228 I 35 I 15 I I 18 I 97 I 273 I 52 I 18 I I I I I Planted June 24, 1944—2-gal. sand—StoneviIIe 2B High temperature greenhouse 8/23-28 I I IsIs4I2ss|40|19 |s|s5l|s0o||42 21 I I Planted September 16, 1943—4-gaI. soil——StonevilIe 2B Check (heated greenhouse) I 10 I 53 . . . . . . .. 15.3 In cold greenhouse 12/21-23 & 12/29-1/4 53$ fifl 5. I 1O I 58 . . . . . . .. 15.7 6. w C1 IT‘ F‘ P1 F] n-1 z Z .0 $ Q F‘ H I111 >4 11> U] > Q 5'1 O C1 F‘ H C '21 17> F‘ P1 MI "U P1 5U v-u 3 F] Z P] [H 1-] W H. v-c O z FRUITING AND SHEDDING OF COTTON IN’ RELATION TO LIGHT 69 Brief period of high temperature. In the second part of table 13, the results of a late-summer experiment dealing with high temperature ef- fects for a single brief period are shown. Eight Stoneville 2B plants were placed in the unshaded greenhouse for five days, during the early part of the fruiting period (August 23 to 28). A peak in the shading curve (not shown) with its apex on August 27 appeared for the plants in the greenhouse. This increased rate of shedding started on August 26 and lasted until August 31. There was practically no shedding from the check plants until August 31. In spite of this early shedding, the plants that were in the hot greenhouse for five days did not show any significant differences in any of the measurements as compared with the outdoor checks, although the figures for fresh weight of plants were slightly higher for the high temperature plants. Further evidence of increased shedding tendency following exposure to high temperatures is shown in fig. 19 where the peak of shedding (in the curve for early summer plants in soil, p-lanted April 22) oc- curred on July 19 following a. week of hot weather in which the maximum daily temperatures were 100° F. or more outdoors. This was the only peri- od of hot weather that has been of sufiicient intensity to cause noticeably increased shedding from outdoor plants, in the course of the five-years’ work here reported. Lloyd (38) recorded high shedding rates on a few occasions which he believed to be due to high temperature during a few previous days. Evidence of varietal differences in response to high temperature. The data for the group planted on April 6, 1944 in the first section of table 13 represented a single group of eighteen plants made up of six varieties. By separating the varieties in this group, the resulting re-combination of data showed considerable variation in the relative effects of the high temperature treatment between the two groups. For example, one group of nine plants (Stoneville 2B, Deltapine 14, and Half and Half varieties) averaged 22 bolls set, 33 bolls shed, and 40 percent set. The other group (Rogers Acala, Mebane Estate, and Qualla) produced only 14 bolls set per plant, with 36 bolls shed, and 28 percent set. The average for both groups, as shown in table 13 was 18 bolls set per plant, with 34 bolls shed and 35 "percent set. With the small number of plants of each variety per treatment, no significance was found in the interaction of varieties and the high temperature. Effect of low temperatures on shedding. During the early winter, 1943, a group of 20 Stoneville 2B plants was set aside for a study of cold ef- fects on boll shedding. On December 21, ten of these plants were placed in an unheated greenhouse. After two days, during which time a minimum night temperature of 45° F. was recorded, the plants were returned to . the heated greenhouse on account of danger from freezing. They were placed in the cold greenhouse again on December 29 and remained there until January 4, 1944. During this second period, a low temperature of 38° F. was recorded on one night and some of the leaves showed injury ' son of the shedding rates associated with the various treatments. 70 ‘BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION from the cold. The plants were kept in the warm greenhouse until Jan- uary 14 when data on fruiting were taken as shown in the last section of table 13. No significant difference was found in the set of bolls or shed- ding following the cold treatment. These results are in accord with field observations made in Texas Panhandle fields near mid-October 1941 when cotton with many green bolls showed no signs of excessive shedding following near-freezing temperatures on several previous nights. Ewing (23), however, noted an increase of shedding on one occasion in two years’ work in Mississippi following low temperatures at night (54° to 56° F.) but there was also high evaporation at the same time and the ef- fects of the low temperature were uncertain. VARIATIONS IN CULTURAL PRACTICES In order to obtain information as to the possible effects of nutritional differences that might arise in the various treatments, with the particular types of culture used in this work, experiments were conducted to as- certain the differences in fruiting that might follow certain definite variations in the amount of nitrogen and certain mineral elements avail- able to the plants. Another purpose of these studies included compari- Al- though most of these data indicate but relatively small differences in shedding rates between the different treatments, marked differences in fruiting were obtained in many cases. With few exceptions, these nutrient variations were designed to simulate ordinary differences in soil fertility that might occur under field conditions, rather than to study the effects of deficiencies or excesses of these elements. Varying Composition and Amount of Nutrient Solution During the early part of the work on the shedding problem, a number of tests were made of the effects of changes in the ratios of the three important fertilizer elements (N, P, and K) on the set of cotton fruit. Nutrient solutions were used that included all possible combinations of two different levels of each of these three elements. In addition, a change in concentration of the nutrient solution, at a time about midway between the opening of the first flowers and the opening of the first mature bolls, was also included as a variable in the studies with sand-culture plants. Results of two of these experiments are shown in table 14. The effects of increased nitrogen supply to the plants is very evident in the consistent increases both in vegetative growth and number of bolls set. These increases are apparent regardless of whether the extra nitro- gen was supplied by increasing the amount of this element in the N-P-K ratio or by applying larger amounts of the complete nutrient solution by extra daily applications. Upon analysis of the data, significance at the 1 percent level was found for the figures on fresh weight of plants, num- Ho. Man-Ana», .. _ . . FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT \ 71 Table l4. Etfects of variations in composition and amount of nutrient solution supplied to cotton plants, in sand culture, on vegetative growth and fruiting. Average per plant in all cases. Compositionl of nutrient solution Fresh _ weight Number Number Percent Up to middle oi’ After middle of of plants of bolls of bolls set fruiting period fruiting period g. set shed “*5! Planted April 10, 1941-—-Early summer series—7 plants per treatment N-P-K N-P-K 390 24 1 9 56" N-P-k N-P-k 387 21 23 43 N-p-k N-p-k 315 V19 21 i 4s N-p-K N-p-K _ 402 24 23 51 n-P-K n-P-K 23 1 1 6 1 4 53 n-P-k n-P-k V 201 15 14 52 n-p-K n-p-K 221 14 16 47 n-p-k n-p-k 216 14 18 44 N-P-K, 2xda.2 N-P-K, 2xda. 554 33 18 65 n-p;k, 2xda. n-p-k, 2xda. 465 28 22 56 Difierence required‘! .. . . . . . . . . . . . . . . . . . . . . 62 2. 3 4.4 . . . . . . . . . . Planted June 14, 1941—Late summer series—8 plants per treatment N-P-K N-P-K 363 a 20 34 37 N-P-K n-p-k 336 17 37 31 N-P-K NN-P-Kv 405 2e 33 44 NN-P-K NN-P-K 535 27 40 40 n-p-k ~ n-p -k 1 69 1* 9 1 8 33 n-p-k N-P-K , 247 14 21 40 N-P-K, 2xda. N-P-K, 2xda. 601 31 32 49 N-P-K, 3xda.3 N-P-K, 3xda. 742 37 44 45 n-p-k, 2xda. n-p-k, 2xda. 302 16 27 37 n-p-k, 3xda. n-p-k, 3xda. 419 24 33 42 Difierence required4. . . . . . . . . . . . . . . . . . . . . . 67 4.4 8 , _ , _ , , _ _ _ _ 1 N = 90 p.p.m. nitrogen P = 45 p.p.m. phosphorus K = 60 p.p.m. potassium = 30 p.p.m “ p = 15 p.p.m. “ k = 20 p.p.m. “ n . NN =180 p.p.m. “ _ _ iOne liter of nutrient solution twice a_ day, 8 A. M. and 2 P. M. s “One liter of nutrient solution three times a day, 8 A. M., 12 N., and 2. P. M. 4Difierence required for significance at 5 percent level. ber of bolls set, and number of bolls shed in both experiments. The minimum differences required for significance at the 5 percent level are shown in the table for each of the two experiments. In spite of wide variations in vegetative growth as shown by the fresh weights and also in the number of bolls per plant, there was hut little difference between the different treatments in regard to the precent- 72 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION set. This indicates that increased vegetative growth, under these condi- tions, was not made at the expense of fruiting activities, but rather that the plants set fruit in proportion to their size. The slightly higher percent-set figures in case of the plants that received the highest amounts of nitrogen (either by extra daily nutrient applications or by stronger solutions) may have been due to a somewhat extended period of vegeta- tive growth. In these cases, each of the high nitrogen plants had a few squares developing at time of harvest. Another‘ outstanding result shown in table 14 is the fact that changes in the concentration of the nutrient solution, during the fruiting period when there were several forms on each plant in the small-boll (sus- ceptible to shedding) stage, did not seriously affect the percent-set. This shifting from one nutrient solution to another produced changes, in vege- tative growth and in numbers of bolls set, associated with the amount of nitrogen supplied. When the change was made from the weak to the strong solution, the plants were larger and the number of bolls set greater than in case of the constant use of the weaker solution. Likewise, when the shift was from the strong to the weak solution, there was a decline in the growth and fruiting figures as compared with continued use of the more concentrated nutrient solution. Wadleigh (54) has shown that changing cotton plants from a nutrient solution of low osmotic concentra- tion to a, higher osmotic concentration, while maintaining the same high nitrogen level, may result in certain (relatively small) differences in shedding rates. Of’ interest, also, in table 1.4 is the significant difference in number of bolls set between the low and high potassium contents in the nutrient solution, in the high-nitrogen plants of the early summer series. The low potassium, in case of both low and high phosphorus, produced a few less bolls than the high-potassium solution and a slightly smaller per- cent-set also accompanied these low-K treatments. The shedding curves (not shown) were similar for all of the treatments in both of these ex- periments and peaks occurred at practically the same time in all of the treatments. The early summer experiment (first part of table 14), referred to above, was repeated during the fall and early winter. The seed were planted on September 4 and the plants were harvested on January 6-7, 1942. In keeping with results presented under “Seasonal Differences in Fruiting and Shedding”, the set of- fruit was much poorer and the percent-set lower than in the case of the summer plants. For the four treatments involving the high-nitrogen nutrient solutions, the average of thirty-two plants showed 6 bolls set per plant, 2O bolls shed, and a 23 percent-set. The corresponding figures for the same number of low-nitrogen-level plants were 5, 16, and 24, respectively. There was no apparent difference be- tween the two groups in regard to size of plants. It may be seen, there- fore, that the fruiting activities of both groups were very similar. At this season of the year, both high- and low-nitrogen plants set a small .1. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 73 tnumber of bolls considering the number of blooms produced. The I egligible effects of the extra nitrogen for greenhouse plants during the ‘fiall and early winter may be due to the insufficient amounts of carbo- iliydrates that are synthesized under the poor light conditions prevalent sgrowth. Much of the inorganic nitrogen taken up by the plants is sup- iposed to be utilized through combination with carbohydrates into nitrog- “enous organic compounds. A third comparison of these same nutrient solutions was made during sfthe spring of 1942 with Stoneville 2B plants started on January 9. In ' case, the four groups of plants (32 in all) with the high-nitrogen lutions had, on the average, 16 bolls set per plant, with 19 bolls shed, d a 46 percent set. The low nitrogen groups gave 11 bolls set, 13 bolls ghed, and 46 percent set. These results resemble those of the summer lants more closely than the fall-grown plants, a condition found previously be true for the spring series. The increase in number of bolls set e to the higher nitrogen level is apparent, although the percent-set gure is the same for both high- and low-nitrogen treatments. In this experiment, there was a-significant difference between the high- d low-phosphorus applications at both levels of nitrogen. For the gh-nitrogen series, the 16 plants receiving high-phosphorus (45 p.p.m.) t; while the 16 low-phosphorus plants averaged 11 bolls set, 16 bolls d, and 41 percent set. For the low-nitrogen series (16 plants), these e figures were 12, 13, and 48, respectively, for the high-phosphorus d 9, 13, and 41, respectively, for the low-phosphorus treatments. No "dy explanation for these differences in phosphorus levels in this one t is apparent, although a different type of river sand was used in n that of other types of sand used in some of the other experiments. Another summer series of plants was employed in a similar test in 42. This group was made up of 56 plants, half of which were Stone- lle 2B and the other half Deltapine 14. Seven different nutrient solu- ons were used involving three levels of each of the three major elements. e check treatment consisted of the middle level of all three major ements. In case of each of the other treatments, either the high- Ht or lowest level of one of the elements was used in combination iwith the middle levels of the other two elements. The three levelsof trogen consisted of 42, 84, and 168 p.p.m.; of phosphorus, 16, 32, and ‘p.p.m.; and of potassium, 20, 40, and 80 p.p.m. The results were very - ilar to those obtained with the Rogers Acala veriety in 1941. Increases ere obtained in fresh weight of the stem and leaves, number of bolls t, and weight of bolls at harvest, in proportion to the amount of nitro- n in the nutrient solution. The variations in phosphorus or potassium d not produce any significant differences in any of the growth or fruit- Hg responses. The percent-set figures varied from 46 in the loW-nitro- filuring these months which resulted in an increase in the vegetative utrient solution averaged 20 bolls set, 22 bolls shed, and 48 percent is case and its phosphorus-fixing properties may have been greater ' 74 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION gen plants to 62 in the highest nitrogen level. Again, the narrow range in percent of bolls set is apparent in spite of widely different proportions of nutrient elements in the solutions used. During the latesummer of 1942, an additional experiment was carried out with 32 Stoneville 2B cotton plants, planted July 31 in 2-gal. jars of ‘sand and harvested Ocetober 19. The plants were “divided into four groups of eight plants each. One of these groups was supplied with a nutrient solution with medium concentrations of the three major elements as listed above. The other groups were supplied with three similar solu- tions in which the nitrogen, phosphorus, and potassium were used, re- spectively, at the highest level with the other two elements in each case remaining at the medium level. In this test, larger plants with about four more bolls (22 vs. 18) per plant and a significantly higher weight of bolls were obtained with all three of the treatments-—high nitrogen, high phosphorus, and high potassium. Here, again the percent-set figures showed no marked variations, however, between treatments since the range was from 57 to 62 percent. i Comparison of One, Two, and Three Plants in Each Container In order to ascertain any possible eifect on fruiting or shedding that might be associated with an increase in the number of plants per con- tainer, a small test was made during the late summer, 1942, with Stone- ville 2B plants in 2-gal. jars of sand. All of these jars were planted with several seed in each and after the seedlings had reached a fair size, all but one were removed from four jars, all but two from a second group of four jars, and three plants were allowed to remain in the third _ group. The average number of bolls produced by these plants per jar was: 18 (one plant), 19 (two plants), and-2O (three plants). The percent- set figures were 56, 53, and 46, respectively. There were no statistically significant differences between these groups. OTHER EXPERIMENTS IN RELATION TO SHEDDING During the course of these studies, several experiments have been com- pleted, the results of which were either inconclusive in ‘character or of minor importance in relation to the shedding problem. Some of these results seem of suflicient interest to include in this report. Efiects of “Chemical Hormones” Since certain synthetic growth-promoting chemicals have been used to prevent dropping of fruits in certain horticultural crops, especially the preharvest drop of apples (24), trials were made of the effect of some of these chemicals on cotton boll-shedding. No apparent effects were found when 10 to 20 mg. of indole acetic, indole butyric, or naphthalene- acetic acids or naphthylacetamide were added six or eight times during FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 75 the fruiting period to each liter of the nutrient solution applied to plants in sand culture. In the greenhouse, cotton plants sprayed with Solutions containing 1000 p.p.m. of naphthalene acetic acid or levulinic acid showed marked epinasty (downward bending) of the leaf petioles within twenty- four hours, followed by excessive shedding of small squares. The epinasty and increased shedding were not apparent when the concentration of these two chemicals was reduced to 1'00 p.p.m. During the summer of 1942, thirty-two Stoneville 2B cotton plants in 2-gal. jars of sand (planted May 12) were divided into four lots, three of which were sprayed eight times between June 24 and August 12 with solutions of 5 p.p.m. of the potassium salt of naphthalene acetic acid, 10 p.p.m. of levulinic acid, and 40" p.p.m. of levulinic acid, respectively. The plants were sprayed thoroughly at each application so that both sur- faces of all leaves were wetted. Data taken on August 27 showed no significant difference in either growth or fruiting between the different treatments nor between the checks and the treatments. The average number of bolls per plant for each of the four groups varied between 21 and" 23, while the percent-set figures varied from 55 to 60. __A.-..__>~Z;....¢..-.._.-___._.;-.L.V_.\;Z>i. _.i.>.l-.>.a)_.............i.-.e_._ .---_._._m.~....__..__»~.___ lb Fig. 20. Sections of cotton fruiting arms showing (center and lower left) three young bolls the ste-ms of which were treated with naphthalene acetic acid in lanolin just before blooming. Abscission of the form was prevented but the boll dried up. In the lower right, a similarly "treated boll is shown developing normally; above, untreated forms. Arrows show point of application of the “chemical hormone.” a 76 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Another test with greenhouse plants, involving the treatment of the stems of squares, flowers, and young bolls with lanolin (wool fat) im- pregnated with naphthalene acetic acid at three different concentrations, was carried out in the spring of 1941. This work involved the treatment of 468 stems of fruiting forms on 30 Rogers Acala plants, including check treatments with plain lanolin, during the period from April 22 to May 8. With the 0.01 percent concentration of naphthalene acetic acid in lano- lin, there was no difference in the amount of shedding resulting from the treatment. When the concentration was increased to 0.10 percent, the percentage of bolls set became smaller and still more serious shedding occurred with the 0.50 percent concentration. There was an indication that the forms treated during bloom were more adversely affected than the squares or small bolls. With the highest concentration of the chemi- cal, many of the treated stems of the fruiting forms remained green and no abscission layer formed. The young boll did not develop in these cases, however, but dried up together with the calyx and remained on the plant, as shown in fig. 20. Wetting of Open Flowers As previously mentioned, other workers (38, 23) have found indica- tions of increased shedding rates for young bolls whose flowers opened on days with showers in the forenoon. _ They concluded that rain destroyed the pollen in recently opened flowers and thereby prevented fertilization. Since these same effects have been noted in these studies and interpreted as due to low light-intensity effects, and since the same effects are found under greenhouse conditions, a test was made in which the flowers were wetted experimentally. This experiment was conducted during the sum- mer, 1942, with sixteen plants of the Stoneville 2B variety planted May 12 in sand. The flowers on eight of the plants were sprayed thoroughly with water from an atomizer twice each day (morning and afternoon) on the day that they first opened (White flowers). The flower wetting was commenced on July 18 and the plants were harvested on August 28. At this time, the check plants had 19 bolls per plant and showed 54 percent-set. The plants with the sprayed blossoms had 17 bolls each with 46 percent- set. There were no statistically significant differences between the check and treated plants, in regard to number of bolls set or the number shed. In view of the fact that many of the sprayed blossoms retained water in the base of the corolla throughout the day, it seemed that this treatment should have equalled the effects of a heavy shower or a more prolonged rain, in‘ deleterious effects on the pollen. Lloyd (38) reports that Orton and Duncan performed a similar experiment in 1907 and obtained a dif- ference of only 6.61 in the percentage of bolls set and that they doubted the importance of rain on freshly opened flowers as a cause of shed- ding. Although Lloyd (38) has reported the bursting of cotton pollen grains in water, there is some doubt as to the actual damage to the pollen from FRUITING AND SHEDDING_ 0F COTTON IN RELATION T0 LIGHT 77 5m Under the microscope, it has been observed that pollen from ntly opened cotton flowers could be caused to burst readily by wetting. Tthe other hand, pollen was kept over night in a Petri dish, in which water placed and condensation occurred on the glass, without damage to '_;grains. It is quite likely,‘therefore, that much of the pollen in an A flower might escape damage from wettingeven during a rainy day. Pruning and Related Treatments opping of plants. In the 1942 spring series of Stoneville‘ 2B plants in _ - culture, the top of.the main stem was removed from eight plants 13m». time of the opening of thegfirst blooms on March 14. This treat- ‘t was applied following suggested benefits from topping of ‘plants in _field in southeastern Louisiana (4). Following topping, the treated ‘lfts made later vegetative growth mostlyfrom the tips of fruiting “ilches. The topped plants (without bolls) averaged 431 g. each at ‘ of harvest as compared with 474 g. for the check plants. The topped ts also had 15 bolls per plant and 45 percent-set while the check plants 21 bolls and 48 percent-set. The average boll on the topped plants ghed 20 g. as compared with 17 g. each for the checks. Obviously, - was no marked advantage in this experiment from the topping tment. émoval of leaves from stem and fruiting branches. In the late fall ‘c1941, the stem leaves were removed from seven Rogers Acala plants i" n the first blooms appeared on the oldest fruiting branches. This tment left the plants with foliage on the fruiting branches only. As p lt, a reduction occurred in, the number of bolls, from 6 on the check ts to 3 on ‘the stem-defoliated plants, at time of harvest. The set of was 25 percent for the treated plants and 21 percent for the checks. ‘the same time, the leaves were removed from the fruiting branches, of n other comparable plants. These branch-defoliated plants also had lls on the average at time of harvest and they showed a 16 percent- Along with this test, smaller numbers of Stoneville 2B plants were , i each of the above treatments separately and in this case branch liation produced more bolls than were obtained from the nontreated t. g_ n account of the indicated possible beneficial effect of fruiting-branch oliation, another experiment was -carried out in the late summer of i '2. Fourteen sand-culture, Stoneville 2B plants were divided into two and the fruiting branches of one lot were defoliated as described ve. The data, taken on November 17, were very similar for the treated nontreated plants. The defoliated plants produced 6 bolls, while the . plants produced 7 bolls. Bolls from each treatment averaged 21 I s apiece. Both lots showed 46 percent-set. the spring of 1945, all leaves and bolls past the shedding stage were oved from the fruiting branches of four sand-culture plants in the 78 BULLETIN NO. 677, TEXAS AGRICULTUTIAL EXPERIMENT STATION greenhouse. Many of the squares that remained produced full-sized bolls and no excessive shedding from these defoliated fruiting branches was noted. Removal of squares from certain fruiting branches. The fruiting forms, consisting of various sizes of squares, were removed from the three lowest branches of five Rogers Acala plants in the summer of 1941. This treatment compelled the setting of the entire crop of bolls on the upper branches. Although the percent-set for plants with the forms removed was 58, compared with 50 percent for the checks, there was no significant difference between the number of bolls set—l9 for the former and 18 bolls for the latter group of plants (checks). A somewhat similar test was performed with sixteen Rogers Acala plants in the late summer, 1941. The squares were removed from all branches except six, spirally located along the stem, on eight of the plants. It was intended by this experiment to ascertain whether a cotton plant would set a given load of fruit even if the bolls had to be borne on a limited number of branches. At harvest, the plants with certain branches defruited had a fresh Weight of 472 grams compared with 363 grams for the checks. The treated plants had only 11 bolls per plant but showed a percent-set of 69 Whereas the checks had 20 bolls with a 37 percent-set. This experiment indicated that boll distribution over the entire cotton plant was necessary for the best yield of bolls. There was a marked im- provement in the percentage of squares that produced mature bolls, how- ever, in the plants with fruiting restricted to certain branches. Removal of bracts from fruiting forms. Evidence of the independence of the cotton boll in relation to closely adjacent plant parts was obtained by removing the bracts (calyx) from immature forms at yarious stages of growth, from medium-sized squares to small bolls.‘ Under greenhouse conditions in the spring of 1944, with Stoneville 2B and Coker 4-in-1 plants, this treatment did not change, the percent-set of the treated forms as compared with those on which the calyx was allowed to remain. No measurements were made of boll-size in this test although the bolls that developed without the calyx appeared as large as those with the bracts removed. Covering of Fruiting Forms In the early fall of 1943, five lots of fourteen forms each, including squares, flowers, and small bolls, on a group of nine plants in the green- house were covered with five different materials: black cloth, black paper, white paper, thin cellophane, and tinfoil, respectively. A small piece of the material in each case was folded over the cotton form and tied around the stem of the form so as to form a hood. There were 42 squares, 11 flowers, and 17 small bolls covered, all on October 5. All forms not covered were removed. By October 10, only one of these forms had been shed. There were nine small bolls recorded as shed by October 14 and by the 23rd a total of 38 small bolls and 1 square had been shed. Thus, there mvmyrr\4 . ,__ w. , FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 79 remained 30 bolls on the 9,plants most of which were past the shedding stage. Among those remaining were 9 covered with black cloth, 7 with black paper, 5 with white paper, 5 with tinfoil, ‘and 4 with cellophane. Having obtained this indication that covering of immature cotton forms with opaque material would not always cause the form to be shed, a further trial of this sort using only tinfoil was made in the spring of 1944. Twenty-seven large squares, flowers, and small bolls selected at random on three Coker 4-in-1 cotton plants were covered on April 1'0. On May 19, fifteen of the twenty-seven forms covered had developed into large bolls. In some cases where the tinfoil was placed on large squares, the flowers had opened and the boll developed, all within this covering. The three plants had a total of 32 large bolls at this time and three comparable check plants had 33 bolls. It seemed therefore that the shed- ding rate was not increased by the tinfoil covering on the immature forms. It was also indicated that the shedding stimulus came from a general con- dition of the plant rather than arising from local conditions in the square or young boll. Further evidence was obtained in this connection when nine cotton plants were shaded for one week in October, 1943, except for certain fruiting branches which were allowed to project through holes in the black cloth covers. In this case, there was excessive shedding of many cotton forms 0n the branches in the sunlight and somewhat heavier shedding of forms on the shaded branches. Comparison of black cloth and cellophane covers. In order to discover any causes of shedding, other than light intensity, that might be involved in the shading of cotton plants with cloth, a small test was made in the late summer of 1943 in which black cloth and cellophane covers were compared. Three Rogers Acala plants were, covered individually with black cloth at 3:00" P. M. on September 7. At the same time, three com- parable plants were covered with a cylinder of cellophane (Eastman acetate sheet, 0.01 inch) having a disc of the same material fastened loosely as a cover. There were also three check plants. The ventilation through all of the covers was so regulated that the temperature in each was two to four degrees higher than the outside temperature under sunlight condi- tions. The temperature in the shade of the leaves of the check plants at 4:00 P. M. on September 9 was 96° F., while under the black cloth and cellophane cover similar readings averaged 97° F. in both cases. The covers were kept in place over the plants continuously until 11:30 A. M. September 13. Some of the leaves that touched the cellophane were speckled with brownish spots, as if scorched. Twelve small bolls and 30 squares were shed from the three black-cloth-covered plants from September 13 to 17. During this same 6-day period, the three cellophane-covered plants shed only 2 small bolls in all and there was no shedding from the checks. The checks and the cellophane treated plants all shed a few small bolls from September 28 to October 7 but only two small bolls were shed by the black-cloth plants in this period. On the latter date, the check plants had 12 good-sized bolls per plant and showed 69 percent-set. The cello- phane-covered plants had 7 bolls per plant and a 42 percent-set. Those "was used in the fall, winter, and spring experiments. 80 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION under black cloth shade had only one boll per plant and these plants showed by far the poorest set (8 percent). This small-scale test indicated that the low-light-intensity effect of the black-cloth shade was the im- portant factor in shedding under these conditions. The somewhat fewer bolls and lower percent-set for the cellophane-covered plants as compared with the checks may have been due to the slightly higher temperatures involved and to the leaf injury associated with the cellophane cover. Mild Shade Treatments Followed by ClearWeather In a few instances in the course of this study, the effects of shedding caused by experimental treatments have not been severe enough to cause significant differences in final yields of bolls, and in a few cases even beneficial effects of such treatments have been noted. For example, half of a series of forty-eight plants (made "up of six varieties, Stoneville 2B, Coker 100 Str. 5, Rogers Acala, Deltapine 14, Mebane, and Delfos 531 B) Fig. 21. Cotton plants growing in 2-gal. jars ‘of sand. The black cloth shade as arranged here caused considerable shedding. However, cloth around the sides of the shaded area also was found necessary to keep out reflected light which on bright days was sufficient to main- tain an intensity‘ above that recorded for cloudy weather. The greenhouse in background Below, a view of outdoor shade from a nearby building. r... . .. Mp1 o mp, Fwy”, , FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 811 _ in 2-gal. jars of sand were shaded for two periods, June 24 to 27 and July 7 to 12, 1942. The shade consisted of a flat roof of black cloth just above the tops of the plants (fig. 21). There was no shade cloth around the sides * of the group of shaded plants. Consequently, the light intensity from above under the center of the shade at noon on a clear day was about 400 f.c. dWith the target of the light meter pointed towards the open sides, however, F between August 2 and 10. i between 7 and 9 bolls per plant. a awhile the five treatments varied from 37 to 45 percent-set. lfhto boll weights, however, the check plants produced the greatest total weight of bolls at time of harvest. The average weight per boll on the intensities from three to six times this figure were obtained. Also during the second period of shade for five days, only about two and a half days were clear. Counts of forms that were shed during the first twenty-four hours following removal of the cloth afterithe second period of shade yshowed 6 bolls and 1 square to have been shed by the shaded plants, f on the average, and 2 bolls and 1 square per plant from each of the checks. Only four cloudy days occurred between the removal of the last l shade and the time of harvest, July 28. The data showed that there l was but little difference between the checks (22 bolls and 47 percent-set) and the shaded plants (20 bolls and. 43 percent-set). This difference was slight compared with the amount of shedding that had been found to result from the shading treatments. Another case of this sort occurred during the summer 1944 series, when 48 plants (twenty-four each of Stoneville 2B and Rogers Acala) in 2-gal. ‘_. jars of soil were divided into six groups and different shade treatments given to five of the groups by placing them in the shaded section of a greenhouse (light intensity about 2000 f.c. at noon) for various periods The treatments resulted in marked shedding at the time. After the plants were replaced outdoors the weather re- mained mostly clear up to the time of harvest on August 21, when the check plants averaged 7 bolls per plant and the shaded plants varied The percent-set for the checks was 35 In regard “check plants was 23 grams while similar averages for the five treatments .1 varied between 14 and 2'0 grams per boll. There was no significantdif- ference between the fresh weights of the plants in the different groups at time of harvest. Excessive Shedding by Plants with No Boll-Load IEarly in June, 1944, eighteen 2-months old cotton plants (three each 7 of six varieties) in 2-gal. jars of sand were placed inside an unshaded greenhouse. The sand mixture in which these plants were growing con- tained a too large proportion of fine particles and the plants suffered from wilting to a serious extent as the water requirements were high under these conditions. Only one of the plants died, however, and the others "ac-apparently became partially adjusted to these conditions. Excessive shed- " ding of the fruiting forms occurred and on July 21 the number of (im- mature) bolls per plant varied from none to eight. At this time all bolls, 82 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION ._ flowers, and large squares were removed and the plants were returned outdoors. It was intended that these plants should be used to furnish a number of small bolls, for another test. The plants made a large amount of new growth and soon many new fruiting forms were developed on all of the plants. Of the 142 blooms that were tagged between July 25 and August . .|_ M-.. 5, only 44 (31 percent) remained on the plants past the small-boll stage. A Considering the fact that there were only two cloudy days (August 7 and 12) and two partly cloudy days (August '5 and 6) between July 25 and August 17, this percent-set is unusually low for plants outdoors in the summer. These particular plants, furthermore, had no maturing bolls to make a load on the plants at this time. Apparently, maintenance of the rapid growth, over a relatively long period following removal of the plants from the greenhouse, utilized most of the carbohydrates synthesized during this period. QUANTITATIVE CARBOHYDRATE ANALYSES OF COTTON PLANTS UNDER VARIOUS CONDITIONS AND TREATMENTS In the course of this study involving the effects of light, soil moisture, and temperature on shedding in cotton, considerable analytical work has been carried out to ascertain the variations in carbohydrate contents of plants subjected to the different treatments. Since all three of the above factors are known to aifect photosynthesis and since variations in the shedding rate have often been attributed to variations in food reserves within the cotton plant, it has seemed necessary to measure the carbo- hydrates in cases where increased shedding rates were obtained. In this work, emphasis has been placed on the carbohydrate content of the leaves since a low-carbohydrate leaf content was found in preliminary tests to be correlated with increased shedding rates. However, other parts of both check and treated plants such as bolls, stem bark, tap-root bark, and tap-root woody tissue, were also analyzed. In practically all cases, the samples were selected to show the immediate effects on the plants within one to six days after initiation of the treatment. The analyses included only a determination of so-called “total carbohydrates” as shown by the reducing power of a hydrolyzed sample of dried and ground plant tissues. In collecting the samples of leaves, care was taken to select at the same time like numbers (usually from 2 to 4 per plant) of comparable leaf blades from corresponding positions on treated and nontreated plants. The fresh weight of the leaf material was recorded and after drying for several hours at 70° to 80° C. in a forced-draft oven the samples were weighed again to obtain the dry weight. The larger veins were then removed and the dry leaf blades were ground in a hand mortar, after which the leaf powder was passed through a 40-mesh screen. One-gram samples of the dry powder were placed in 125 ml. Florence flasks to each of which 50 ml. of 2 percent hydrochloric acid were added and the flasks FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 83 were placed in a boiling water bath for 30 minutes. The flasks and contents were then cooled, neutralized with sodium hydroxide (using the natural red color in the hydrolyzed solution as an indicator), cleared with neutral lead acetate and deleaded with potassium oxalate. The filtered, clear filtrate was made up to 100 ml. Ten-millileter aliquots of this extract were then added to 20 ml. of a modified Fehling’s solu- tion made up of equal parts of Solution A containing 12 grams per liter of copper sulphate and Solution B containing 44 grams of Rochelle salt and 16 grams of sodium hydroxide per liter. This mixture was heated in the boiling water bath for five minutes and the unreduced copper .was determined by titrating the free iodine, obtained upon addition of po- tassium iodide, with a dilute solution (about 12 grams per liter) of sodium thiosulphate, using soluble starch solution to make the end-point more definite. Each supply of Fehling’s solution was standardized with anhydrous dextrose. The total carbohydrate contents (as anhydrous dextrose) shown in the tables are expressed in percentages of dry weight of plant material. This method is similar to a procedure previously used in a study (15) of the carbohydrate content of plants affected with certain virus diseases in which a 2-hour period of acid hydrolysis was employed. The method has provided a convenient and relatively quick estimation of the carbohydrate content of certain plant parts and it has been sensitive enough to give progressively highs; readings for cotton leaves collected from morning to afternoon. The values reported in this study should not be considered as exact quantitative measurements of the absolute amount of the carbo- hydrates present. Using dry leaf powder from the same samples, the total carbohydrate content has been found considerably lower by this method than by an alcohol-extraction and diastase method.* The same relative differences between samples, however, were shown by either method. The carbohydrates measured in this way probably include the reducing sugars, disaccharides, starch, and dextrins, together with intermediate and closely related compounds. Analyses of the bark and woody portion of the stem (sections about six inches in length taken just above the soil) were also made in cases of many treatments. These analyses were made as in the case of the leaf material except that the woody part of the stems was ground in a food- chopper with fine cutter before drying and the bark samples were ground finely in the food-chopper after drying. The bark and wood samples were not passed through the sieve. A few analyses of young bolls were also made, and in these cases the bolls were cut and dried and the part that could be finely ground in a mortar was sieved out for analysis. In all cases the samples under comparison were collected at the same time and those from treated and check plants were carried through the analytical procedure side by side under the same conditions. *The comparative analyses we're made by Dr. D. R. Ergle. 84 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Elfects of Shade on Carbohydrate Content of Cotton Leaves and Stem Bark Although it is Well known that 10w light intensities are usually ac- companied by reduced photosynthetic activities in plants, a number of analyses of shaded cotton plants have been made here in order to corre- late shedding tendencies with variations in carbohydrate constituents of certain plant parts. Some plants are known to require only a certain fraction of ordinary sunlight for maximum photosynthesis (48). With cotton, insufficient work has been done along this line to state the light intensities necessary for highest carbohydrate production. However, certain analyses of leaves that had been shaded with white cloth, that reduced the sunlight intensity about 70 percent, have shown fairly con- sistent but slight reductions in total carbohydrates. Black cloth that allowed passage of about 4 percent of direct sunlight reduced the carbo- hydrates of the leaves to a much greater extent. The results of several analyses of cotton leaves that were shaded during the forenoon are shown in table 15 in comparison with analyses of leaves from nonshaded plants. Additional data involving the effects of low light intensities on carbohy- drates in the leaves together with the results of analyses of the stem bark of the same plants are given in table 16. These samples were taken from cotton plants of different varieties and under various conditions as to time of year, temperature, age of plants, and shading treatment. These results indicate that reductions in light intensit from around 10,- 000 or 12,000 f.c. to about 3000 f.c. were associated with decreased carbo- hydrates in the leaves and stems. The most striking reductions in the carbohydrate content of both leaves and stems, however, was found in plants that had been subjected for a few days to light intensities of 500 f.c. or less. It may also be seen in these two tables that the carbohydrate Table 15. Elfect of white cloth and black cloth shade on carbohydrate content of cotton leaves Percent total Sample taken carbohydrates Date Size of Material in sample " ‘ planted container - Hour \Vhitc Black Date C. S. T. Check cloth cloth 6/16/41 2-gal. 10/1 3 PM 12 leaves—4 plantsl 6.5 4.4 4.1 6/16/41 2-gal. 10/6 1 PM 12 leaves—4 plants 4.7 5.1 2.6 5/20/42 l-gal. 7/13 1 PM 10 leaves—5 plants 8.6 9.7 5.9 5/20/42 l-gal. 7/13 4 PM 10 leavcs—5 plants 12.0 11.3 7.6 5/20/42 l-gal. 7/14 1 PM 5 leaves—5 plants 8.6 7.6 4.9 5/20/42 l-gal. 7/14 4 PM 5 leaves—5 plants 8.9 7.2 5.7 1/21/44 3-gal. 4/27. 3 PM 2 leaves———2 plants 7.3 4 1 2.6 Average, all tests 8.1 7.1 4.8 Percent of check . . . . . . . . 88 59 lThree leaves from each of 4 plants, making a total of 12 leaves per sample. '4 ~1 ‘an mil .' _ _ as, 1 a; ~ 6/15/44 1 PM In laboratory room 1 day 5O I 6.6 I 1.7 I . . . . . . . . . . . . . . . . . . . . 6/15/44 10 AM In laboratory room 5 days 50 I 3. I 1.4 I 8.5 3.2 Planted 4/6/44, 4-gal. soil, 3 plants per treatment - . I 6/27/44 11 AM I In laboratory room 1 hour - I 5O I 4.8 I 4.5 I . . . . . . . . . . . . . . . . . . . . 6/29/44 10 AM I In laboratory room 2 days I 50 I 5.3 I 1.5 I . . . . . . . . . . . . . . . . . . .. 7/ 1/44 10 AM I In laboratory room 4 days I 50 I 5.4 I 1.1 I . . . . . . . . . . | . . . . . . . . . . 7/ 3/44 10 AM? l In laboratory room 6 days I 50 II 2.1 I] 0.0 II 17.1 I 3.3 Planted 4/6 /44, 4-gal. soil, 2 plants per treatment ' I 7/10/44 I 3 PM Shaded greenhouse 5 days I 2000 3.6 I 1 2 12.4 I 3.6 7144 4 /16/ Planted 4/5/44, 2-gal. sand, 3 plants per treatment Black cloth shade 1 d th 2 e 4 6 s 500 Planted 4/5/44, 2-gal. sand, 3 plants per treatment 7/18/44 4 PM Black cloth shade 1 day 500 2.0 0.0 9.2 3.4 7/19/44 4 PM Black cloth shade 2 days I 500 . . . . . . . . . . 0.6 . . . . . . . . . . 2.4 7/21/44 3 PM Black cloth shade 4 days I 500 3.6 0.0 | 4.4 2.0 7/24/44 9 AM I 500 I . . . . . . . . ..I 0.2 I . . . . . . . . ..I 1.4 Black cloth shade 6 days Planted 6/12/44, l-gal. soil, 6 plants per treatment Black cloth‘ shade 1 1 98 {LHDPI 0L NOILVTEIH NI NOLLOO DNICIGEIHS CINV DNLLIIIH 86 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION content of these plant parts may be lowered markedly within one day after application of the shade treatment. The results further indicate that mobile carbohydrate reserves may be quickly used by the cotton plant under low light intensity and they may require replacement regularly in order to maintain a normal level. I Analyses of plants under the continuously low-light-intensity conditions previously described showed lower carbohydrate contents in both the leaves and stem bark, both in the morning and afternoon samples, as compared with plants in direct sunlight. Efiects of Wilting on Carbohydrate Content of Leaves and Stem Bark ’ The results of carbohydrate analyses of plants that were subjected to various wilting treatments are given in table 17. Periods of wilting from 1 to 8 days were in practically all cases found to cause a marked reduc- tion in the carbohydrate contents of the cotton leaves. This reduction was found to occur quickly following the treatments, as shown by the results of analyses at the end of the first day of wilting. In most cases, analyses of leaves wilted for two to four consecutive days showed still greater reductions in carbohydrate contents. The fourteen leaf samples from wilted plants showed (table 17) an average carbohydrate content 54 percent less than the average content of the nonwilted leaf samples. The only leaf sample which did not show a decrease in carbohydrate con- tent that might be\ attributed to the wilting treatment was taken on July 14, 1944, when a maximum temperature of 100° F. was recorded. On the average, the samples of stem bark showed a very slightly smaller total carbohydrate content in the wilted plants as compared with the checks. In many cases, however, a somewhat higher content was found in the bark of the wilted plants, indicating that the reduction found to have occurred in the carbohydrate content of the leaves did not extend to the bark of the lower stem’. The average for all analyses in table 17 shows a decrease of 8 percent in the bark of wilted plants that might be due to the wilt treatment. ' Effects of High Temperature on Carbohydrate Content of Leaves and Stem Bark Analyses of cotton leaves taken from plants on clear days in the sum- mer in an unshaded greenhouse showed marked reductions in the total carbohydrate content as compared with samples from similar plants kept outdoors. Under these conditions, the temperature in the greenhouse was usually 10 to 15 degrees higher than comparable readings outdoors and the samples were taken on days when the maximum greenhouse temperature was about 108° F. In the tests made during the spring of 1945, the high-temperature plants were kept in a partially closed greenhouse where the temperature remained between 100° and 110° F. with bright sunshine. “These results were compared with those from other plants of a~' ‘YET FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT t: , . \ 87 317.. Efiect 0f wilting on carbohydrate content of leayes and stern bark of cotton plants. Percent total carbohydrates fple taken I , j? Wilting treatment J Leaves Stem bark our “ C. S. T. i Check l Wilted Check Wilted Planted 4/6/44, 4-gal. soil, 2 plants per treatment 3 PM Wilted once a day for 5 days ll 3.6 II 2.1 12.4 7.4 i” 3 PM Wilted continuously for 5 days I 3.6 I 1.8 12.4 12.1 ~» | Planted 4/6/44, 4-gal. soil, 2 plants per treatment Wilted continuously for 3 days I 4 PM II 7.6 1.2 I 11.2 8.0 .1‘ Planted 4/5/44, 2-gal. sand, 3 plants per treatment ‘LP 3 PM Wilted 3 days (high daily T.) 2.2 2.6 9.4 9.2 i Planted 4/5/44, 2-gal. sand, 3 plants per treatment 3 PM Wilted 3 days (high daily T.) 3.6 0.0 4.4 6.0 Planted 6/12/44, l-gal. soil, 6 plants per treatment E44 4 PM Wilted 1 day 9.1 5.3 11.1 11.4 3 PR4 Wilted 2 days . . . . . . . . 1.9 . . . . . . .. 10.9 4 PlVI Wilted 4 days 7.7 3.9 12.4 9.3 ‘ Planted 4/5/44, 2-gal. sand, 2 plants per treatment 4'44 s AM Wilted 4 days 2.7 1.1 9.0 12.2 z . Planted 4/5/44, 2-gal. sand, 2 plants per treatment 4 PM i Wilted 3 days l 4.1 \ 1.7 l 12.8 11.9 Planted 4/5/44, 2-gal. sand, 3 plants per treatment 144 s AM Wilted 5 days 1.9 1.1 9.8 13.2 ’ Planted 4/6/44, 4-gal. soil, 2 plants per treatment 1/44 3 PM Wilted 4 days 6.6 1.7 17.1 11.9 '[44 9 AM Wilted 8 days 1.4 1.7 8.0 10.1 Planted e/12/44, l-gal. SO11, 4 plants per treatment 10 AM Wilted in greenhouse since 7/25 3.4 1.6, i 9.5 6.6 I (mostly cloudy days) ‘ Planted 4/5/44, 2-gal. sand, 4 plants per treatment ‘F144 4 PM Wilted 2 days 6.1 3.7 I . . . . . . . . . . . . . . . . ‘[44 8 AM Wilted 2 days 5.4 2.1 7 6.2 I 88 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION the same lot that were kept in a well-ventilated greenhouse in which the temperature rarely exceeded 85°} to 90° F. In most cases where the plants were kept under high-temperature con- ditions for several days (summer 1944), a decrease was found in the car- bohydrate content of the leaves and stem bark (table 18). Additional data on the effects of relatively brief periods of high temperature on the rela- tive increase in carbohydrate content of cotton leaves during the fore- noon are shown in table 19. These data show a higher content in leaves from plants in a ventilated greenhouse than in those from comparable plants in the high-temperature greenhouse. The increase in carbohydrate content during the day under high-temperature conditions was less in the older plants (4-gal. soil cultures) that were nearing maturity than in the younger sand-culture plants. In this table, it may be seen that the carbohydrates in the soil-plant leaves were somewhat more than doubled during the forenoon on clear days under normal greenhouse temperatures. At the same time, similar leaves from plants in the high-temperature greenhouse showed only a 44 percent increase over the carbohydrate content found in the morning samples. With the sand-culture plants, the leaves from check plants increased 83 percent and those under high tempera- ture increased 68 percent in carbohydrate content during the forenoon. In the caseof both soil- and sand-grown plants, the younger leaves (near top of plant) appeared to be inhibited in carbohydrate accumulation under high-temperature conditions to a lesser degree than the older leaves (farther down the stem). Table 18. Eflect of high daily temperature for an extended period on total carbohydrate content of leaves and stem bark of cotton plants. Sample taken Percent total carbohydrates Treatment l I l Leaves Stem bark Date Hour C. S. T. Check l High T. ll Check l High T. Planted 6/26/44, 4-gal. soil, 2 plants per treatment I I I I 9/15/44 4 PM P In greenhouse after 8/23 ‘I 11.5 ‘I 3.4 i] 13.3 || 9.8 Planted 6/24/44, 2-gal. sand, 2 plants per treatment 9/15/44 4 PM In greenhouse after 8/23 6.1 I 3.9 i 12.7 5.5 Planted 7/5/44, l-gal. soil, 2 plants per treatment I 9/15/44 4 PM In greenhouse after 8/23 I 10.8 3.8 g 10.4 11.8 Planted 6/26/44, 4-gal. soil, 1 plant per treatment I 9/25/44l 4 PMl In greenhouse after 9/21 | 3.9 I] 1.1 1| 6.5 ll 4.8 Average, all tests l 8.1 l 3.1 | 10.7 ll 8.0 Percent of checks 38 | . . . . . . ..| 75 hmltlllm ‘Y Pr‘ w,» Ivyw-v qr"- . r lStem leaves except as noted. ‘~’Plants 1n well-ventilated greenhouse. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 89 Table 19. Total carbohydrate content of cotton leavesl following exposure to high temperatures in a partly closed greenhouse for a few hours. < Percent total Sample taken carbohydrates Plants and samples Date Hour 1945 C. S. T. Check? High T. Plants in Soil 3/9 9 AM 5.8 5.3 Planted 12/5/44, 4-gnl. soil, 3 plants per treatment, p a 2 leaves per plant per sample 3/9 3 PM 9.5 5.0 3 /22' 9 AM 4.9 4.5 Upper -- leaves 3/22 1 PM 7.7 8.5 3/22 9 AM 1.5 1 .9 Middle leaves 3/22 1 Phi 2.0 4.0 ~ 3/22 9 AM 0.6 0.8 . Planted 12/5/44, 4_gal. soil, 6 plants Lower - per treatment, 1 leaf per plant leaves 3/22 1 PM 2.9 0.7 per sample 3/26 8 AM 2.6 4.0 Upper - . leaves 3/26 1 PM 9.2 5.6 Middle 4/2 s AM s11 2 .3 fruiting . . branch l 4/2 1 PM 8.5 5.0 4/2 8 AM 4.8 5.6 Middle leaves 4/2 1 PM 7.5 7.3 _ Average, morning samples 3.3 3.6 Average, afternoon samples 6.8 5.2 Increase during day—-percent 106 44 Plants in Sand ‘ 4/2 8 AM 5.0 3.7 Upper ' leaves 4/2 . 1 PM 8.2 8.8 4/2 8 AM 3.4 , 1.6 Middle leaves 4/2 1 PM 5.8 3.2 Planted 1/5/45, 2-gal. sand, 4 plants per treatment, 1 leaf per plant 4/4 9 AM 3.2 4.9 per sample Upper leaves 414 2 PM 6.8 5.9 4 /4 9 AM 2.5 3.6 Middle leaves 4/4 2 PM 2.8 3.5 4/6 8 AM 3 .4 3.1 Middle leaves 4/6 2 PM 8.6 7.1 Average. morning samples 3.5 _ 3.4 Average, afternoon samples 6.4 5.7 Increase during day—-percent 83 68 I I 90 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Increase During the Day in Carbohydrate Content of Leaves of Dilferent Ages In view of the high rates of shedding that are usually observed in cotton plants that have already set a number of bolls and that are ap- proaching maturity (late in the fruiting period), a series of carbohy- drate analyses was planned which might indicate the relative photo- A synthetic activity of the old and young leaves on the same plant at this ‘I stage of growth. crease in amount of carbohydrate materials accumulated during the fore- noon. (young) and middle (older) leaves. ternoon than in the morning while the content of middle leaves on the same plants increased 7 percent during the same period. Since the ' leaves compared were from the same plants and under similar environ- In order to compare leaves of different ages, those near the top of the stem were compared with lower stem leaves as to the in- I Table 2'0 shows the results of fourteen comparisons between upper Considering all samples, the car- bohydrate content of the upper leaves was 37 percent greater in the af- mental conditions, it would appear that the greater age of the middle stem leaves was an important factor in the apparently lower photosynthetic activity. It is of interest also to note in table 20 that the leaves from the oldest plants, particularly those in soil culture which had begun to show yellowing of the foliage, were found to have a high carbohydrate content and these plants translocated only a small part of the carbohy- drates from the leaves during the night. The carbohydrate accumulation in the soil-grown plants (table 20) oc- curred shortly after most of the shedding of the last small bolls had taken place, at the time the first bolls were starting to open. Similar accumulation of carbohydrates was also found in younger plants (early summer 1945) in soil culture that were in a stunted condition and had ceased growth about the time the first blooms opened. The leaves 6f these relatively young plants showed a carbohydrate content around 11 percent both in the morning and afternoon. This accumulation of large amounts of carbohydrate material in the leaves of the old soil-plants _‘ may have been due, at least in part, to the depletion of nitrogen from the soil. However, a somewhat similar condition was found in the sand- culture plants (May 2 samples, table 20) which were still receiving the usual daily applications of nutrient solution.~ In this connection, samples taken on May 24 of upper leaves from plants of this same sand-culture series, which were placed outdoors on May 8 and had dormant terminal buds since that time, showed a very low carbohydrate content both in the morning (3.2 percent) and afternoon (3.5 percent). At that time (May 24) about half of the bolls on these plants had opened and no new vegetative growth was apparent. This would indicate an unusually low photosynthetic activity in the leaves of these old plants. By June 15, small new leaves had developed on these plants as secondary growth and- l‘ these leaves showed a carbohydrate content of 4.6 percent in the morning and 8.0 percent in the afternoon. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 91 file 20. Total carbohydrate _content of upper and middle stem leaves of cotton plants, 1n the morning and afternoon. Percent total Sample taken carbohydrates Plants and samples Date Hour Upper Middle 1945 C. S. T. leaves leaves _ _ 3/27 8 AM 14.2 8.9 FPJanted 12/5/44, 3-gal. soil, 6 plants, 1 leaf per plant 3 /27 1 PM 18.5 11.6 , per sample. (Plants yellowish) 4/13 8 AM 8 3 8.3 4/13 3 PM 8 O 8.0 3/22 9 AM 4 9 1 .5 Plants green 3/22 1 PM 7.7 2.0 4 /3 8 AM 5 3 5.7 Plants green 4/3 3 PM 6 9 7.0 4/6 8 AM 5 6 4.7 Plants lanted 12/5/44, 4-gal. soil, 6 plants green 4/6 2 PM 11.2 6.8 per treatment, 1 leaf per plant per sample 4/6 8 AM 8.8 6.9 Plants yellowish 4/6 2 PM 17 .8 6.5 4 /13 8 AM 12 .4 8.6 Plants green 4/13 2 PM 15.7 10.0 4/13 8AM 19.3 16.0 Plants yellowish 4 /13 2 PM 21.0 20.9 4/2 8 AM 5.0 3 .4 4/2 1 PM 8. 8 5.8 4 /4 9 AM 3.2 2. 5 _ , 4 /4 2 PM 6.8 2.8 ' 1 ted 1/5/45, 2-gal. sand. 4 plants per treatment, * 1 leaf per plant per sample, in greenhouse 4/23 9 AM 3.7 4.9 4/23 3 PM 6.5 3.6 5/2 8 AM 5.5 6.2 5/2 4 PM 6.4 7.8 4/23 9 AM 6.8 6.5 anted 1/5/45, 2-gal. sand, 4 plants per treatment, 4/23 3 PM 7.7 5.9 1 leaf per plant per sample; plants outdoors 5/2 8 AM 9.5 9.1 5/2 4 PM 10.6 9.7 Average, morning samples 8.0 7.2 Average, afternoon samples 11 0 7.7 Increase during day-percent 37 7 Analyses of Leaves from Plants Grown in the Greenhouse During the Summer Leaves and stem bark of plants kept in the unshaded greenhouse in e summer for the duration of the fruiting period showed a medium to 92 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION high carbohydrate content in both morning and afternoon samples, in the few samples taken during the 1944 test. These samples were taken near the time of harvest when the oldest bolls were starting to open. Analyses of Other Parts of the Plant in Relation to Shedding Stem bark vs. root bark. In case of some of the early analyses in this work, total carbohydrate determinations were made of both lower-stem bark and root bark. The carbohydrates in these two materials were found to be quite similar quantitatively and they were also found to vary to practically the same extent following shading treatments of six-days duration. Consequently, the taking of root-bark samples was omitted from this work and analyses were made only of the bark on the lower stem, 5 to 7 inches above the soil line. Woody portion of stem. Preliminary analyses of the central cylinder (wood) of the lower stem and taproot of cotton plants used in these ex- periments showed the carbohydrates in this part of the plant to remain relatively stable during the few-days shading treatments that were ap- plied to induce shedding. These analyses were therefore discontinued in favor of the stem-bark data which were found to undergo changes in carbohydrate content much more rapidly than the woody parts of the stem. ~ Young bolls and bracts. Many analyses of young bolls, from a few days after flowering up to formation of, fibers strong enough to interfere with cutting by a knife, were made to ascertain any possible correlation be- tween carbohydrate contents of the boll and shedding tendencies. Shading for a few days was found to cause distinct reductions in the carbohydrate content of the bolls in many instances. With wilted plants, no significant differences were found. It was also difficult at times to compare bolls of the same age on wilted and nonwilted plants, due to the tendency of those on the wilted plants to ripen prematurely and thereby resemble physiologically bolls on the check plants that were much older. In gen- eral, the variations in carbohydrate content of the young bolls were not as consistent, in relation to the shedding responses to the treatments, as the variations in carbohydrate content of the leaves. These results indicate that ‘the amount of carbohydrate synthesized by the plant and available for the competitive utilization throughout the plant is one of the most important internal conditions concerned in shedding. Young bolls (about 5 days after blossoming) that were shed from full- grown nontreated cotton plants were found to have a considerably lower total carbohydrate content (6.4 percent) than comparable bolls (11.0 per- cent) that remained on the same plants. Similarly low carbohydrate con- tents were found in small bolls that were caused to be shed by shading of the plants. In one test, no difference was found between the carbohydrate contents of small bolls that wereshed from wilted plants as compared with those picked from nonwilted (check) plants; while at another time FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 93 the shed bolls showed only a little over one-half the amount (3.5 percent) present in unshed bolls. At the same time, bracts were removed from half-grown bolls (on both check and wilted plants) that were not shed 1 but-no difference was found in the carbohydrate content of these bracts (3.'l"and 3.9 percent, respectively). The carbohydrate content of the in- A terior of these bolls was 24.8 and 25.2 percent, respectively. DISCUSSION In this study, variations in the amount of light in the environment of the cotton plant have been associated with profound differences in fruiting behavior. In most cases, marked difierences in vegetative growth l. have accompanied changes in fruiting. The effects on fruiting processes Q have been seen in various ways but chiefly in the number of fruiting p. forms developed on the plant and in the relative number of forms that " have been shed prior to maturation of the bolls. Evidence’ is presented which ; indicates that periods of cloudy weather during the fruiting period consti- tute a hazard in the production of the cotton crop. In many instances, {iv-varietal differences have been found among the responses of cotton to these variations in the light factor. Recognition of this inherent quality by the iplant breeder may lead to the improvement of commercial strains so that p, they may be better adapted to specific growing conditions. Of the three factors (light, drought and heat) that have been studied as ___causes of shedding in cotton, light has appeared the most important, under ,the conditions of this work. Considering the large amount of shedding (induced and the apparent lack of serious concurrent damage to other parts ‘ of the plant (such as loss of leaves), periods of low light intensity appear to be a natural and highly effective cause of shedding. In comparison with high-temperature treatments, shedding brought about by low light P intensity has been more sudden and of greater intensity (fig. 22-A). High shedding rates, similar to those found under poor light conditions, have been caused by wilting (fig. 22-B), although plants that have been wilted severely enough to cause considerable shedding usually show a certain amount of leaf damage that was not apparent in shading experiments. In Nigeria, Thornton (52) found that shedding during the dry season was accompanied by loss of leaves while no loss of foliage was associated with shedding during the Wet period. It is particularly significant, too, that most I I of the treatments and conditions that involved impaired light conditions (shading, fall and winter conditions, close spacing, and shortened day length) resulted similarly in increased shedding rates and impaired fruit- ing capacity of the plant. ._ As "shown by carbohydrate analyses, particularly of leaf material, a fhpossible relationship is indicated between the three common causes for (shedding studied here-—light, drought, and high temperature—and the “ synthesis or utilization of carbohydrates by the plant. The carbohydrate content of leaves, stem bark and root bark was found to be depleted when 94 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION the plants were ‘kept under low light intensities for a few days. Wilt.‘ ing was also found to lower the photosynthetic ability of cotton leaves, as shown by the relatively small increase in the carbohydrate content of leaves on wilted plants. The apparent (often slight) increase sometimes found in the carbohydrate content of the stem bark after a period of wilting may have been due to interference with sugar movement, especially since wilting of greenhouse plants appeared to cause a measureable de- crease in diameter of the stem. r Although the carbohydrate content of cotton leaves was found in some cases to increase during the first hours of exposure to abnormally high temperatures‘; considerable depletion of carbohydrates was noted in the leaves, and to a lesser extent in the stem-bark, of plants that had been IlllllllllllllllllllTlllllll1111111 45 — — S ‘° “ ' CHECK f, as - ._ Q so - ' - 5 25 __ s11»: s DAYS a i ‘5 2o - I w I5 - 2 a IO — _ z _ -._ 5 - _ ' 1 1 1 1 1 1"'1-+-1~.1/1’ - - 1 1 1.1 l 2o 22 24 26 2e 3o 1 a 5 7 9 11 13 1s 17 19 21 2a AUGUST SEPTEMBER 1944 a 50 - CHECK ---------- --' ---- -- "wlurzo FOR THREE oAYs-n- SHADE THREE oAYs—--- - I - . a .. q, SHADE ‘I1 WILTED asron: , m 4Q _. ,\ -., EACH WATERING .. 2 —> AND s- ‘l '~._ g - WILT , \ ' - 1.1. 30- , l. - u- I-' an Q . a: .20 —- / _ Q : 2 " / ' g IO -- / -l l l I I l l l 1 l5I7I92l23252?293l2468lOl2l4l6 AUGUST SEPTEMBER ‘ I944 Fig. 22. Comparisons of shedding curves following shade, high temperature, and wilting treatments. A. Three curves each representing 8 plants in sand culture. Note the increase in early shedding due to the 5-day shade and high temperature (greenhouse) treatments, both applied on August 22-27. B. Each of the four curves represents the average of 7 plants in soil. Each of the three treatments (wilted for 3 days, shaded for 3 days, and wilted for a few hours between each watering after August 16) shows a high rate of shedding before the check plants (dotted line) had shed any forms. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 95 dept in an unshaded greenhouse for several days during the summer (table ). In this connection it should be noted that only moderate increases .1 shedding rates due to high temperatures were found, in comparison T th the effects of shade or wilting. Likewise, the reduction in the ruiting capacity of high-temperature plants, in comparison with lower- mperature controls, has not been especially severe. Possibly, high temperatures may promote more rapid vegetative growth and also a more. pid respiratory rate, as suggested by Wadleigh (55), which could result gs:- a severe drain on the carbohydrate reserves and which may have been ‘responsible for the more severe shedding and smaller bolls reported here. has been shown by Armstrong and Albert (5) that removal of leaves gErom the fruiting branch also reduces the size of cotton bolls. _ Under field conditions, it would be difficult to differentiate between the ieffects of cloudiness and those of too high rainfall upon shedding. The purves presented by Harland (26) for cotton in the West Indies show that ‘avy shedding took place during and following heavy rainfall in late ctober and early November and he suggested that root asphyxiation , ay have caused the peaks in the shedding curve. On the other hand, arland noted that flower-bud shedding occurred with plants under con- olled moisture conditions in containers at the same time as with plants der field conditions. Experimental flooding of soil, however, has not roduced sufficiently intense shedding in most cases to support adequate- i the soil-saturation theory. In cases where increased shedding rates ave been obtained by flooding, most of the shedding .has occurred in e young square stage (2, 43). In this connection, it should be re- mbered that high rainfall does not necessarilyinean“ excessive cloudi- ess, nor vice versa. For example, Canney (11) has pointed out that e northeast coast of Brazil has four times as much rainfall an- ually as the coastal region of Portugese East Africa where the “annual oudiness” is twice as much as in northeast Brazil. For practical pur- >= - es, however, it is certain that cotton in areas of fairly high rainfall nditions than would be encountered in areas of low rainfall. +In the case of high temperatures, it may not always be possible to dif- rentiate between heat effects and those of wilting, since water loss from e plant is much greater at unusually high temperatures than under rmal conditions. Although wilting is not apparent, a plant under high mperature conditions may suffer effects similar to those- obtained from ble wilting, due to the water-stress within the plant. Hastings (27) g4? stated that the unfavorable effect of high temperatures is aggra- ated by low soil-moisture contents. Schneider et al. (46). found with pple leaves that a marked reduction in photosynthesis and transpira- on, together with an increase in respiration, took place before wilting .as evident. Heinicke and Childers (31), working with apple leaves, ound a gradual drying of the soil to be accompanied by an appreciable uction in the rate of transpiration and of photosynthesis. Alekseev (3) ported a decrease in CO2 absorption by apple leaves when the fresh .1 ring the growing season is subjected to many more fluctuations in light, 96 Fig. 23. Defoliated cotton plants (spring series in greenhouse, 1944) showing differences in growth and fruiting between the following treatments: A. heck, B. Short day (8 A. M. to 3 P. M.), C. Close-spaced, D. Shaded 4 days early in fruiting period. Six varieties are shown in each series, from left to right: Sto-neville 2B, Rogers Acala, A. D. Mebane Estate, Delta- pine 14, Qualla, and Half and Half. Note the better set of bolls in checks than in any of the treatments. Also, the shade resulted in larger plants (D) with many more young forms at time of harvest. u‘. \ "44l‘2li4 m! I‘ n! iVimlv ‘minim. .14‘ n; FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 97 ‘l’ weight of the leaf was reduced below 80 percent of the maximum moisture content. ,_ The well-known tendency for cotton plants to shed many young bolls ; late in the season, usually after a crop of bolls has been set and many i of the bolls are fairly mature, deserves special mention in connection with these studies. At this time, the maturing bolls probably make a heavy 1 demand for available sugars. As shown by the carbohydrate analyses in table 19, older leaves appear to be less active in carbohydrate manufac- 5 ture than young leaves. A decrease in the total carbohydrate-s in tobacco was found by Dennison (14) as the leaves became older. Since there is 4 usually but little leaf growth after the plant has set a fair load of bolls, 1 most of the foliage has become well advanced in age by the timethis late- l season shedding becomes prominent. Evidence is also shown in table 19 that photosynthesis in the older cotton leaves was more likely to be im- f paired by high temperature than was the case in younger leaves, as shown i by the smaller carbohydrate accumulations in old leaves during the day. F As may be seen in figs. 5-C, 16-A, and 19, sharp peaks of shedding may l not always occur as the plants reach maturity. Apparently, cotton plants ‘l1 do become more sensitive to external shedding stimuli at this time, how- t; ‘ever, (figs. 3-A, 3-B, and 5-B). As commonly noted, young bolls, a few days after flowering, are more l, susceptible to shedding than fruiting forms in earlier or later stages of ' development. However, an increase in the severity or length of a treat- ment that induces shedding has been found to result in abscission both ‘a of younger (squares) and older cotton forms (half-grown bolls). When l1 outdoor plants in the summer were placed in the laboratory, a relatively f, large number of squares was shed before the peak of shedding of the small squares was reached (table 9). In some of the greenhouse experi- ments during the winter, however, considerable shedding of the smallest squares (up to 5 mm.) took place along with the loss of the small bolls. At the same time, a smaller number of older squares was shed. These i1 results indicate further a common cause for shedding of forms of all ages, from the smallest square up to the partially matured boll. Bailey and Trought (6) expressed the opinion that those conditions which caused shedding of the smallest flower buds induced buds of other sizes to shed A also. It has been noted that the most damaging results, insofar as final set of bolls is concerned, followed the shading and other low-light-intensity treatments that were applied fairly early in the fruiting period (table 8).- In the case of this and similar experiments, however, the treatments , have been late enough to cause shedding of large squares and at least a few young bolls. As a result, it seems that shedding of the larger fruit- ping forms (half-grown squares or older) may be more devastating to the A fruiting activities of the cotton plant than shedding of small squares. As stated previously, during certain experiments under fall or winter green- F house conditions excessive small-square shedding had been recorded but 98 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION the yield of bolls from these plants was equally as good as from plan in comparable experiments Where shedding of these smallest forms not so prominent. This conclusion is in keeping with results of certa' studies in Mississippi Where Hamner'(25) found no reduction in yield from removal of small squares (about 3 days old) but where Dunnam et a (19) obtained a reduction in yield of cotton in proportion to the per? centage of larger squares (6 to 14 days old) removed. Possibly stimulus towards increased vegetative growth is more intense followin removal of large fruiting forms than in the case of the smallest squares. From a physiological standpoint, shedding caused ~by some environ? mental factor may not be at all analogous to the loss of fruiting form by hand removal or through insect injury. In case of the former, the shed» ding occurs as a result of a condition within the plant which does n; allow more fruit to be set at that time. Favorable conditions for frui ,1. ing would have to be restored before normal fruiting could proceed. I the case of mechanical removal or insect injury, on the other hand, th general condition of the plant would be as favorable for fruiting as i was before the loss of the forms. The plant would be able, therefori to proceed immediately with the development of the remaining form and any change in fruiting tendencies might not be at all serious. Ex. cessive shedding after a period of cloudy weather is a definite indicaii tion that the plant is physiologically unable at that time to set and; maintain additional fruit and its growth activities can be expected to be predominantly vegetative. A surge of vegetative growth might be initi-j, ated at this time and maintained for the remainder of the season to the detriment of the fruiting processes. It has been observed throughout this work that shedding does not“ always occur the same number of days after the causative treatment Shedding can be induced more easily with large vegetative plants, such as those in the greenhouse in the winter, than with outdoor plants in the» summer. This may be due to differences in carbohydrate reserves within the plant. As may be recalled, shedding from either low light intensity,’ or wilting has occurred only after one or two days of the treatment. This, allowed adequate time for certain food reserves to be depleted. In all; probability, one or more days would be required for the abscission layer?" to form and for the adjacent cells to separate, following the shedding stimulus. Lloyd (38) caused small bolls to shed within 24 hours by mechanical injury and “the more drastic the operation, the more rapid} the response.” In the case of bolls that were shed when about 14 daysii old, Lloyd allowed “a true response period of six to ten days as the ‘ap-gj proximate limits of requirement.” The time required for shedding, fol» lowing application of the stimulus, may also depend upon the age of th‘ plant since Ewing (23) noted that the “period of persistence” for bollél shed early in the season was from 9 to 16 days after flowering whereas, late in the season most of the bolls were shed about 4 to 6 days after; flowering. i FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 99 High carbohydrate contents of leaves and stems do not always pre- vent shedding from taking place, however, as shown by the relatively high contents of the stem-bark of plants following periods of wilting, by the fairly high levels of carbohydrates in cotton leaves under sum- mer greenhouse conditions and by the apparent reserve in older plants’ ‘ that show late-season shedding. In the latter case, the carbohydrates may be in a form unsuitable for translocation, or the effects of drought, high temperature, or age of vegetative organs may result in interference with the translocation process. Any factor that tends to inhibit a con- stant supply of soluble carbohydrates to the young bolls would most likely cause increases in shedding rates. The food "requirements of other parts of the plant, such as leaves, stems, and roots, may also play an important part in determining the proportional amount of carbohydrates that are made available to the bolls or squares. It is quite possible, also, that the older bolls (past the shedding stage) may compete for available sugars more successfully than the younger bolls. Some of the most serious effects, as far as shedding and reduction in number of bolls per plant are concerned, have been obtained in cases where a stimulus to shedding has been followed by another after several days. Evidence of this can be seen in table 6 and in figs. 13 and 22-A. Such recurrence of stimuli for shedding undoubtedly occur frequently under field conditions in the periodic occurrence of cloudy weather, dry E E? é periods between rains, or a few unusually hot days separated by cooler .5 periods. On the other extreme, the least noticeable effects upon shed- g ding and fruiting have been found after a single moderate treatment E 5 E- has been followed by favorable environmental conditions for the remainder _of the fruiting period. In some cases, even a beneficial effect on fruit- ing has been recorded following a single mild shade treatment; these effects may be similar to those noted by Eaton (21) following removal by hand of the first squares. The results obtained here, however, involve for the most a part only one or two shading treatments during the life of the plant. In ’ most crop environments, poor light conditions with low intensities un- doubtedly may occur much more frequently and with equally damaging , results each time. - -Working with cotton plants in a shaded greenhouse during the sum- I mer, Wadleigh (55) concluded: “after a plant had set a number of bolls suflicient to deplete its nitrogenous reserves, all subsequently formed young bolls abscised” and this was followed by abscission of young squares and abortion of the terminal buds on fruiting branches. In the ‘fstudies reported here, similar cases of high shedding rates during the Flatter part of the fruiting period have been recorded for plants in the summer outdoors, both in sand that was supplied constantly with a high- iuitrogen nutrient solution and in soil from which the available nitrogen must have been more or less exhausted. When the nitrogen content of Qthe nutrient solution was changed (from low to high or from high to glow), near the middle of the fruiting period, no marked changes in the 100 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION shedding curves were noted. ‘Further growth and fruiting of the plan however, took place in proportion to the amount of nitrogen suppli following the shift in solutions, indicating absorption of nitrogen by roots in proportion to the amount available. The increased yields th were obtained (table 14) by increasing the amount of nitrogen suppli were accompanied by renewed vegetative growth and development new foliage which may have made possible the increase in number of bo _ per plant. It,is of interest to note that, in spite of the sudden chan in amount of nitrogen supplied, the final yield of bolls varied direct with the amount of nitrogen and, even with the considerable differenc found in the amount of fruiting and size of plants, the final percenta of bolls set was remarkably alike for all treatments. In these studies, certain varieties of cotton, notably Stoneville Half and Half, Coker 4-in-1, Washington, Deltapine 14, and Roldo Rowd_ have been found to be relatively less sensitive to unfavorable light m ditions imposed in various ways than certain other varieties such I; Rogers Acala, A. D. Mebane Estate, Qualla, and Lone Star. The varieties have maintained higher fruiting capacities (indicated by nunf ber of bolls per plant and weight of these bolls) when performance und winter growing conditions was compared with summer productivenes Also, under shade treatment, Stoneville 2B was found to show bett/l fruiting tendencies than Rogers Acala. Similar differences were‘ note " between these same varieties, in regard to close-spacing and hig q temperature treatments, and variations in the total number of hours r?‘ sunshine per day but the data in these cases did not show significant difi ferences in interactions between treatments and varieties. However, y many cases the averaged data for a group of plants have shown differenc in favor of the first group of varieties mentioned above. In many cases, treatments used caused considerably more variation between individual plants than occurred among the- checks. Therefore statistical significanc~ could not‘ be demonstrated. No significant differences were found in th_ reaction of ten different cotton varieties (table 11, 5th part) to a single; period of wilting for five days, under greenhouse conditions; which might; indicate less inherent difference between these varieties in regard y‘ drought resistance than in regard to ability to withstand unfavorable light; conditions. As a result of these tests, it seems quite likely that the popularity among growers of some of the varieties, either old or new, may be due to the ability of the strain to withstand certain unfavorable; influences of poor light in the environment, that may occur under field; conditions much more frequently than the grower may suspect. ‘I In view of the evidence presented, it seems likely that weather con-:_ ditions frequently may affect the amount of sunlight available to cotton» plants under field conditions many cotton-growing areas to the extent, that excessive shedding may occur. Harmful shedding would most likeli occur in dense stands of rapidly growing plants such as often occur in fertile soil and favorable moisture conditions. When these hazards ar7 combined with those of drought, and when consideration is given to the‘. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 101 ger of shedding from high temperatures, it can be easily imagined that ewlcrops of cotton escape strong shedding stimuli, even before a fair l p of bolls has been set. Cotton that has received a favorable amount rainfall would undoubtedly suffer at times from poor light conditions jl~ to cloudy weather. Also, dry-land cotton in many regions must often ‘eel the effects of drought. Fields under irrigation are likely to lie in at -= of high daily temperatures. Optimum conditions for cotton fruit- j.» , therefore, may be obtained only rarely under field conditions. The use varieties that withstand unfavorable environments to‘ the highest de- e would appear most desirable and profitable as insurance against low "elds from these nonpreventable circumstances. The importance of lvorable light conditions for cotton, presuming that irrigated regions * ally have an abundance of sunshine, is implied in the statement by Todd 53): “Cotton was certainly at first a rain crop . . . . .; but thequan- Ty ofcotton grown under irrigation in every part of the world is now g large. Indeed, it appears that the best growing conditions can only » secured under irrigation.” i thas been estimated (45) that one more good boll on every stalk of n in four southwestern states, Texas, Oklahoma, Arkansas, and iana, would increase the annual yield of lint from this area by 0 bales. This result might be obtained byplanting more extensively higher-yielding cotton varieties that are least sensitive to certain un- orable environmental conditions and which have desirable fruiting ities. In the light of some of the experimental work herein re- l, it would seem possible to improve materially the yield of cotton er Texas conditions by planting varieties that are best adapted to . particular area in which they are grown. It appears important to con- n’ the effects of certain factors, especially variations in sunlight in- associated _with changes in weather, in selecting varieties of cot- ‘ebest adapted to a given region. SUMMARY . The cotton plant has been found sensitive to variations in amount "light as shown by vegetative growth and amount of fruit set under dif- t light conditions. - " Periods of low light intensity or short-day periods of full sunlight resulted in impairment of the fruiting capacity of the plant and an in- e in vegetative growth. ,3. These effects of light on fruiting and shedding in cotton have ap- yred more important than those of drought or high temporature. a Inferior fruiting and, in most cases, increased rates of shedding been found associated with fall-winter environmental conditions, ' g and close-spacing of plants, and reduction in number of hours v shine per day. 1 Periods of cloudy weather during the fruiting period or treatments 102 BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION involving similarly low light intensities for a few days have been f lowed by increased rates of shedding of fruiting forms. 6. These variations in the amount of light available to the plant h‘ been associated more closely with the relative number of bolls maintai _ to maturity than with the formation of flower buds (photoperiodism). 7. Certain varieties of cotton, such as Stoneville 2B, Coker 4-in, Deltapine 14, Half and Half, Roldo Rowden, and Washington, were foul‘ in general, to be less sensitive to unfavorable variations in light con 'tions than certain other varieties tested, such as Qualla, A. D. Meb Estate, Rogers Acala, and Lone Star. 8. Wilting of cotton plants continuously for a few days at a time w; found to cause excessive shedding. However, brief periods of wilting (7 to one or two days), although occurring frequently, resulted only‘ in cu“ tailment of vegetative growth, fewer and smaller bolls per plant, - usually no marked increase in shedding rates. 9. The addition of nitrogen to jars of soil in which cotton plants we" growing resulted in increases in yield of bolls when the soil moisture c l tent was adequate. With frequent wilting, or with soil of a continuous low moisture content, however, these benefits from extra nitrogen we not obtained. In most cases, no marked differences in shedding rates found between high- and low-nitrogen plants as a result of water-str treatments. 10. Certain treatments involving excessively high soil moisture ' sulted in increased shedding of cotton forms, especially when poor drai age was provided in jars of soil. Often, the high shedding rate was p ' ceded by wilting of the leaves and in some cases it was followedby dea_ of the plant. ' l 11. High shedding rates and decreases in weight of bolls per pla were also found associated with high temperatures (around 100° F. above), either during extended periods of high daily temperatures or fog lowing briefer periods of exposure to such conditions. i 12. Under outdoor conditions in the summer, the number of bolls p, plant varied in proportion to the amount of nitrogen supplied. Consiste = increases in fruiting, however, were not obtained from additional nitroge applications under fall-winter conditions in the greenhouse. ’ 13. Differences in fruiting resulting from the addition of extra pho‘ phorus or potassium (above adequate, basic amounts) in sand culture we usually small and the shedding rates were found to remain fairly unifo under conditions of widely different nutrient (N, P, and K) levels. 31 best fruiting was usually obtained with properly balanced nutrients. i’ 14. Certain treatments of the stem of cotton squares or flowers wit “chemical hormones” prevented formation of the abscission layer but di not prevent abortion of the young boll. 15. Wetting of open flowers twice daily with an atomizer caused on A slight increases in the percentage of young bolls shed. i FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT 103 16. No outstanding impairment of fruiting activity was caused by jremoval of bracts from squares, topping of plants, removal of leaves from fifruiting arms, and covering of squares with such materials as black cloth, ilnfoil, or cellophane. A 17. Chemical analyses of the leaves (and, in some cases, stem bark) lfrom plants that were subjected to conditions of low light intensity, ‘idrought, or high temperature have shown a low carbohydrate content or a low rate of photosynthesis to be associated with these “shedding-inducing treatments. ACKNOWLEDGMENT The author wishes to acknowledge the interest of A. B. Conner, formerly director, in the initiation and support of this work. The following have jbeen especially helpful in caring for the experimental plants: Norman G. l-"fDuren, T. J. McLeaish, Marion Wilkinson, E. A. Womack, and Sylvester iBteen. Valuable advice and aid in statistical analysis of the data have received from Dr. L. M. Blank, Professor C. B. Godbey, and Dr. E. Lyle. Dr. P. J. Talley and Dr. D. R. Ergle have assisted in the prep- firation of the manuscript. l LITERATURE CITED Adams,'Frank, F. J. Veihmeyer, and Lloyd N. Brown. Cotton irrigation investigations in San Joaquin Valley, California 1926 to 1935. Calif. Agr. Exp. Sta. Bull. 668. 1942. Albert, W. B., and G. M. Armstrong. Effects of high,soil moisture and lack of soil aera- tion upon fruiting behavior of young cotton plants. Plant Physiol. 6: 585-591. 1931. Alekseev. A. M. Influence of water content of leaves upon photosynthesis. Bot. Jour. U.S.S.R. 20: 227-241. 1935. According to Biol. Abstr. 11 (8) N0. 18225. 1937. Anonymous. Pinching the cotton plant. Editorial in Cotton Trade Jour. (New Orleans). > 21 No. 40: 2. 1941. Armstrong, G. M., and W. B. Albert. A study of the factors which influence the growth 11161510 development of cotton buds and bolls. S. C. Exp. Sta. 43rd Ann. Rpt. p. 40-42. Bailey, M. A., and T. Trought. Growth, bud-shedding and flower production in Egyptian » cotton. Min. Agr., Egypt, Tech. and Sci. Serv. Bull. 65. 1927. Balls, Wiglléawrence. The cotton plant in Egypt. Macmillan & Co. Ltd. London, 202 DD- - Beckett, R. E., and J. W. Hubbard. The shedding of 4-lock and 5-lock bolls in upland ,5 cotton. U. S. Dept. Agr. Tech. Bull. 277. 1932. _Berkley, E. E. Studies of the effects of different lengths of day, with variations in " temperature. on vegetative growth and reproduction in cotton. Ann. Mo. Bot. ’ Gard. 18: 573-604. 1931. " Berkley, Dorothy Megowen, and Earle E. Berkley. Super optimal and thermal death temperatures of the cotton plant as affected by variations in relative humidity. Ann. _ Mo. Bot. Gard. 20': 583-604. 1933. .. Canney, Ernest Everett. Rain-grown cotton and climate. Shirley Inst. Mem. 3 (24): " 1924. Carver. W. A. Cotton varieties for Florida. Fla. Agr. Exp. Sta. Bull. 285. 1935. - Cook. O. F. The abortion of fruiting branches in cotton. U. S. Dept. Agr., Bur. Pl. Ind. ’ Circ. 118: 11-16. 1913. ; Dennison. Raymond A. Growth and nutrient responses of Little Turkish tobacco to long a and short photoperiods. Plant Physiol. 20: 183-199. 1945. _ Dunlap, A. A. The_ total nitrogen and carbohydrates. and the relative rates of respira- tion, in virus-infected plants. Amer. Jour. Bot. 17: 348-357. 1930. i Dunlap, A._A. Cotton fruiting studies. Tex. Agr. Exp. Sta. Ann. Rpts. p. 91, 1939; p. -- 87, 1940: and p. 68, 1941. ,1 Dunlap, A. A. Low light intensity and cotton boll-shedding. Science 98: 568-569. 1943. ' Dunlap, A. A. Light, drought, and heat as factors in cotton boll-shedding. Abstr. Phyto- ‘ r-path. 34: 999. 1944. “ Dunnam, E. W., J. C. Clark, and S. L. Calhoun. Effect of the removal of squares on -- .__yield of upland cotton. Jour. Econ. Ent. 36: 896-900. 1943. Q Earle, F. S. Diseases of cotton. Ala. Agr. Exp. Sta. Bull. 107. 1899. ~ p J Eaton. Frank M. Early defloration as a method of increasing cotton yields, and the re- f; lation of fruitfulness to fibre and boll characters. Jour.‘ Agr. Res. 42: 447-462. 1931. 104 22. 32. 33. 34. 36. 37. 38. .39. 40. 41. 42. 43. 44. 45. 46. 47., 48. 49. 50. 51. 52. n D s41 5s. 5s. 'Wadleigh, Cecil H. Nutrition studies with the cotton plant. In 50th Ann. Rpt., BULLETIN NO. 677, TEXAS AGRICULTURAL EXPERIMENT STATION Eaton, Frank M., and Neil E. Rigler. Effect of light intensity, nitrogen supply, fruiting on carbohydrate utilization by the cotton plant. Plant Physiol. 20: 380-41 I 1945. ‘ Ewing, E. C. A study of certain environmental factors and varietal differences i' fluencing the fruiting of cotton. Miss. Agr. Exp. Sta. Tech. Bull, 8, 1918, Gardner, F. E., P. C. Marth, and L. P. BatJer. Spraying with plant growth substanc to prevent apple fruit dropping. Science 90: 208-209. 1939. I-Qamner, A. L. Fruiting of cotton in relation to cotton fleahopper and other insec _ which do similar damage to squares. Miss. Agr. Exp. Sta. Bull. 360. 1941. '~ Harland, S. C. Manurial experiments with Sea Island cotton in St. Vincent, with som notes on factors affecting the yield. West Ind. Bull. 16: 169-202. 1917. - c Hastings, Stephen H. Irrigation and related cultural practices with cotton in the Sal River Valley of Arizona. U.S.D.A. Circ. 200. 1932. g Hawkins, R. S. Variations of water and dry matter in the leaves of Pima and Acall. cotton. Univ. Ariz. Tech. Bull. 17: 1927. Hawkins, R. S., S. P. Clark, Geo. H. Serviss, and Chas. A. Hobart. Varietal differenc" in cotton boll shedding as correlated with osmotic pressure of expressed tissue fluids. Jour. Agr. Res. 48: 149-156. 1934. . '_ Hawkins, R. S., R. L. Matlock, and Charles Hobart. Physiological factors affecting fruiting of cotton with special reference to boll shedding. Univ. Ariz. Tech. Bull. 46. 1933. -. Heinicke, A. J ., and N. F. Childers. The influence of water deficiency in photosyn- thesis and transpiration of apple ileaves. Proc. Amer. Soc. Hort. Sci. 33: 155-159._ 1935. . Johnson, Burt, and Cecil H. Wadleigh. Certain ecological factors and the cotton plant.“ Ark. Agr. Exp. Sta. Bull. 376. 1939. 1 Kearney, Thomas H., and Robert H. Peebles. Heritability of different rates of shed- ding in cotton. Jour. Agr. Res. 33: 651-661. 1926. l’ Kearney, T. H., and W. A. Peterson. Egyptian cotton in the southwestern United‘ States. U.S.D.A., Bur. Pl. Ind. Bull. 128. 1908. King, C. J., and H. F. Loomis. Agricultural investigations at the United States Field-a Station, Sacaton, Ariz. 1925-1930. U.S_.D.A. Circ. 206. 1932. 5_ Knight, R. L. The effect of shade on American cotton. Emp. Jour. Exp. Agr. 3: 31-40.‘ 1935 ' “ Konstantinov, N. N. Light intensity as a factor in the development of various forms of X cotton. Compt. Rend. (Doklady) De l’Acad. des Sci. de YURSS 47: 607-609. 1940. ‘ Lloyd, Francis E. Environmental changes and their effect upon boll-shedding in cotton.‘ Ann. N. Y. Acad. Sci. 29: 1-131. 1920. Mason, T. G. Growth and abscission in Sealsland cotton. Ann. Bot. 36: 457-484. 1922. l McDowell, C. H. Growing cotton under 1rr1gat1on 1n the Wichita Valley of Texas. Tex.‘ Agr. Exp. Sta. Bull. 494. 1934. McDowell, C. H. Irrigation requirements of cotton and Valley of Texas. Tex. Agr. Exp. Sta. Bull. 543. 1937 Novikov, V. A. The influence of the intensity of illumination on the development of the Compt. Rend. (Doklady) de l’Acad. des Sci. de FURSS 11: 397-400. _ cotton plant. 193 . - ~ Prescott, J . A. The effect of water on the cotton plant. Bull. 14 Sult. Agr. Soc. (Cairo). '- 1924 , Reinhard, H. J. Hibernation of the boll weevil. Tex. Agr. Exp. Sta. Bull. 638. 1943. Richardson, T. C. As reported in “Demand for Cotton”-—An Editorial, Morning News, August 31, 1941. Schneider, G. William, and N. F. Childers. thesis, respiration, and transpiration of apple leaves. 1941 Bur. Pl. Ind. Bull. 279. Shirley, Hardy L. The effects of light intensity upon seed plants. In Biological Effects of Radiation II (Edited by B. M. 1936. Shive, John W. In Letter from Department of Plant Physiology, Rutgers University, New Brunswick, N. J., dated March 22, 1943. Steyaert, R. L. Le Port et la Pathologie du Cotonnier. Influence des facteurs meteoro- logiques. Pub. de_ l’Inst. Nat. pour l’Etude Agron. du Congo Belge. Sci. Ser. No. 1 9. 1936. Templeton, J. Watering and spacing experiments with Egyptian cotton. Egypt, Tech. and Serv. Bull. 112. 1932. Thornton, T. Flower bud and boll shedding of cotton in the Ilorin province, Nigeria. Trans. 3rd Int. Congr. Trop. Agr. (1914) 1: 379-384. 1915. Todd, John A. The world's cotton crops. A. & C. Black, Ltd. London, 460 pp. 1915.A k r . Agr. Exp. Sta. Bull. 368: 46-49. 1938. Wadleigh, Cecil H. Growth status of the cotton plant as influenced by the supply of w nitrogen. Ark. Agr. Exp. Sta. Bull. 446. 1944. Young, P. A. Cottons resistant to wilt and root knot and the effect of potash fertilizer in East Texas. Tex. Agr. Exp. Sta. Bull. 627. 1943. grain sorghum in the Wichita‘; The Dallas .-. Influence of soil moisture on photosyn- i Plant Physiol. 16: 565-583. . Shantz, ‘H. L. The effects of artificial shading on plant growth in Louisiana. U.S.D.A., 1913. i Duggar). McGraw-Hill, New York and London. ‘l Min. Agr., i l