~ NOVEMBER ‘I963 CUTTUN PRODUCTIO IN FAR WEST TEXAS with emphasis 0n IRRIGATHJN AND FERTILIZATION TEXAS A8 _ ,1, - " ‘ \ This sentence should read: "For 2 to 2-é-bale yields, no less than 100 pounds and no more than 150 pounds of nitrogen per acre each year is required (3933 following legumes), " instead of "(more following legumes) ." The Trans-Pecos region is the only area exas where cotton is grown under full irri- n, Annual rainfall, ranging from approxi- ly 12 inches in eastern Pecos County to 8 Q: in El Paso County, is too slight and un- 'n to be of any real value for cotton produc- l_ Farmers generally consider the effects fnfall detrimental rather than beneficial to n. f-Irrigated agriculture, therefore, is limited ose areas where either surface or ground p = of acceptable quality are available for =tion. The major area using surface waters e Rio Grande Valley above and below El 1, with some irrigation at Presidio. A small _f1ge of alluvial soils in the Pecos River Valley is irrigated with river water. Perennial g1" at Balmorhea provide water for about f acres in that area. By far the larger por- of irrigated acreage is by pump irrigation. ,4 15 more or less distinct irrigated areas in ,mp irrigation, while the other eight areas on supplemental pumping. Total acreage or previously in irrigation comprises about _ million acres. The map of the Trans-Pecos ‘resource area gives the approximate loca- if the 15 irrigated areas, Figure 1. These ‘ and the approximate acreage in each, with .f irrigation (surface or pump) and a brief ption of soil types are listed in Table 1. jtton is by far the chief cash crop grown 1 West Texas. Approximately 73 percent total irrigated acreage is in cotton. This tage was considerably larger before cot- reage restrictions were introduced. Most cotton grown is of higher fiber quality- i ple Acalas and extra long-staple Ameri- tian types. The hot summers, low fty, low rainfall and bright sunlight pro- " almost ideal climate for quality fiber ion. Average yields are high and are .8 primarily by cool spring and fall tempera- Jand an average frost-free period of only i200 days. though per-acrielyields are high in com- , with other Texas areas, the potentials 3 n production have not even been ap- ‘_ ‘vely, associate agronomist, Substation No. 17, ‘ , Texas; associate agronomist, Substation No. 9, e-Texas; and superintendent and research coordina- bstation No. 17, El Paso, Texas. ans-Pecos, seven are entirely dependent‘ COTTON PRODUCTION IN FAR WEST TEXAS with emphasis on IRRI GA TI ON AND FER TILIZA TI ON D. E. Longenecker, E. L. Thaxton, Jr. and P. I. Lyerly* proached. Average yields now range between 1% and 2 bales per acre for the Acala types. These cottons, with proper management, are capable of yielding from 21/; to 3 bales per acre. Even higher yields than this are possible with skip-row production. Failure to obtain these potential yields can be attributed to many factors, most of which fall into the classification of farm management. Chief among these are inefficient irrigation practices and improper fertilization. This publication presents the results of 4 years’ field tests with Upland cotton by the Texas Agricultural Experiment Station on maJor Trans-Pecos soils. The effect of irrigation fre- quency and fertilization on many aspects of cot- ton production will be discussed, including yields, earliness, lint percent, boll and fiber character- istics, water-use efficiency and disease incidence. Design and Location of Field Tests The 11 tests over the_ 4-year period, 1959-62, were of the same basic field design: Each test was of randomized block design with six com- plete replications and 144 total plots, covering approximately 3 acres. Plots were four rows wide and 40 feet long. Yield data and boll Contents Summary -------------------------------------------------------------------------------------- " 2 Introduction --------------------------------------------------------------------------------- r- 3 Design qnd Location of Field Tests ---------------------------------------------- ~~ 3 Effect of Irrigation and Fertilization 0n ¢°ll°I1 Yield‘ ------------------ —- 5 Substation I7 --------------------------------------------------------------------- -- 5 Fabens ----------------------------------------------------------------- ------------- -- 6 Pecos ---------------------------------------------------------------------------------- -- 7 Fort Stockton ---------------------------------------------------------------------- -- 9 Van Horn --------------------------------------------------------------------------- ~- 9 Skip-Row Yields .................................... -- - ---------------------------------- -- lo Effect of Irrigation and Fertilization on Water Use Efficiency...i..... ‘I2 Effeq- qf Irrigation and Fertilization on Earliness ...................... .. l2 Effect of Irrigation and Fertilization on Boll Characteristics ...... .. I3 Substation I7 ............................ ------------- ------------------------ -- l4 Fabens ................................................................................. -- l5 Pecos ................................................................................... -- l6 Fort Stockton ...................................................................... __ ‘l8 Van Horn ........................................................................... .. l8 Effect of Irrigation and Fertilization on Fiber Properties .......... .. I9 Fiber Length ........................................................................ _. 2O Fiber Fineness ..................................................................... _. 21 Fiber Strength ..................................................................... __ 22 Effect of Irrigation and Fertilization on Incidence of Verticillium Wilt .............................................. __ 23 Acknowledgments ....................................................................... __ 24 Need for Other Nutritional Elements _______________________________________ __ 24 3 NEXV MEXICO p HlQH PLA|N3 F? I 4.WILD HORSE »3ELPAS0 { ' I r ' I 8 i l ' ' " ""1 , HUDSPETH ‘CULBERSON ipggygs VIA ‘ v ' , l. ELPASO VALLEY ~, l l0 2.PRESlDIO I ‘ 2 3.1.090 FLATS I l I l 5. DELL CITY QBALMQRHEA a JEFF DAVIS “:’|\5§0& (‘o a '\ ‘r. P5008 PUMP ~__ / ¢ ‘v a. eAi-"isrow “~~~_ 4 \ \Q\ Pl-ATEAU 9. PYOTE IO. GRAN DFALLS ll. IMPERIAL l2. BAKERSFIELD l3. COYAN OSA I4. FT. STOCKTON l5. BELDING l l PRESIDIO i I I genswsrsa I l I l Figure 1. Map of the Trans-Pecos area showing location and approximate extent of irrigated portions. TABLE 1. IRRIGATED AREAS AND SOILS OF THE TRANS-PECOS LAND RESOURCE AREA WITH RELATED INFORMATION A r ximate P m or $3.0 u p Nature and texture Area irrigated surface * . . . . of surface soils acreage irrigation . ALLUVIAL SOILS El Paso Valley 90,000 Both Stratifield alluvial sands, loams and clays Presidio 6,000 Both Stratifield alluvial loams and clays Barstow 40,000 Botlt Alluvial loams and clays Grandfalls . 13,000 Both Alluvial loams and clays Imperial 15,000 Both Alluvial loams and clays Bakersfield 9,000 Both Alluvial loams and clays UPLAND SOILS Pew; 160,000 Pump Silt loams and loams Cqyqnosq 80,000 Pump Loams and clay loams Balding 18,000 Pump Loams and clay loams Fort Stockton 7,000 Both Loams Van Horn (Lobo Flats) 10,000 Pump Clay loams and clays Van Horn (Wild Horse) 10,000 Pump Sandy loams and clay loams Dell City 20,000 Pump Silt loams Balmorhea 10,000 Both_ Sandy loams and clays Pyoie 4,000 Pump Sands and loamy sands Total 492,000 4 A ‘les for determination of boll and fiber char- 'stics were taken only from the center two Each test included 10 fertilizer treat- ts, (11 in 1962) with six rates of nitrogen four rates of superphosphate. All of the phate and half of the nitrogen (ammonium Ste) was sidedressed in early June. The p” Jhalf of the nitrogen was sidedressed in 11 Y- gEach replication was split for two fre- cies of water application, designated as sum- irrigation frequencies. This method was for several reasons; (1) irrigation prac- from area to area varied considerably, {most growers were irrigating at definite in- foals because it best fitted farm operations T§(3) many of the tests were located on lands ooperating growers, with irrigations per- ed by the grower. Five of the tests were l, e El Paso Valley on alluvial soils, four at '8 and one each at Fort Stockton and Van on upland soils. All tests except one were q tested by hand three times to obtain earliness ffect of Irrigation and Fertilization on Cotton Yields '1‘ Station 17 -j‘_.Alll1Vial soils of the Rio Grande Valley are ly loams or clay loams with a montmorillo- lclay fraction. These soils are highly fertile giare generally underlain at various depths by sand which provides good underdrainage. f ' major shortcoming is the tendency of the f, soils to deteriorate structurally under inuous row cropping. This causes crusting jfclodding and makes tillage operations diffi- When in good physical condition, these A will produce high yields of most farm crops ’ng reasonable tolerance to salinity. Prac- s, y all valley irrigation is level. runs are re- ely short, and cotton is planted on beds after i 'gation. , SlThTGG tests (1959. 1960 and 19611 were con- at Substation No. 17 near El Paso. Two ighe tests (1959 and 1960) were conducted e same plot of Gila loam soil. The third was Qila clay loam. Both fields had been deep- ed to about 30 inches a year or two before were initiated and this is believed to have considerable effect upon the results. Sum- {irrigation frequencies on all tests were every f d every 16 days, beginning in early June ending early in September. Acala 1517C the cotton variety .1 grown. ;In all three tests the effect of differential tion and fertilization on yields, boll char- “stics and fiber quality was similar. For ii reason, the data from these tests have been gined, and average yields for the 3-year 1 are shown in Figure 3. a Figure 2. Tractor-mounted fertilizer apparatus used to apply various rates of nitrogen and phosphate to experimental test plots. Yield curves for 10-day and 16-day summer irrigation show that the response of cotton to applied fertilizer was not greatly affected by irrigation frequency. Maximum yields (approx- imately 21%; bales per acre) with 16-day irrigation were obtained with the 120 N -60 P205 rate. High- est yields with 10-day irrigation were obtained at the 240-60 rate; however, the 50 pounds of additional seed cotton obtained per acre with 10-day irrigation would not pay the cost of the extra 120 pounds of nitrogen. In these tests, therefore, it would not have been profitable to apply more than 100 to 120 pounds of nitrogen per acre, regardless of irrigation frequency. Greatest response to nitrogen was obtained with the lowest applied rate — 60 pounds of nitrogen per acre. At the 800 pound nitrogen level, yields were somewhat reduced at both irrigation fre- quencies. The effect of superphosphate on yields can be observed at the 60 and 300 pound rates of nitrogen. At the 60-pound nitrogen level, the addition of 60 pounds of P205 per acre was of doubtful value. Yields were increased slightly with 16-day irrigation but were decreased with 10-day irrigation. At the 300-pound nitrogen 4500 '- 4000 '- U (I O O s ? IO-DAY SUMMER IRRIGATI@ FREQUENCY —- —— —--— IG- DAY SUMMER IRRIGATION FREQUENCY saco cotton YIELDS. Les/Acne N ‘s’ I I o1 A A A A A A A l A A N-r o so so n20 nao 24o aoo soo soo 30o q so so o so n 2o | so FIDO-IO 6O 6O POUNDS OF N AND P205 APPLIED PER ACRE Figure 3. Three-year average yields of 1517C seed cotton as affected by irrigation frequency and soil fertility level at Sub- station 17 (1959-61). 5 Figure 4. Deep-plowing operation on alluvial soils in the El Paso Valley. level the effect of phosphate was more apparent. Yields were rather sharply reduced by lack of applied phosphate at both irrigation frequencies at this high nitrogen level. This probably re_ fleets an unbalanced nitrogen-phosphate condi- tion which was corrected by 60 pounds of P205 per acre. Results of these and previous tests at this location indicate that this phosphate effect was caused more by very high nitrogen than by low phosphate, and is not likely to be significant at recommended 100 to 120-pound rates of nitro- gen. Soil tests have shown sufficient available phosphate for normal cotton growth on all but very sandy El Paso Valley soils. Failure of the cotton to show a yield response to more frequent (10-day) irrigation on these soils indicates that sufficient water for crop needs was supplied with the 16-day frequency. With 16-day irrigation, cotton grew shoulder-high and showed no signs of suffering for water at any time during the three seasons. Stalks were taller and more rank with 10-day irrigation but pro- duced no more bolls per plant. Fruiting was de- layed and verticillium wilt infection was more severe; both of these factors will be discussed later. 4000 r- 3500*- 3000 — l0- DAY SUMMER IRRIGATION FREQUENCY 2500 '- v- - -- -— l5- DAY SUMNER IRRIGATION FREQUENCY 3:50 servos YIELDS. LBS/ACRE t scoot Q ______4___V_)TJ___:____JI___ __;__M_,__ i“ _ r 1 A | l | l l N-—~) f) m $0 5Q I30 I80 240 300 300 3% 300 Flor“) Q 0 6O 60 O 60 I20 '80 6O 6 POUNDS OF N AND P206 APPLIED PER ACRE Figure 5. Two-year average yields of 1517C seed cotton as affected by irrigation frequency and soil fertility level at Fabens (1959, 1.961). 6 cwith deeper root systems and better water p Q V In these tests, the cotton was planted double-row (cantaloupe) beds in late April. Se lings emerged quickly, seedling disease dama was negligible, and the cotton grew off rapidl It is believed that the previous deep-plowing pe mitted development of extensive root syste y» which were able to obtain moisture from consi erable soil depths. Ten-day irrigation at t location resulted in three; unnecessary summ, irrigations and needless application of about p inches of water, which contributed appreciably production costs. I Fabens - 1 £ In both 1959 and 1961, identical field tes were conducted on the Buckland farm ne‘ Fabens, 15 miles east of the previously de-scrib tests on similar alluvial loam (1959) and cl loam (1961) soils. Climatic conditions were ali at the two locations. Irrigation design and i I gation frequency were also similar, with le 600 to 800-foot runs and summer irrigation f quencies of 10 and 16 days. The only appare difference was in the physical condition of t soil. The fields at Fabens had never been dee plowed and were in generally poorer structu p condition than those at Substation 17. The t ton, also, was planted about two weeks earlier I Fabens in colder soils. Young plants suffer severely from seedling disease and although fi stands were good, almost all roots had be damaged to some extent. a Figure 5 gives the 2-year average seed y ton yields for the Fabens tests. Both tests gay similar response to irrigation frequency, but th differed greatly from those at Substation 1 In both tests, more frequent irrigation resulted V significant increases in yields, averaging alm 300 pounds of seed cotton per acre. Overall yiel were somewhat lower at Fabens than at Subs tion 17 . 7 It appears that the response to additio water at Fabens can be attributed primarily, poorer soil physical conditions and perhaps se ling disease damage. Although no actual m surements of root systems were made, obser tions showed that roots at Fabens were not nea 1 so deep or extensive as at Substation 17. M of the roots were in the top 2 feet of soil. S t low-rooted crops generally respond to more f quent irrigation because (1) the root syste being less extensive, have less soil volume f t. which to extract water, and (2) surface soils i. out more rapidly because of evaporation. ti versely, shallow-rooted cotton is more susce ible to flooding damage and “wet feet.” marked response of cotton to deep plowing, of I noted on alluvial soils, is apparently associa- tration, although actual data is mainly lacking. The effect of fertilizers on yields at Fab showed some similarity to Substation 17. W’ 10-day irrigation, the 180 N-60 P205 rate most profitable). Higher nitrogen rates had 1i p ,0nal effect on yields. The 16-day frequency ntly did not supply enough water for most _nt use of fertilizer. Maximum yields with irrigation schedule Were obtained at the 60 P205 level. Higher nitrogen rates acted a ease yields somewhat. the ‘GO-pound nitrogen level, the effect rphosphate was the reverse of that at tion 17. Yields were increased with 10- igation and decreased with 16-day irriga- There is no logical explanation for this y . At the 300-pound nitrogen level, super- hate additions had no appreciable effect on i at either irrigation frequency. s mmarizing the results of these and other g n the E1 Paso Valley, it appears difficult e general recommendations as to best irri- A frequency for cotton. Crops on similar- ing soils may show marked differences in pose to water even during the same year. physical condition, depth of rooting and date fnting all may affect yields, even though fer conditions and insect control are good. I» be concluded, however, that where: soil conditions are conducive to deep root- ind other factors are not limiting growth, i tton yields can be produced with only four “Le summer irrigations — much less than 1s lly applied. actors often contributing to higher yields se soils are maintenance of deeper furrows ion for more adequate and uniform irriga- if nd later planting (late April) to avoid seed- isease damage in cold soils. Most growers naware that seedling disease can destroy ts and cause cotton to be relatively shallow Without any visible evidence or appreci- oss of stands. A deep-plowing operation will benefit yields on most medium and Xtured valley soils by breaking up lmper- layers and by turning under structurally orated surface soils. Subsoiling also helps. d en alfalfa or other legumes are grown ‘nally and phosphate applied to the le- L .:~ a no additional phosphate should be needed i on at the present time. With continuous a moderate phosphate application (50-60 . of P205 per acre) every 5 to 6 years may irable to avoid possible unbalanced nitro- gosphate conditions. For 2 to Zlé-bale ‘ no less than 100 pounds and no more than funds of nitrogen per acre each year is re- (more following legumes). This nitrogen g preferably be sidedressed in late May or June. Sandy soils may require split appli- to avoid excessive leaching losses. d.“ - v,_*<-l‘ e largest body of upland soils under irri- g in the Trans-Pecos area is the Reeves- n seriesl, composed of grey to brown, highly p-i loams and silty clay loams, underlain t gypsiferous caliche deposits at depths ely re-named “Hob-an.” Figure 6. Typical cotton roots from Fabens test showing shallowness of rooting and absence of taproots. Greatly restricted taproot (center plant) penetrated crack through compacted soil. of 2 to 4 feet. However, these soils in the Pecos area contain a high percentage (60 to 65 percent) of silt and tend to disperse easily when wet form- ing hard thick crusts and clods. The Pecos area is entirely pump irrigated. Saline ground waters average about 2,700 ppm total salt. Cotton yields have generally been good in spite of this because soils are permeable and salts are easily removed by leaching. Cotton is usually lister-planted and worked onto beds as the season progresses. Surface irrigation is either sloping or level, with generally longer runs than in the El Paso Valley. Much of the Pecos pump area has been developed from virgin soils during the past 20 years. Because waters were saline and soils were permeable, many farmers have developed the practice of frequent irriga- tion combined with heavy nitrogen applications in cotton production. Four consecutive years of cotton tests (1959- 62) were conducted on these soils at Substation 9, Soils and Crops Unit, Pecos, Texas. Field design Figure 7. Cotton seedlings showing damage from seedling disease. Infected plants (left) may survive but root systems will be shallow compared to healthy plants (right). . 1 4500'- ‘ooo- Zgj -_-Q> i 3500- / 3°00 - ‘l-DAY SUMNER IRRIOATION FREQUENCY N u O O -— —- — —— lQ-DAY SUQER IRRIGATION FREQUENCY seen cotton YIELDS. Les/Acne IO 3 O >- l L h l | 1 N-—D O 6O SO PIOI—> O O 4r n 1 n 1 l 240 300 300 300 300 O 60 I20 I80 6O 6O POUNDS OF N AND P200 APPLED PER ACRE Figure 8. Three-year average yields of 1517C seed cotton as affected by irrigation frequency and soil fertility level at Pecos (level irrigation, 1959-61). was identical to El Paso Valley tests. The first 3 years (1959-61) tests were established 0n level land with short (400-foot) runs. The 1962 test Was on sloping land (0.4 percent slope) with 800- foot runs. Results for the first 3-year period are given in Figure 8. Summer irrigation frequencies were every 7 and every 14 days in all Pecos tests. Figure 8 clearly shows that cotton yields were greatly reduced by frequent (7-day) irri- gation on level runs. These yield reductions oc- curred at all fertility levels and were without question caused primarily by adverse soil physical conditions resulting from frequent irrigation. With 7-day irrigation the soils stayed wet rather than moist. Cotton grew shoulder to head high, had a definitely lighter green leaf color, and a generally less thrifty appearance. Cotton get- ting 14-day irrigation was chest high, had a better leaf color and was more bushy and healthy looking. More crusting and clodding occured with 14-day irrigation because soils dried out more between irrigations. Because of the formation of large cracks, soil aeration was much better, however, and this factor appears largely respon- sible for the significant differences in growth and yields. There seems to be no doubt that the combin- ation of level soils and frequent irrigation had strong adverse effects on many aspects of cotton growth and physiology. Data on boll character- istics, to be presented later, bear out these con- clusions. . The response to nitrogen applications was significant with both irrigation frequencies, bu.t was greater with 14-day irrigation. Maximum yields with 2-week irrigation were obtained at the 240 N-60 P205 level, an increase of 620 pounds per acre over the no-fertilizer plots. With weekly irrigation, maximum yields were reached at the 120 N -60 P205 level, a total increase of 500 pounds per acre. Apparently the cotton was able to utilize applied fertilizer more efficiently with less frequent irrigation. There was no significant yield response to phosphate except a reduction at the 300 N-O P205 level with both irrigation fre- 8 I system which brings about the adverse effec quencies, indicating an unbalanced N-P2O5 condi tion. The net result of too frequent irrigatio,’ on level runs was an average loss of over 60 pounds of seed cotton per acre in addition to th cost of applying 6 extra summer irrigations, l combined loss of approximately 100 dollars acre. x In 1962 the test was conducted on 0.4 percen slope in an adjoining field; '”()ne other differen i was the inclusion of a 180 N-O P205 fertilizi treatment to better evaluate the effect of supe phosphate. Results are shown in figure 9, an are quite different from those of the previo . 8 years. In this test, highest average yields"‘we I obtained ,with 7-day irrigation. These diffe ences were» apparent at 9 of the 11 fertility trea a ments and were statistically significant at the percent level. a With 0.4 percent slope, very little crusti ' or clodding occurred. Water was confined a the bottoms of the furrows, and beds were we ted by lateral capillary movement, commonl called “subbing.” With this management sy tem the soil was irrigated without appreciab flooding, soils were kept moist without crusti or clodding, cultivation was easier and the remained loose and friable. Cotton grew tall and somewhat more dense with 7-day irrigati but there» were no visible effects of over-ir gation such as occurred on level soils. It can ryf? concluded from these tests that any irrigati‘ of flooding, crusting, clodding or waterloggi of these soils will contribute to lower yields a other growth effects. A It appears that sloping irrigation should l most suitable in this area provided that adequa and fairly uniform water penetration can be q tained without excessive tail water losses. Pr visions for return flow may have to be incorpo ated into the systems. Most desirable irrig tion frequency and degree of slope have not é been determined. Work at the Trans-Pecos su station is in progress to determine this. I Response to nitrogen on sloping runs w good with both irrigation frequencies, with hig est yield increases obtained under 14-day i i gation. Maximum yields occurred with 1 4000- 0| (I 8 \ \ 1 1 1 i x ‘I-DAY SUMNER IRRIGATION FREQUENCY — — -— — M-DAY SUMNER IRRIGATION FREQUENCY SEED COTTON YIELDS. LBS/ACRE 3 O O "s: is arms POLHDS OF N AND PtOU APPLIED PER ACRE V Figure 9. Average yields of 'l517D seed cotton as affected irrigation frequency and soil fertility level at Pecos (0.4 per slope, 1962). nds 0f nitrogen per acre regardless of irriga- n frequency. Higher nitrogen rates resulted significant yield reductions with 14-day irri- tion and may indicate that the crop did not eeive sufficient water for maximum response to yilizer. The only apparent effect of super- osphate on. yields Was the previously noted ld reduction with 14-day irrigation at the 300 I P205 level. This did not occur with 7-day ‘gation, for reasons not clear at the present e. These tests indicate that nitrogen rates for la 1517 in this area should never exceed 180 a ,= nds of nitrogen per acre. The good permea- 'ty characteristics of these soils, while excel- l t for salinity control, undoubtedly are con- ive to high nitrogen leaching losses. For this on NH3 or ammonium-type forms should be A re desirable owing to their greater resistance ‘leaching. Split nitrogen applications would go appear to be advisable. There does not seem be any immediate need of yearly phosphate lications for cotton. Phosphate recommenda- i s as made for the El Paso Valley would prob- ; y be applicable here also. if Stockton Only one test was conducted in the Fort a 1kton area, and this was on the Chandler A s at Belding in 1960. Although fertilizer Water recommendations for the area cannot permade on the results of one test, the findings informative for other reasons. In this test, 10-day summer irrigation was ied out on all plots from early June until mid- y. At this time half of the plots continued the same schedule while the other half was a back to irrigation every 20 days. Last irri- ion for all plots was on Sept. 13th. The ~se of this irrigation arrangement was to F “irmine the effect of reductions in late sum- A irrigation frequency on yields, earliness, boll i. acteristics and fiber quality. The soil in the test area was a dark brown, y» well-aggregated loam, classified as Reeves F; deep phase, but differing greatly from the pceja; at Pecos. It was representative of a large y of soils in the Belding-Coyanosa area which. among the most productive soils in the en- Trans-Pecos region. The test was on level t ;k 1517C cotton was planted on single-row in late April. From emergence to maturity, r in all plots was strong and vigorous, un- tedly the finest-looking crop of the entire series. Yield data are reported in Figure 10. These results clearly show that drastic cut- in late summer’; irrigation frequency can a serious effect on yields in this area. Eli- tion of three late summer irrigations reduced seed cotton yields an average of 350 pounds cre, a highly significant reduction. As will be shown from earliness, boll char- t, istic and fiber quality data, these yield re- 4500f ‘O00’ é l \ \ l. \ \ / i l \ l 1 I I l | l l \ l l | 1 | l SO00 >- IO-DAY SUMMER IRIGATION FREQUENCY -— —— —- -— lO-DAY IRRIO. TO IID-MLYiEO-DAY IRRIO. TO HID-SEPT. 2500- szso cowou YIELDS. Les/Acne | n {L} n u -> o so so mo foo 24o 3'00 aoo =3» also PzOa—>O o so oo co o .00 uzo I00 J0 POUNDS 0F N MO P108 IPFLED PEI ACRE Figure ‘I0. Average yields of 1517C seed cotton as affected by irrigation frequency and soil fertility level at Fort Stockton (I960). ductions apparently were the only adverse effect attributable to less water in late summer. All other aspects were favorable. Yields with 20- day late summer irrigation still averaged 21/2 bales per acre. Less favorable late-season wea- ther in 1960 probably would have lessened the difference between irrigation frequencies. This aspect is discussed more fully under “earliness.” This was the only one of 11 tests in which nitrogen applications gave no significant yield increases. Cotton growth, yields and soil-test information all indicated that available soil ni- trogen was adequate for crop needs without fur- ther additions. Cotton was shoulder to head high in all plots with a deep green leaf color. Corn had been grown the previous season. There is some indication that additional phos- phate was used to good advantage by the crop, particularly at the 60-pound nitrogen level. At this nitrogen rate, 60 pounds of P205 per acre gave significant yield increases with both irri- gation regimes. These increases were not evi- dent at the 120 and 180-pound nitrogen levels, however. Further tests on these soils would be necessary before specific recommendations could be made. ‘ Van Horn The final test to be reported was established in the Wild Horse section of Van Horn on the Chandler Farms in 1962. This test was on uni- formly deep Reeves sandy loam of only moderate fertility. Acala 1517D cotton was planted April 27 on single-row beds. The field had an approx- imately 0.1 percent slope with 800-foot runs. Summer irrigation intervals were every 12 and every 18 days, to conform with the grower’s schedule. The 18-day plots got five summer irri- gations; the 12-day plots received eight. Stands were good, the cotton grew off rapidly, and wea- ther conditions all season were favorable. Yield data are given in Figure 11. This test emphasizes a situation that dif- fered appreciably from any of those previously described. Soil physical conditions here appar- 9 4000 y- /° \ ll (I ° s s o I T . O \ nz-mv suuuza lame. rneoucucf -— —- —- —- ll-DAY SUMNER RRlO- FREQUENCY sun cotton new; LBS/ACRE IO I0 El O O I l l l l N—) 0 6O 6O I20 I00 I00 PtOl-)0 0 6O 6O O 60 POUNDS 0F N AND P20! APPLIED PER ACRE 300 300 300 300 0 6O I20 I00 Figure 11. Average yields of 1517D seed cotton as affected by irrigation frequency ‘and soil fertility level at Van Horn (0.1 percent slope, 1962). " ' ently were not a limiting factor. The soil was friable, infiltration rates were very good, and the cotton in all plots was vigorous and deep rooted. Even though three fewer irrigations were used, the 18-day irrigation schedule produced signi- ficantly higher yields. With the 18-day fre- quency, plants were about shoulder high, com- pared with above-head height in the 12-day plots. However, the largestalks of the 12-day plots did not set a heavy boll load and maturity was definitely delayed. Reductions in yield oc- curred at all fertility levels, and fertility differ- ences appear to be partly responsible for differ- ences in yield. Judging from the almost 3-bale yields on the 18-day plots, soil moisture did not appear to be a limiting factor in this test. The 12-day sche- dule supplied excess water which. caused rank vegetative growth and apparently leached away nitrogen, which prevented the setting of a heavier boll load. Light also was probably a factor re- ducing boll set in this tall cotton. The conclu- sions concerning nitrogen were based on several facts: (1) Yield reductions on the 12-day plots were greatest near the head ditch where infiltra- tion was greatest, and (2) at least 5 or 6 inches of water were applied per irrigation to these permeable soils. Also, considerably greater re- sponse to nitrogen was evident with less frequent irrigation. On these permeable soils, too fre- quent or too heavy irrigation can add to produc- tion costs ilaknd also may reduce yields by leach- ing away applied nitrogen. A iHighest yields, 3 bales per acre,were ob- tained with 18-day irrigation at the 120 N-60 P205 level, proof that heavy nitrogen applications are not necessary for good yields, even on per- meable soils. Phosphate significantly increased yields at the 60 and 180-pound nitrogen levels with 18-day irrigation but not with more frequent irrigation. This emphasizes the often complex interrelationships among water, nitrogen and phosphate in irrigated cotton production. More tests in this area will be necessary before defin- ite irrigation or fertilizer recommendations can be» made. l0 l fornia has been grown in skip-row fashion. T i’ a It may be concluded from results of thes tests that it would be exceedingly difficult rf1 recommend specific summer irrigation frequen ies for best cotton production, even within an particular area. Many factors operate which in fluence to some extent the effect of irrigatio frequency on yields in any given year. Mos of them have been discussed to some exten These include soil texture, irrigation system de sign, soil structure, date of planting, depth o rooting and climatic variability, as well as other For these reasons, it appears inevitable that th determination of best irrigation frequency o any particular farm, with any particular; seq of management practices, must rest with the grower himself. No farmer can afford not to l. t doing some experimenting on his own. A changl in slope, in length of run, in cotton variety, i type of bed, even in depth of furrow, may alte the response of cotton to water. it Skip-row Yields It has been known for many years that our side cotton rows often produce higher yiel than solid-planted rows. The explanation pears to result partially from the effect of ext i light, but is undoubtedly caused partly by root competition for water and fertilizer on oug side rows. Whatever the cause, the effect been recognized in irrigated cotton areas, a f much cotton from the Mississippi Delta to Cal involves leaving one or more unplanted rows regular intervals throughout the cotton field. Recent (1962) easing of USDA regulatio on skip-row planting now makes it possible r as little as one unplanted row in a regular sk pattern to be deducted from alloted acreage a presents the possibility of greatly increased yiel per planted acre. The question involved A whether or not the increased yields from outsi rows is great enough to offset the extra cos of preparing and maintaining skip areas. Mal growers have been disappointed with skip-r methods. Others continue to plant all or p 2500 - OUTSIDE ROWS 2000 - Ill I i’ \. 3 I 50° _ INSIDE Rows J . ,0\ /.__ __ __ __g _ __ I \ > ___ __- 0/ ,_ _. '5' |ooo- J ‘I-DAY suuuzn IRRIGATION FREQUENCY — ——--—l4-DAY sumaen mnnsnuou rncousucv soo - M *4 L u-s o o I00 20o 2 30o 30o P20o—-)O so o o o so POUNDS 0F N AND P205 APPLIED PER ACRE Figure 12. Comparative lint yields from inside and ou A rows of 1517C cotton as affected by irrigation frequency and fertility level at Substation 17 (1958). OUTSIDE ROWS __ --o-\ INSIDE ROWE /"' "" _ ‘F’ _ _ _ _ lO-DAY SUMMER IRRIGATION FREQUENCY —- -— — — IG-OAY SUMMER IRRIGATION FREQUENCY r l a r 1 l n L r '1 ' 6O 5O I20 llO H4O BOO BOO BOO BOO ‘=3’ ‘ O O 6O 6O 6O O 5O I 2O I BO 6O POUNDS OF N AND PlOl APPLIED PER ACRE figure I3. Comparative lint yields from inside and outside of 1517C cotton as affected by irrigation frequency and soil ‘ level at Substation l7 (1961). heir allotments in this manner. The prob- is not only one of management but involves factors as well. During the past 5 years, three evaluations kip-row plantings were made in the El Paso ey. All three gave essentially similar results, only two will be reported here. In 1958 a Lination fertility-irrigation test with 1517C lg on alluvial clay loam at Substation 17 Was ted so that one outside row of each plot bor- y... an adjacent 2-row skip area. These out- 1 rows were harvested separately, and Figure Tompares the yields of lint with inside (or -planted) rows. Summer irrigation fre- Qjcies were every 7 and every 14 days. In the fertility-irrigation test at Substation 17, "ously reported, was designed so that one 'de row of each plot bordered an adjacent g-row skip area. Lint yields from outside ‘ssinside rows are reported in Figure 13. In tests the skip areas were» maintained in a-free borders by occasional disking and re- ring to obtain best control of irrigation y; In each of these tests, outside rows yielded an average of almost 600 pounds more lint per planted acre than inside rows, calculating all rows in the field as border rows. In other words, a system of plant two, skip two or plant two, skip four would have yielded well over a bale more cotton per planted acre. These increases would have been obtained even though average solid- planted yields in both fields were well above 2 bales per acre. In both tests the cotton was tall and vigorous, and boll set on outside rows was great enough to cause the plants to bend over into the skip areas in late summer. This oc- curred because most lateral branching and boll set on these rows was on the skip» side, a strong indication that light plays an important role in skip-row production. Nitrogen increased outside row yields in 1958 but had little effect in 1961 on soil that had recently been deep-plowed. In the 1961 test, outside rows on the no-fertilizer plots yielded an average of over 4 bales per planted acre. This indicates the tremendous yield potential of these finer textured alluvial soils when conditions are right for maximum production. Researchers in Arizona have calculated that the first 14 bale increase in skip-row planting is absorbed by increased production costs. In these tests, then, assuming a plant-two system, profits would have been increased by 3/4, bale per planted acre. Tall, healthy, deep-rooted cotton is essential for good success with skip-row planting. The heavy boll loads required on outside rows de- mand that the plants be large and vigorous enough to set and carry this load. For this rea- son, best success with skip-row cotton is obtained on soils of high production potential. These tests show that high yields can be obtained with- out special irrigation or fertilizer treatment on ire i4. Skip-row planted cotton in the El Paso Valley. I out additional fertilizer or water. On productive soils, this practice can greatly increase yields per planted 11 60- \ U O I IO-DAY SUMMER IRRIGATION FREGJENCI — -— -— —- IG-DAY SUMMER IRRIGATION FREQUENCY N O I PERCENT SEED COTTON, FIRST HARVEST "o" I 12o nao 24o aoo soo soo 3'00 o so so so o so o so POUNDS or u mo mo; APPLIED PER ACRE N-—) P2OU-—+ O O 5O Figure 15. Three-year average percent of 1517C seed cotton harvested at first picking as affected by irrigation frequency and soil fertility level at Substation 17 (1959-61). productive soils. Skip-row planting, under good management, presents opportunities for greatly increased yields and profits in southwest irrigated areas. It also permits easier access to the cot- ton for insect control and defoliation spray pro- grams with ground rigs. Effect of Irrigation and Fertilization on Water Use Efficiency The term “water use efficiency” has been widely used in irrigated crop production to des- cribe the efficiency of irrigation with respect to crop yields. In hay and. forage production it has been defined as “the pounds of dry matter produced per inch of water applied.” In cotton production, it has come to mean the pounds of lint produced per inch of water applied. It is "particularly important in comparing different cotton types or varieties but has significance in all irrigated areas from the standpoints of water conservation and production costs. The term includes rainfall also, but for all practical pur- poses, rainfall in this area can be disregarded. Actual measurements of applied water were not made in these tests, but reasonable approxi- mations will serve for purposes of this discussion. In all tests, “water use efficiency” was much BOP b O I U O I PERCENT SEED COTTW, FIRST HARVEST '8 I lO-DAY suuuzn IRRIGATION mzoucucv '0' — —- —— IG-DAY suuusa IRRIGATION FREQUENCY n I I I I l I I "-5 O 6O 6O I20 I80 240 300 300 300 3Q’) PzOl-fi 0 O 6O 6O 6O 6O O 6O I20 I00 POUNDS OF N AND Pam APPLIED PER ACRE Figure 16. Two-year average percent of 1517C seed cotton harvested at first picking as affected by irrigation frequency and soil fertility level at Fabens (1959, 1961). 12 fertilization on earliness is of considerable I lower with more frequent irrigation. Pounds of lint per inch of applied water ranged from .9 high of 4O at Substation 17 in 1961 with 16-da . I irrigation to a low of 15 at Pecos in 1961 with 7 p day irrigation. Even in those tests in whic yields were increased with more frequent irri gation, the water use efficiency still was signi? ficantly higher with less ‘frequent irrigation Nitrogen, by increasing yields, improved wate use efficiency. I With limits, of course, greater efficienc: is always obtained with less water in irrigate cotton production, even with indeterminate t. l: Acala cottons. No information is currently avail able in this area as to how little water is actuall needed to produce a profitable cotton crop». On" bale of 1517D per acre was produced at the Pec Station in 1962 with only one summer irrigatio A Further investigations are needed along this lin Effect of Irrigation and Fertilization on Earliness Cotton growers are well aware that earl harvested fiber usually returns the highest pric a because of better grades and generally high fiber. quality. This is especially true in the Tran Pecos area with its limited growing season, whe l high quality cottons of indeterminate fruitin habits are grown almost exclusively. i For this reason the effect of irrigation and, portance. In these tests, earliness was dete mined by the percentage of the total yield at fir harvest. Generally three pickings were mad beginning when 2O to 40 percent of the b0 were open. Figures 15-19 give the percentag of seed cotton at first harvest for the tests a Substation 17, Fabens, Pecos and Fort Stockt, respectively. These data were not obtained . Van Horn because of a shortage of hand la for harvesting. a These results are conclusive evidence t I more frequent irrigation has a definite delayip effect on crop maturity. In all 10 tests, ex summer irrigations delayed maturity by si’ ficant amounts regardless of soil fertility lev 50- §_ §O—- -._____*_-—"" A O I 0| O I T-DAY SLHMER IRRIGATIM FREOIENCY — —— — -- M-DAY SMMER RRIGATION FREGIEY PERCBIT 8E0 COTTON, FIRST HARVEST a 8 I I O I - |-q 60 I20 I60 240 500 300 300 MOI-fig ‘O0 6O G0 6O 6O O 8O I20 POUNDS OF HAND PzouAl-‘PLIED PER ACRE Figure 17. Three-year average percent of 1517C seed c_ harvested at first picking as affected by irrigation frequency » soil fertility level at Pecos (level irrigation, 1959-1961). ~ percentages at first harvest were lower as f"; as the actual amounts harvested. This effect particularly noticeable in the 3-year average p; at Substation 17 where yields were not in- f; ed by extra water and at Pecos (Figure 17) ire extra water greatly decreased yields. Even abens and Fort Stockton, where more fre- To irrigation increased yields significantly, the ntages at first harvest were lower. i ' It can be concluded from these data that ‘frequent summer irrigation would definitely ten maturity in this area. The Fort Stock- test (Figure 19) strongly indicates that late- tmer irrigation frequency holds the key to fness in the Acala varieties. In this test, sion of three late summer irrigations reduced 5i significantly, but final yields still were 1 bales per acre. With 20-day late-summer ation, harvest was completed by November ' With continued 10-day irrigation, final har- j was completed in early December. With erse late fall weather (early freeze) many ese bolls would have failed to open. However, 9- yield reductions show that some caution ' ld be exercised in reducing late-summer irri- yjon frequency. These data are not sufficient 'ke specific recommendations. It appears A on most medium to fine-textured soils some ctions in late-summer irrigation frequency -_~possible without appreciable losses in yield. reas with early first frost date (Van Horn Dell City), the benefits of earlier maturity i. d more than offset slight yield reductions. "rally had a. slight additional delaying effect i; aturity when expressed as percentage of yield. However, there were no significant rences in amounts harvested at first pick- data not shown). The overall effect of nitro- »_ appears to be an increase in total yield ght about by p-rolonging of the fruiting ‘n Increases in yield due to nitrogen are fore usually in. the form of heavier late ests, which can not be called a delay in l rity. This is borne out by data from no- _gen plots in which first harvest yields were i but late harvest yields were very light. ' - lack of nitrogen tends to make Acala cot- _ more determinate» in their fruiting habits ‘ ducing late-summer top growth and con- fently reducing the late-boll load. The only apparent effect of superphosphate rliness was at Pecos in 1962, where phos- i- at high nitrogen levels appeared to induce- “er yields at first harvest. There is no evi- in any of these tests that earliness can be- “ ved by phosphate at recommended nitrogen ffect of Irrigation and Fertilization A on Boll Characteristics 3Little information is available concerning the i‘ of fertilizers and water on cotton boll cteristics under southwest irrigated condi- _The data show that nitrogen applications v T a o v l l l t l T-DAV SUMMER IRRIGATION FREDUENCV -— -— —— — Il-DAY SIIIER IRRIGATIXI FREUJENCV PERCENT sun cotton nus-r mrv: a 8 I I I . l l 9 l l \ | 1 L l I so so 12o 30° 30° 5°° o so so o no no _., no no 24o PJoI-ag o so so FGINDS OF N AND PQDI APPLIED PER ACRE Figure 18. Average percent of 1517D seed cotton harvested at first picking as affected by irrigation frequency and soil fertility level at Pecos (0.4 percent slope, 1962). ' tions. This information should be of consider- able importance not only in basic cotton research but also in the practical aspects of scientific cotton production. Boll characteristics of any particular cotton variety are large-ly hereditary in nature and subject only to relatively minor var- iations in response to environmental changes. For that reason, these variations, when they occur, are some-times indicative of serious physiological disturbances. In addition, changes in boll char- acteristics can have an appreciable effect on lint yields per acre and are therefore of direct in- terest to the cotton grower. The four boll characteristics to be discussed here are (1) seed cotton weights per boll, (2) lint weights per boll, (3) seed weights per boll and (4) lint percent. Lint percent is closely related to but not synonymous with “gin turn- out,” which is always somewhat lower because of the removal of burs, trash, etc. in the ginning operation. Lint percent of 1517C cotton can vary anywhere from 39 to 33 percent as a re- sult _of changes in environmental conditions. As- suming 2-bale cotton yields, a variation of only 3 percent could mean a difference of 30 pounds of lint or about 10 dollars per acre. Lint percent or “gin turnout” depends on the relative 10'- ~——- _- so- \ u 0 » u O v lO-DAV SUMMER IRRIGATIOI FREQUENCY — — —— IO-DAV IRRIQ TO IlD-JULY-ZO-DAY IRMQ. 1'0 MID-SEPT. PERCENT SEED COTTON. FIRST HARVEST a 8 l I 41 i» its s50 fine o co uzo no (‘u-so no .2, mo 23o non-so o 0o 0o co rouvos OFIAIIDHOeAFHID Puma: Figure 19. Average percent of 1517C seed cotton harvested at first picking as affected by irrigation frequency and soil fertility level at Fort Stockton (1960). l3 - ,-_’o\ / /'/ 9- _ _ _‘_ _ ~ / .. o - / A LATE HARVEST Z \ / \ / \ / \0—-———I/ / \ .../ /‘\ f.’ .__ - K / / \ \ \ ,/ IO-DAY sumizn IRRIGATION rnzoueucv - —- - lG-DAY SUMMER IRRIGATION FREQUENCY WEIGHT or sczo covron PER BOLL (gins) GI i O i i L l i i i I i {i N——-§ O SO SO I20 ISO 240 300 300 300 300 P10a—¥0 O 6O 6O SO 5O O 60 I20 I80 POUNDS OF N AND P209 APPLIED PER ACRE Figure 20. Three-year average weights of 1517C seed cotton per boll at two harvest dates as affected by irrigation frequency and soil fertility level at Substation 17 (1959-61). weights of lint and seed per boll. For that reason, the effect of fertilizers and water on lint and seed weights is important. Lint yields per acre also are affected by boll size and the number of bolls per plant. A variation in lint percent could be partly or entirely offset by changes in boll weight or number of bolls per plant; the opposite also could be true. A lower lint percent together with smaller bolls or fewer bolls per plant could mean a considerable loss in pounds of lint per acre. In all field tests, 20-boll samples were care- fully selected from each fertilizer and irrigation treatment -- 10 bolls from each of two replica- tions. These were taken at two different dates corresponding to mid-harvest and late harvest periods. The samples were ginned and evaluated for boll and fiber properties at the fiber labora- tory at Substation 17. No actual data were taken on stand counts or number of bolls per plant. Planting rates and row spacings varied slightly from location to location. For that reason, evaluations of boll characteristics will be made at each individual location. Comparisons between locations would not be entirely justified. Substation 17 Figures 20-23 show the effects of irriga- tion frequency and fertility level on boll weight, lint weight, seed weight. and lint percent, re- spectively, at Substation 17. All data are 3-year averages. 5.- MID- HARVEST ,..._._---4____*_.._.__4 _._---»-__ ,_ __,__- ____ I [- Qms) LATE- HARVEST n I IO-DAY SUMMER IRRIGATION FREQUENCY V -— -— — — IC-DAY SIJAIIER IRRIGATION FREQUENCY WEIGHT OF LINT PER BOLL \ o I i i l i i i i l mi I-—)O GO SO 240 300 BOO 300 30° F|OI—)O O 6O 5O O GO I20 IIO POUNDS OF N AND P10! APPLIED PER AUIE Figure 21. Three-year average weights of lint per boll of 1517C cotton at two harvest dates as affected by irrigation fre- quency and soil fertility level at Substation 17 (1959-1961). 14 Q I io-oiv simiicli inmoinou msoimicv (,1 — —-—|o-oiv amen mmairiou maiciicv 1mm’ or sczo PER sou. (gi-is.) u if. O 1 I i i | i i i 4g N—b 0 60 6O I20 IIO 240 300 300 I» K I H0140 0 6O O0 l0 0 IO I20 I” _ 60' POIIIDSWNANIJPIOIAPPLIDPERAK Figure 22. Three-year average weights of seed per U of 1517C cotton at two harvest dates as affected by irrigation f V quency and soil fertility level at Substation 17 (1959-1961). Boll weights (Figure 20) are by nature su ject to considerable variation. Different bol on the same plant, maturing at the same tim may differ considerably in seed cotton weigh This is caused only partly by location of the b0 on the plant. Minor variations in boll weig curves should therefore be disregarded, and onl the more obvious conclusions should be drawn. Greatest differences in boll weights were du to irrigation frequency. cotton weights were significantly heavier wit more frequent ( 10-day) summer irrigation. Th‘ differences averaged 0.4 gram per boll. A simil but less noticeable effect was evident at la harvest. Differences in boll weight at this da ’ averaged 0.18 gram in favor of more freque irrigation. Y Figures 21-22 show that these same diffe ences occurred in Weights of both lint and s‘ per boll, with the greatest effect being on --., weights. As will be seen later, this same eff ~- was true to some extent in all tests where cott was deep-rooted and healthy, and was particular noticeable at Fort Stockton where late-season so moisture was very low in some plots. It can I concluded with reasonable certainty that 0 ‘ 36- \ , /‘\ \ ib-NARVEST \,_ ____--n~\ LINT PERCENT in 0i 1 IO-DAY SUMMER IRRIGATION FREQUENCY — — -— — lG-DAY SIHIER IRRIGATION FREQUENCY l1 i i i i 0 ir-s so i5» 2'40 ‘ alao i» PtOl-DO O GO H) GO U0 I20 POUNDS OF N AND P10! APPLIED PER ACRE Figure 23. Three-year average lint percents of 1517C co at two harvest dates as affected by irrigation frequency and fertility level at Substation 17 (1959-1961). MID-HARVEST __ ___ __¢.._ ... _._o ,¢\ /‘ V , \ ,»~\ r At mid-harvest, se . \ LATE-HARVEST ,__ __ _ IO—DAY SUMMER RRIOATIGI FREOIENOY — -— — —’ IS-DAY SUMMER IRRIGATION FREQIENCY A | r 1 I l L I 6O I20 I00 240 300 300 300 300 60 8O 6O O 6O I20 IQ 60 POUNDS OF N AND P10! APPLIED PER ACRE gure 24. Two-year average weights of 1517C seed cotton _l at two harvest dates as affected by irrigation frequency fertility level at Fabens (1959, 1961). it 0f more frequent irrigation during the ing season is to increase boll Weight, with f r effect 0n seed than on lint. Seed Weight ilwses at Substation 17 averaged 0.27 grams ill at mid-harvest and 0.12 grams at late est. Corresponding lint Weight increases f: 0.10 and 0.05 grams. he effect of more frequent irrigation in in- p} d seed Weights is reflected in the lint per- data (Figure 23). Mid-harvest lint percents ‘ged 36.7 for 16-day irrigation as compared ‘.2 for 10-day irrigation. The-se differences L small but significant. Late harvest differ- i; were much less (0.2 percent). The effect of maturity date on all boll char- ‘tics should be noted. Late-harvested bolls much lighter in Weight at both irrigation Qncies. Both seed and lint Weights Were Lint percent Was also appreciably lower. _ effects Were strongly evident in all tests at g tions and are partly associated With lower ratures during late-season boll development. his reason alone, growers should strive to anagement practices so as to obtain earlier C‘ 1W- ‘ itrogen increased both boll Weights and eights appreciably at mid-harvest, With _ t increases occurring at the highest nitro- evels. This response to nitrogen Was evi- _~ n all tests at all locations, as Will be observed |O~DAY SUHNE, IRRIGAYION FREQLENCY --— -—- —— —|5-DA'AV SUIIIEQ“ IRRIGATION FREQUENCY I | l l | I If) 6O I20 ISO 240 300 300 300 300 O 6O 6O 6O BO 0 6O I20 I80 c; POUNDS OF N AND P100 APPLIED PER ACRE i I re 25. Two-year average weights of lint per boll of 1517C I two harvest dates as affected by irrigation frequency Jertility level at Fabens (1959, 1961). MID-HARVEST / , /’_ "' / __ flEfimfisL ../ WEIGHT OF SEED PER IOLI. h I lO-DAY SUIIIIER IRMOATION FREQUENCY —-—— —— -— IG-OAV SUIIIER IRRIGATION FREQUENCY 0| I l . . u—> o eh so |'2o roe 21o =5» =5» zoo aoo non-w o so so o so 12o no SO l 6O POUNDS OFNAIDPQOIAPPIJEDPERAORE Figure 26. Two-year average weights of seed per boll of 1517C cotton at two harvest dates as affected by irrigation fre- quency and soil fertility level at Fabens (1959, 1961). later. In contrast, lint Weights per boll were rel- atively unaffected by nitrogen. This too Was essentially true in all tests. Lint Weight appears to be the most stable of all boll characteristics measured in these tests at any given harvest date. It is affected only slightly by irrigation frequency and scarcely at all by fertilizers. It is concluded that the principal effect of nitrogen on boll Weight can be attributed to its action in increasing the Weight of seed. The effect of phosphate on boll characteris- tics at Substation 17 Was variable and difficult to evaluate. Some slight increases in seed Weight at high nitrogen levels may have been due to phosphate applications. There Was no consistent effect of phosphate on either lint Weight or lint percent. Fabens Figures 24-27 show the 2-year average boll characteristic data for the Fabens tests. Some distinct differences Were evident between these results and those at Substation 17 for reasons not entirely clear at this time. More frequent summer irrigation increased boll Weight slightly at mid-harvest but not at late harvest. Additional Water also did not increase 39- I- 5 so - O I ll] Q Z J35- lO-DAV swucn IRRIGATION rnzoueucv "‘\ \\ / a4 - - - - - lo-oAv suuum nruumn raeouzucv \‘\ \ / / \ v ill L . . . . 4g. u—>o so so no aoo zoo soo 30o non-w o so co o so no no rouuosornmnnounncznmnca: Figure 27. Two-year average lint percents of 1517C cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Fabens (1959, 1961). 15 a") MID-HARVEST //‘\ \ \ , ,0 .4___ / \‘_____.,/ ~| z g___._.. LATE-HARVEST T-DAY SIMIIER IRRIGATION FREQUENCY -— -— — —- IQ-DAY GJNIAER IRRIGATION FREQUENCY WEIGNT OF SEED COTTON PER IOLL( a | l I 1 | 1.80 240 300 300 300 300 O GO I20 I80 I l I N——)0 5O 60 I20 GO 6O 6O 60 POUNDS 0F N AND P10! APPLIED FER ACRE Figure 28. Three-year average weights of 1517C seed cotton per boll at two harvest dates as affected by irrigation frequency and soil fertility level at Pecos (level irrigation, 1959-1961). seed weights at mid-harvest, and late harvest seed were definitely lighter with 10-day irri- gation. As will be seen from the tests at Pecos, Fort Stockton and Van Horn, the Fabens results were exceptional and indicate that some factor was disturbing normal plant functions to some extent. This could have been caused by excessive soil moisture or “Wet feet” effects at intervals during the fruiting period. Other effects were more or less similar to those at Substation 17. Later maturing bolls showed definite reductions in boll Weight, lint Weight, seed Weight and lint percent. These reductions occurred at all fertility levels and at both irrigation frequencies. The effect of nitrogen and phosphate on boll and seed weights was more apparent here than at Substation 17 . Boll Weights were increased ap- proximately 10 percent by the highest fertilizer rates at both irrigation frequencies. Both nitro- gen and phosphate were effective in this respect. Seed weights were increased in similar manner. One effect of more shallow rooting appears to be a greater response to applied fertilizer on these alluvial soils. The almost complete lack of re- sponse to fertilizer by outside skip rows at Sub- station 17 in 1961 adds weight to this conclusion. As at Substation 17, lint weights per boll were relatively unaffected by either nitrogen or phos- phate. The increases in seed weight with no changes in lint weight resulted in sharp reductions in lint 3- -. .- -4~ _ ’ ""' — — -*~ -_ 1- - g- ._ lulrumvzsffl __ __. __ -0- " _ ~*-’ ‘ ~-¢~ -' _._ -~ ___ H, , - - J __ _ _ § ergo/w ___--~*——+ "‘\_\\ ,_/_--.-t W1 I 2 l- 2 l- E L T-DAV SUMMER IRRIGATION FREQUENCY 1 m -— -— — - N-DAY SUMNER IRRIGATION FREUJENCY O ii l 2 Ill 3 0 | | l l l 1 | ; -——>O GO SO I20 IBO 240 BOO 3% 300 BOO PIOl-QO O GO SO 8O 5O O 5O 12D lOO PQJNDS OF N AND P20! APPLIED PER ACRE Figure 29. Three-year average weights of lint per boll of 1517C cotton at two harvest dates as affected by irrigation fre- quency and soil fertility level at Pecos (level irrigation, 1959-1961). 16 \\ /’,o___.__.. \§____._.._.___.0/ ,_ ‘LEE-HARVEST __,/ a r ‘I-DAY SUMNER IRRIGATION FREMIENOY —— -— —-—I4~DAV SUMMER IRRIGATION FREQUENCY velour OF sszn PER DOLL (gm!) ll - til I l | | 6O 6O I20 , , , .- u-->o s80 soo soo aw PaOa-w o o so no m ~ 6O 6O POLIIOS OF N AND P10! APPLIED PER ACRE _._ y. .- Figure 30. Three-year average weights of seed per boll’ l 1517C cotton at two harvest dates as affected by irrigation f _ quency and soil fertility level at Pecos (level irrigation, 1959-1961 percent as fertility rates increased. Lint percen were reduced 11/2 to 21/; percent by heavy fe l lization, regardless of irrigation frequency. Th effect was observed at Substation 117 but A much more apparent here. Pecos The considerable effect that different s conditions can have on boll properties is clear evident in the four tests at Pecos. As previousl described, the tests from 1959 to 1961 were c0 ducted on level runs with disastrous effects v yields from excessive soil moisture and adver soil physical conditions. The 1962 test on 0 percent slope gave greatly different yield resul because neither high soil moisture- nor poor ~- physical conditions were limiting growth. ' Figures 28-31 give the 3-year average o» characteristics data for the tests on level ru Figures 32-35 give similar data for the 1962 te on 0.4 percent slope. Acala 1517D, grown in t 1962 test, has similar boll properties to 1517 s and comparisons between the tests are justifi The effect of soil conditions on boll weig (Figures 28 and 32) was obvious. In the lev run tests there were no significant differences boll weights due to irrigation frequency. Ho. SI- 31- \ \ /.\\ um mvzsr //\\ \\ // \\w”+_~_°_—“”'// \\¢_.__ q/ 3C»- P--—-""" "\\ /¢\\ / \\___ _____4—-—— NIB-HARVEST ll u I LATENARVEST LINT PERCENT ll 0 I ‘I-IIAY SLIAIIER IRRIGATION FREOIENCY -— —-———l4-DAV SMIER IRRIGATION FREQUENCY s:- V J N-i O 60 6O IEO IIO E40 300 300 300 Pl0l—0 0 6O 60 6O G0 C0 IEO O POUNDS 0F N AND PIOI APPLE!) PER ACRE Figure 31. Three-year average lint percents of 1517C col at two harvest dates as affected by irrigation frequency and fertility level at Pecos (level irrigation, 1959-1961). um-a-umsr 1 r T‘ 1 i \ z.’ \ AYE—NARVE$L 'c\ ; \ / \ \- I / /‘\\\ / T-DAY SLMMER IRRMIATIGI FREQUENCY — — -—- -— IO-DAY SUMMER IRRIGATION FREQUENCY V n cit so tic nio zio aim s60 s50 ado , o so so so so o so IZO I80 ' ~ rooms or n AND m» Arman m: ACRE 32. Average weights of l5l7D seed cotton per boll “rvest dates as affected by irrigation frequency and soil jjrl at Pecos (0.4 percent slope, I962). 0.4 percent slope, frequent (7-day) irri- lproduced definitely heavier bolls at both F est and late harvest, and these differ- ere evident at all fertility levels except rogen at late harvest. The-re seems little (hat more normal growth processes were _ ted on the sloping runs. On level runs, ormal physiological processes were dis- particularly with frequent irrigation. t weights per boll also showed consider- ferences between the level and sloping 5§On level runs (Figure 29) 7-day irrigation _- in significantly smaller weights of lint ‘l at both mid-harvest and late harvest. ‘erse was true on sloping runs. As indi- Substation 17, more frequent irrigation 5i contributes to slightly heavier lint ‘ per boll. weights per boll likewise were affected {differences in irrigation design. On level i ‘gure 30) there were no significant differ- tween seed weights with 7 and 14-day f‘: at mid-harvest. On sloping runs f 34) mid-harvest seed were definitely t_ with 7-day irrigation. Clear-cut differ- tween these tests were also present in i - est seed weight data. lie striking differences in boll character- ere strongly reflected in lint percents : 31 and 35). Although higher lint per- . mums? ‘ f ; “.- i i i i __ __ i '— _.\ \ ‘ “T urc-vunvcsr 3 J ? Q - i z \ \ z , ,0- f . r g, ‘i ;;efl>' '1 *4 , ’¢0* "r-mv mamnou rncoucucv -- — - -—|4-uv 30mm: wanna-non mznozacv I I I l i so nio no no zoo soc soo soo e eo eo so so mo. no reunos or u no no; APPLIED m: ACRE l_ 33. Average weights of lint per boll of l5l7D cotton ,- est dates as affected by irrigation frequency and soil i; I at Pecos (0.4 percent slope, I962). MID-HARVEST aMJ b 1 ‘I-DAY SUMMER IRRIGATION FREQUENCY — — — —I4-DAY SUMMER IRRIGATION FREQUENCY WEIGHT OF SEED PER BOLL ( u | I O/l l 5O t | | | A | 4i____| N——) O 60 I20 I80 240 300 500 300 S00 P:0l—)0 D 60 6O $0 0 6O I20 IIO 5O POUNDS OF N AND P10! APPLIED PER AORE Figure 34. Average weights of seed per boll of l5l7D cotton at two harvest dates as'affected by irrigation frequency and soil fertility level at Pecos (0.4 percent slope, I962). cents occurred with less frequent. ( 14-day) irriga- tion on both level and sloping runs, the differences due to irrigation frequency were much greater with level irrigation. These averaged 1.3 and 1.0 percent for mid-harvest and late harvest, respectively, on level runs as compared with 0.5 percent and no difference on sloping runs. This further indicates the definitely adverse ef- fects of frequent irrigation on level fields on this soil type. The effect of maturity date on boll char- acteristics again was apparent in the Pecos tests. Later maturing bolls showed significant reduc- tions in boll weight, lint weight, se-ed weight and lint percent regardless of irrigation design or irrigation frequency. As in the Substation 17 and Fabens tests, nitrogen applications resulted in generally larger boll weights and seed weights per boll. Lint weights per boll again were little affected by nit- rogen . Consequently, lint percents were accord- ingly lowered by higher rates of nitrogen. Re- ductions in lint percent due to nitrogen were greatest on sloping runs. There was little evidence that phosphate had any effect on boll characteristics at the 60-pound nitrogen levels. However, additions of phosphate at the highest nitrogen rates showed some ten- dency to increase both boll and seed weights on the level border tests (Figures 28 and 30). 37F o—~*’\ \ LINT PERCENT u 0| l I -—i—7-DAY SUMMER IRRIGATION FREQUENCY —- —-— -— —l4-DAY SUMMER IRRIGATION FREGJENUY 33- .1 ~_><'> ei: sb uio tin zio sbo :50 aim ado nos-u: o so o so vzo ueo 6O 6O 6O POUNDS OF N AND P20! APFLED PER ACRE Figure 35. Average lint percents of l5l7D cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Pecos (0.4 percent slope, I962). 17 HID-HARVEST ~| I IO-DAY SUMMER IRRIGATION FREQUENCY / — ——- —I0-DAY IRRIG. TO lIlD-JULV—20-DAY IRRIG. T0 MID-SEPT. \ wsmur or szso cotton wen oou. (ants) Q I i i I I i i l I|->> O 60 60 I20 240 300 300 300 W00 FKQS-‘PQ 0 60 6O 6O 0 6O I20 I80 POUNDS 0F N AND P105 APPLIED PER ACRE Figure 36. Average weights of 1517C seed cotton per boll at two harvest dates as affected by irrigation frequency and soil fertility level at Fort Stockton (1960). Fort‘ Stockton The tests at Fort Stockton on deep, fertile Reeves loam in 1960 was primarily concerned with the effects of differential late-summer irri- gation frequency on cotton growth and yields. However, boll characteristics as well as yields and earliness were affected by the elimination of three late-summer irrigations. Figures 36-39 show the results of this test on boll weight, lint weight, see-d weight and lint percent, respectively. Lengthening the interval between irrigations to 20 days in late summer did not greatly affect boll weights or seed weights at mid-harvest. Late- harvest boll and seed weights, however, were greatly reduced in weight. Lint weights were not greatly affected, therefore most of the loss in boll weight can be attributed to lighter seed. Apparently, depletion of soil moisture in the 20- day plots in September and October was great enough to reduce strongly the moisture content of late harvest seed. Some leaf wilting was ap- parent on these plots before defoliation in early October. No wilting occurred on 10-day irri- gated plots. The lower late-harvest seed weights on the 20-day plots had a significant effect on lint per- cent (Figure 39). Cotton on these plots showed exceptionally high lint percents both at mid-har- vest and late-harvest. These averaged 38.3 and 38.9 percent, respectively, compared with 37.3 and 36.9 percent for 10-day irrigated cotton, an c!‘ i ‘F F IO-DAY SUMNER IRRIGATION FREWENCY I E - — ~—-—IO-DAY IRRIG. TO lllD-lllYim-DAY IRRIG- T0 HID-SEPT- 8 ‘g A ALL l- a‘ N——-)0 01) 00 2:0 S00 350 I Si!) h0l—b0 O 60 6O 6O I20 I00 POUNDS OF N AND PIOI APPLIED PER ACRE Figure 37. Average weights of lint per boll of 1517C cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Fort Stockton (1960). 18 MID-HARVEST . __ ’ IO-DAV SUMMER IRRIGATION FREQUENCY — —— --I0-DAY IIRIG. TO IAID-JULV-ZO DAV IRRIG. TO MID-SEPT. i WEIGHT OF SEED PER DOLL u a r | I .1 . i I i I I i i N-—)8 6O 8O I20 1R0 240 300 300 300 PlOl-D O 60 60 60 O 6O I20 60 POUNDS OF N AND PlOl APPLED PER ACRE Figure 38. Average weights of seed per boll of 1517C at two harvest dates as affected by irrigation frequency and. fertility level at Fort Stockton (1960). -‘ overall increase of 11/2 percent. Calculat, indicate that about 10 percent of the seed co yield reduction from less water (Figure '10) therefore due to loss of moisture from the and not an actual lint loss. A Although specific recommendations ca) be made without further studies, there is a ’ sibility that many Trans-Pecos cotton could profit from some moderate reduction irrigation frequency during August and ember. This would particularly apply to v rank cotton on finer textured soils. Benefi 1 be expected would be earlier harvest, hi, grades and higher gin turnout. Caution sh be exercised to avoid drastic water reduc with consequent yield losses as occurred in y test. » Although nitrogen had no effect on yiel this test, it did increase seed weights (Figure; and late-harvest lint weights (Figure 37). effect of phosphate on yields, previously n) can be seen in increase-d boll weights and q weights, particularly at the highest nit u’ levels. é Van Horn Results of the last test, on deep sandy ~l soil at Van Horn in 1962, substantiate prep conclusions in many respects. Data are 1 . - i \ an \\\\\./_; \ \ //\\\ //L£I"ELNALIE_ST*‘\\ \// \\ ‘\>\/\ \,/ A \\,/. \~\ / \ 3a \ {AID-HARVEST / \ \+___._/ \ ‘IIIWNARVEST LINT PERCENT u ~4 LATE- HARVEST on m lO-DAY SUMNER IRRIGATION FREOUDICY — —— ——IO-OAV IRRIG. T0 MID-JULY—ZO-DAY IRRIG» T0 HID-SEPI 1+ Q i i i N—<)0 60 C0 I20 I80 240 300 300 300 - PzOo-—)O O 6D 60 6O 6O 0 60 IEO POUNDS 0F N AND P10! APPLED PER ACRE ‘ Figure 39. Average lint percents of 1517C cotton v harvest dates as affected by irrigation frequency and soil ~ level at Fort Stockton (1960). I _-”. IAIDJIARVEST fl _,e_____ » ’g—” '.' z /'-- __.,’ LATE-HARVEST / \ / \ T?‘ //“ \ \./ \‘__ __ fl -e\ \ / ’ \ . I \ z f< I IZ-DAV SUMMER IRRIGAYION FREQUENCY -——-——-lO-DAY SUMMER IRRIGATION FREQUENCY r r ,4? ' I m} I l I CO SO I20 IBO 240 300 300 300 3M7 o - e0 o so |2o loo 6O 6O 6O FUNDS OF N AND PIOI APPLIED PER ACRE re 40. Average weights of 1517D seed cotton per boll {harvest dates as affected by irrigation frequency and soil ilevel at Van Horn (0.1 percent slope, 1962). ures 40-43. In this test, irrigation fre- ‘es from mid-June to early September were 12 and 18 days. = in other tests, more frequent summer irri- i, resulted in definitely heavier boll we-ights -harvest, with the effect still noticeable harvest. It also increased lint weights weights appreciably at mid-harvest. rcents were considerably lowered at all y levels at late harvest but only at high levels at mid-harvest. u" is test shows that more frequent irri- may result in heavier boll, lint and seed > , yet yields of seed cotton may actually Q =1: Extra" Water contributed to he-a- , lls but fewer bolls per plant, even though wnts were larger. This condition probably fin many fields where shading in tall rank often results either in excessive shedding and small bolls or in a tendency to 5 vegetative, with less total flower and boll ltion. At present, lack of sufficient light ff: to be the primary cause of lower yields in f1 se cotton. Answers to the problem may i‘ less frequent irrigation, (2) skip-row _, (3) wider row spacings or (4) fewer foot of row, any or all of which might =..-.- 1ve. l, e again, nitrogen increased both boll and Leights at mid-harvest, While again having _ fect on lint weights per boll. At late har- _ MID-HARVEST l. .___ ..- ——' ,_-O— -0- o» ”/0~‘\ Zf/k _ Lltt-uuwcer Il-DAV suuuzn unmemou rnrovcncv - - ---re-oev sumrcnilmegrlou rncouzncv = - t» » 24.0 - e50 3i» =60 =6» Q O0 6O 8O O 60 IIO IIO V POUNDS W N AND ROI APPLED PER AGRE i. 41. Average weights of lint per boll of 1517D cotton st dates as affected by irrigation frequency and soil y“ I at Von Horn (0.1 percent slope, I962). IIID- HARVEST ll-DAV SUMNER IRRIGATION FREQUENCY f; 3 a 4 I K HI l n -- - -——Il-DAY suuuzn mnrenton rnzoucncv l: . F‘ '8 '5 E g s O I l l l | I l l I l -—) $0 60 I20 I80 240 300 300 300 300 vibe-g O 60 0O 6O 60 O IO IIO IIO POUNDS OF N AND PIOI APPLIED PER ACRE Figure 42. Average weights of seed per boll of 1517D cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Van Horn (0.1 percent slope, 1962). vest, however, boll Weights and seed weights dropped significantly at fertilizer rates above the 180 N-60 P205 level. No logical explanation is immediately available. It may have been due to production of more bolls per plant at high nitro- gen levels on this sandy soil. Phosphate had no effect on lint weights but showed a tendency to increase boll and seed Weights slightly, particularly at high nitrogen levels. The effects of maturity date on boll characteristics again were evident. Effect of Irrigation and Fertilization on Fiber Properties The quality of cotton fiber is currently based on measurements of the three characteristics, length, strength and fineness, each of which is highly genetic in nature. In most presently grown varieties, these fiber characteristics are the result of many years of breeding research and are not subject to appreciable variation caused by environmental conditions. Within any particular variety these three characteristics vary within narrow limits, and for this reason, variety name a, 37- 38- U u I IE-OAV SIHNER IIIIATIN FIEUQCY urn’ rcnccur 2 I — — -— -—lI-OAV 31H IRNIOATIGI FREQUENCY 33- I O N——IO CO IQ I00 I00 300 300 240 300 P10! O O IO 6O 6O 6O O 60 I20 I00 —. POIINOQOFNINOPIOIAPPLEDPERAORE Figure 43. Average lint percents of 1517D cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Van Horn (0.1 percent slope, I962). 19 L12 - MID~NARVEST ___ ¢ zo- 0/ \ / .16 - IO-DAV QUMMER IRRIGATION FREOUENC — ———-—IS-DAY SLNMER IRRIGATION FREQUENCY FIBER LEN6YN—2.S$ SPAN (HOMES) HOE til i I I i i i l i l ; N—) 0 6O 60 I20 240 300 300 300 O0 Q O 6O 3 P 0 -)0 80 I20 I80 l ' POUNDS 68F N AND P205 APPLIED PER ACRE Figure 44. Three-year average fiber length (21/1 percent span) of 1517C cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Substation 17 (1959-1961). has come to be almost synonymous with a specific quality of fiber. In order to determine how much, if any, variation in fiber properties could occur because of irrigation and fertility differences, fiber sam- ples from all 1959-61 tests were analyzed in the fiber laboratory at Substation 17. Analysis of 1962 samples had not been completed, so these results could not be reported. Fiber Length Fiber length measurements are complicated by the fact that any ginned lint sample contains strands varying greatly in length. The most significant measurement of fiber length, or staple, has for many years been the “upper half mean,” a measure of the average length of the longest half of the fibers in a given sample. In recent years a new digital fibrograph has come into use which measures a similar value called the “21/5 percent span length.” For all essential purposes, “upper half mean” and “21/3 span length” are equivalent and comparable. Figures 44-46 give the 2 or 3 year average 21/2 percent span length data for the Substation 17, Fabens and Pecos tests, respectively. Figure 47 gives the 1-year data from the Fort Stockton test. I28- L24 - /°\ \ flit-HARVEST Y/ \‘___,___. .20 - IO-DAV §IIMER IRRIGATIGI FREQENOY — — -— —l6-DAY SUMMER IRIDATICN FREQIENCY men LENOTN—2J$ SPAN (mcnzs) 5 1 :‘$“l) ‘i, gPOIJGDS OEIET: AND Plllgg APPLIED ACRE 3°00 328 fig Figure 45. Two-year average fiber length (21/1 percent span) of 1517C cotton at two harvest dates as affected by irrigation frequency and soil fertility level at Fabens (1959, 1961). 20 \ / \ mu vunvzsr / \ / " \ z/K \ us - . \ fie HARVEST ‘ / i s / \‘* - —~\ ~x~v x/ \ I J4- ‘I-DAY SUMMER IRRIGATION FREQENCV — -— -—-—l4-DAV SUMMER IRIIOATIQJ FREGIENGY FIXR LENGTH -—-—2.I$ SPAN (INCHES) LII F 1 . ' i Q l i i i i Am _ lT-Si: oi: no soo soc moo-m o o m I20 POUNDS OF N AND P20! APPLIED PER ACRE Figure 46. Three-year average fiber length (21/2 percent sp. of 1517C cotton at two harvest dates as affected by irriga frequency and soil fertility level at Pecos (level irrigation, 19 1961). ’ In all tests, one fact becomes immediat apparent. Late-harvested fiber was somew shorter than mid-harvested fiber, regardless irrigation frequency or fertility level. This eff 9 of maturity date on fiber length was signific in all tests, differences averaging between 0. and 0.04 inches in each. However, these red tions in length with later maturity were s and might escape detection by cotton classers. It is important to recognize that matu T I date does have an effect on fiber length in t area. Most late-harvested bolls are set, and th fiber length determined, in late August and e q September. The generally cooler temperatu and less intense sunlight at this time of year i? be responsible for the shorter fiber. Whate the cause, this again emphasizes the importa of striving for early maturity. - The effect of irrigation frequency on fi length was not the same in all tests. The Fa and Pecos tests showed no significant differen at either maturity date. Those at Substation; showed a slight but definite length increase f 10-day irrigation at mid-harvest but not at 1.20 — iie / A \ ' - \ /°_ '_ '7 '_\ K / ‘ / //\\ __ / \mo-n4nvzsr / us r- 0/ \ / _ V l?‘ W" unsmmvssr i \ \ _-_ - \ /__ __ _\ E \ / ‘ 5i L12 - \\ /— - - -¢\ / u \ / w. \ \ / \ T - \ / v E \\ / a l.l0 - \ \/ -I IO-DAV summer: IRRIGATION rncoucucv § ' — —-- -—I0-DAV inmo. 1o uio-.|uu—2o-oAv lame. 1o MID-SEPT. c LOH; (fl I i i i I i i | _l u—T so so izo 24o aoo w: n»; rum-so o o so izo 6O . 6O FWNDS 0F N AND P10! APPLIED PER ACRE Figure 47. Average fiber length (21/1 percent span) of 1, cotton at two harvest dates as affected by irrigation frequenc’ soil fertility level at Fort Stockton (1960). 1 Q LA s minvzst /._ _ _ IO~DAY SIIAIIER IRRIGATION FREQUENCY — — —- —I8-DAY SUMNER IRRIGATION FREQUENCY .|_ l i i i ;_ I I20 2 300 300 300 300 O 6O I80 O0 O0 FUNDS OF N AND P100 APPLIED PER ACRE 48. Three-year average micronaire (fineness) of 1517C two harvest dates as affected by irrigation frequency and lty level at Substation 17 (1959-1961). “t. Results at Fort Stockton, where late- “: soil moisture was very low in certain plots j» a significant difference (favoring 10-day “ 'on) at late harvest but not at mid-har- It appears that the effect of irrigation on fiber length may be toward slightly (fiber with more frequent irrigation but differences, when they occur, are small obably not of any practical importance. Lber length also appeared to be relatively ._; ted by soil fertility level. A tendency at _ toward increases in length at higher fert- ‘vels was small and not apparent in the ’ sts. ‘fFineness 'eness of 1517C fiber, as measured by the ‘aire, ranges from a high of about 4.2 at f. est to a low of about 3.0 at late har- §~The micronaire measurement is performed f ‘ly to determine fiber maturity. For the ‘l: of high quality yarn, mills prefer that tton have a micronaire- value of 3.5 or l‘ Values lower than this usually indicate egree of immaturity, with consequently __ spinning qualities. Strength of individual . is also closely related to fineness. um-unrvzst , /"* [*1 .__ ; _.\ ,- *\.._n""._"""_ "—O~v-vo-wv-O”? lD-DAY SUMNER IRRIDATDN FREGJENGY —-———l6-DAY SUANER IRRIGATION FREQUENCY i; :.. _i "i. '5: :1: i’ =é2 is v POLMDS OF N AND P10! APPLIED PER ACRE Y, . 49. Two-year average micronaire (fineness) of 1517C two harvest dates as affected by irrigation frequency and v' level at Fabens (1959, 1961). ‘I-DAY SUMNER IRRIGATION FREQUENCY — -- -— —I4-DAV SIMMER IRRIGATION FREQUENCY NICRONAIRE READING no u I l m i i I i i L i i g Nib!) SO 5O I20 IBO 240 300 500 SIX) 3% P:Oo-—)O O 6O 60 60 0 60 I20 I00 G0 POUNDS OFWANDHOIAPPLIED PERACRE Figure 50. Three-year average micronaire (fineness) of 1517C fiber at two harvest dates as affected by irrigation frequency and soil fertility level at Pecos (level irrigation, 1959-1961). Figures 48-50 show the 2 and 3 year average micronaire measurements of fiber samples from the Substation 17, Fabens and Pecos tests, res- pectively. Figure 51 gives 1-year data for the Fort Stockton test. All values reported are aver- ages of four separate measurements. It can readily be seen that irrigation fre- quency had little or no direct effect on fiber fine- ness in any of these tests. This was true of both mid-harvest and late-harvested cotton. Minor fluctuations in the curves must be attributed partly to variation inherent in the micronaire procedure and in boll sampling. However, insofar as its effect on maturity is concerned, irrigation frequency had a significant effect upon fiber ‘fineness. Tests at all four locations showed definitely lower micronaire readings (finer fiber) at late harvest than at mid-harvest. The effect of more frequent irrigation on delay of maturity (previously discussed) would therefore also con- tribute to fiber immaturity. This fact is of con- siderable importance where high quality cottons such as 1517C are grown in areas with limited growing season, such as the Trans-Pecos region. It appears that bolls set progressively later than September 1 in this area will produce fiber that tends to be more and more immature, regardless of water and fertilizer management. The factors responsible are probably both climatic and phy- siological. § 1 IO-DAV SIDAIER IRRIGATION FREQUENCY u I — — —-—-IO-DAY RRID. TD ND—JULYi20-DAV IRRIG. T0 NlD-SEIW. MICRONAIRE READING ....—:é is a‘; #%=~‘~ =2" =2: POUNDSOFNANDPIOIAPPLIEPERAGE Figure 51. at two harvest dates as affected by irrigation frequency and soil fertility level at Fort Stockton (1960). 21 Average micronaire (fineness) of 1517C fiber I06 — FIIER STRENGTH (S 0F I5|7C STANARD) lO-DAY SUMNER IRRIGATION FREWENCV — — — —-I6-DAY SUMMER IRRIGATION FREWENCY .1 n l | l I l | N—-)O 60 6O I20 300 I I | I80 240 300 300 300 PlOl-v O 0 $0 60 60 60 O 60 I20 I80 POUNDS OF N AND PIG! APPLIED PER ACRE Figure 52. Three-year average strength of 1517C fiber at two harvest dates as affected by irrigation frequency and soil fertility level at Substation 17 (1959-1961). The effect of nitrogen on fiber fineness is also related t0 boll maturity date. Both mid-har- vest and late-harvest data at all locations show that nitrogen additions did not directly affect micronaire readings at either irrigation frequency. Nitrogen did, however, increase yields signifi- cantly in all but the Fort Stockton test, and these yield increases from the higher nitrogen plots were largely in the form of heavier late-season pickings. In this respect, then, nitrogen also could be said to contribute to fiber immaturity. These data again emphasize the importance of proper nitrogen fertilization. Without ques- tion, cotton requires some nitrogen for most pro- fitable production. Yet nitrogen in amounts above that required for good yields seldom results in higher yields, always adds to production costs and may contribute to fiber immaturity, invariably found in late-maturing bolls. There was some indication (Figures 49-51) that phosphate additions at the 300-pound nitro- gen levels had a slight tendency to improve micro- naire readings. This was probably because of greatly unbalanced fertility conditions, since 04- uvc-umlzg, A \ \ IO-DAV SUIIIER IRRIGATION FREQUENCY —— —— ——-—l6-DAY SUMNER IRRIGATION FREQUENCY mu svnzvwm (z or lunc sum-um) 2 I O/l 1 n n _|_ n n | 1 I l N—) O 8O 80 I20 IIO I40 300 S00 300 S00 Pl0l—)0 0 6O 00 C0 IO O O0 I20 I00 PGMOSOFNAMSPROIAIPLIDPHAOII Figure 53. Two-year average strength of 1517C fiber at two harvest dates as affected by irrigation frequency and soil fertility level at Fabens (1959, 1961). 22 no effect of phosphate was observed at the Pf pound nitrogen rates. Fiber Strength . The Acala 1517 cottons have always be noted for their exceptionally strong fib Strength is of utmost importance to the spinn’; industry, for it largely determines the stren and durability of finished= goods. With the - vent of processed fabrics, strength has beco even more important, because the “wash a wear” and “crease resistant” processes have» somewhat weakening effect on cotton fiber. ..F these reasons, cotton breeders are striving» c0 stantly to develop strains with greater and grea fiber strength. Within any particular cotton varie strength of individual fibers is an indicator- fiber maturity. Fully mature fibers invaria are stronger than less mature fibers whose . walls have not been fully developed. Stren and fineness are therefore closely related. The situation is complicated by the fact t’ the current stelometer method of measuring fi strength actually measures the breaking point I a given weight of parallel fibers, rather thant strength of individual fibers. Thus, the st . meter values may be misleading. The situa ‘y may be explained as follows: A heavy-gage "1 . may be much stronger than a fine-gage " yet if enough fine-gage Wire is combined to f0 an aggregate as thick as the heavier wire, . strength of the combined fine wires may greater. This, as will be seen, can be true cotton fiber. I Figures 52-55 show that the stelome strength values for the samples from the a at four locations. These values are repo here as percentages of the breaking stren of standard 1517C samples, which were run l currently. As in the micronaire procedure, so variability was inherent in the stelometer ; surements, which largely accounts for the zag nature of the curves. Figures 52-54 show the effect of fiber f' ness on stelometer-breaking strength. Ag’ gate samples of finer late-harvest fiber sho somewhat higher breaking strength than i" mature mid-harvest fiber. This was- proba because of the larger number of individual fig" making up the late-harvest stelometer samp This tends to obscure the fact that individ less mature, late-harvest fibers are usu weaker than those maturing earlier in this a Figures 48-50 show that late-harvested fiber considerably finer than mid-harvest fiber, ‘I breaking strengths of late-harvest samples w only slightly higher. This strongly indicates l ' strength for the late-harvest fiber, although q actual magnitude of the differences cannot . determined. ' Neither irrigation frequency nor soil fed lity level had much direct effect on fiber stren Overall values showed no significant differe y ‘ tests at Substation 17 and Fabens, al- ‘i there was a tendency for more frequent to produce slightly weaker fiber. At a here soil physical conditions on level runs versely affected by frequent irrigation, ‘was a definite weakening effect on mid- fiber, although actual percentage differ- ere not great. Only at Fort Stockton, _idely different late summer irrigation, reciable differences in strength noted. ' tic reductions in late-season irrigation 1'» in significantly stronger mid-harvest t all fertility levels. These differences idisappeared at late harvest. of Irrigation and Fertilization on Incidence of Verticillium Wilt yre is considerable research evidence to f: t both irrigation frequency and nitrogen 'ons have a significant effect on the in- of Verticillium wilt, caused by the soil- ungus V. albo-atmm. This pathogen at- wide range of cultivated crops and is y: ly damaging to irrigated cotton on the fxtured soils of the Southwest. This qcan exist in cultivated soils for many :3: but it multiplies rapidly in the presence gvceptible crop like cotton. It enters the ;_- ough the- roots and gradually spreads out the plant. It apparently damages ‘f t chiefly by plugging the sap-conducting Qwith some metabolic product. This shows y‘ Although the plants seldom are killed, -=~ causes leaf wilting and leaf necrosis, l bsequent shedding of all leaves and L in severe cases. This usually occurs in ‘Vmer during the boll-maturation period. edding causes virtual cessation of normal nctions. The two most damaging effects pure of fiber to mature in bolls already set iuctions in yields caused by the absence “al late-season boll set. Fiber from verti- f, infected plants is always of inferior l. At present, there is no satisfactory con- ithis disease. e effect of irrigation frequency and soil } level on severity of verticillium wilt was 1 in" the tests at Substation 17 during 1959- e test area was moderately infected with g ore 1959, but the disease became more Eton certain plots as a result of irrigation ilization during the test period. To study unts of visibly wilt-infected plants were all plots in September 1959. The plots it replanted in the identical locations in l study the cumulative effects of two years _‘ar management. Figure 56 shows the number of infected plants per plot for s tment in each of the two years, ex- ~ as percentages of the total plant popu- eral conclusions can be drawn from these I It is apparent that adverse effects 'sh discolorations when infected stems . nun srncmrn (t or lll7O srmomo) ‘I-DAY SIMIIER IRRIGATKM FEOUENCY -— -— -— —- li-DLV $UNMER IRRIBAYION FREQUENCY N-Jo 6b 6o non-v o o so | | l A 300 300 300 300 6O I20 IIO I00 ‘ O0 0 POUNDS OF N AND ROI APPLIED PEI ACRE Figure 54. Three-year average strength of 1517C fiber at two harvest dates as affected by irrigation frequency and soil fertility level at Pecos (level irrigation, 1959-1961). caused by high soil moisture were major factors affecting the increased severity of the disease. In both 1959 and 1960, plots getting more fre- quent irrigation showed the highest percentage of infection. Frequent irrigation accounted for an average disease increase of 37 percent in 1959 and 58 percent by the fall of 1960. This cumu- lative effect indicates that continued frequent irrigation can cause rapid buildup of verticillium within only a few years’ time. The data also show that applications of nit- rogen may be instrumental in increasing the in- cidence of the disease. Although not so evident ‘in 1959, the effects of nitrogen became more ap- parent in 1960 with more frequent irrigation. The combination of high soil moisture and high nitrogen brought about a rapid increase in the percentage of infected plants, either by a buildup of the organism in the soil or by making the plants more susceptible to infection, or both. Under certain conditions, the application of phosphate also tended to increase wilt in- cidence. This was evident in 1960 with frequent irrigation at high nitrogen levels. The effect of phosphate was variable, however, and not so consistent as that due to nitrogen. These re- O a I u u I I A men srmaorn (I. or l5l7O smmuo) no-uulvzsn IO-DAV sumsn inmomou rneowncv — —~— —l0-DAV IIRH. TO IID-NLY—IO-DAV IRRHI. YO HID-SE?! IOr- V "ltd; ob .5 44f solo sob ado scio ado 9:01-00 0 SO Pmmnswnmohmfiingélge O GO IIO IIO Figure 55. Average strength of 1517C fiber at tvvo harvest dates as affected by irrigation frequency and soil fertility level at Fort Stockton (1960). 23 PERCENT INFECTED PLANTS lO-DAV SUMMER lRRlGlYlCN FREQUENCY '0 -— -—— ———|5-DAY SJMIER IRRIGATION FREQUENCY 1 0 l l | I l l l | n 1 N——§0 60 O0 I20 lUO 240 300 300 300 330 PcOi—>0 0 60 5O GO 60 O 6O I20 IBD POUNDS 0F N AND PIOO APPLIED PER AGE Figure 56. Severity of verticillium wilt (percent of infected plants) as affected by irrigation frequency and soil fertility level at Substation 17 (1959-1960). sults and those of other research workers present strong evidence for conservation of both water and fertilizer on wilt-infected soils. Once soils are severly infected it becomes almost impossible to grow cotton profitably, and the disease persists as long as susceptible crops are grown. Need for Other Nutritional Elements Soils of semi-arid desert regions such as Far West Texas are formed primarily through the physical processes of rock weathering and contain large amounts of primary minerals. These minerals decompose slowly under irrigated cropping and supply the crops with small but ample quantities of various minerals required only in trace amounts for normal plant growth namely, iron, manganese, zinc, copper, boron and molybdenum. Potassium, calcium and magnes- ium, required in relatively large quantities, are also derived from these primary minerals in amounts more than sufficient for normal growth. In addition, all waters used for irrigation contain appreciable amounts of soluble salts which supply Figure 57. Upland cotton severely infected with verticillium wilt in the El Paso Valley. adversely affected. 24 Both yields and fiber quality are large additional quantities of potassium, sodiu calcium, magnesium and sulfate, as well as tr amounts of boron and other elements. Irri tion waters alone can add from 300 to 600 poun each of soluble sodium, calcium, magnesium a sulfate, as well as 100 to 150 pounds of potassi per acre per year to cultivated soils. There is virtually no possibility that any the so-called secondary elements (calcium, m’ nesium, sulfate) could become deficient in Tra i Pecos irrigated soils. Numerous field tests , this area and in New Mexico and Arizona al have failed to show any response of cotton. additions of potassium. Likewise, many‘ I tests have shown more than ample quantities i soluble potassium to be present in these so' regardless of soil texture. It is highly unlik that any except possibly very sandy soils W0 require additions of potassium for normal co { growth. More potassium is added in the wa each year than is removed in the cottonseed. y lint is virtually pure cellulose, and contains . fertlizer, secondary or trace minerals. To be certain that cotton did not requ’ additions of any of the trace minerals (ir manganese, zinc, copper, boron, molybdenum) ’ normal growth and fruiting, each of the 19 and 1960 tests discussed here contained 24 pl‘ in which these six elements were added in ad tion to nitrogen and phosphate. FERRO Corp., material FN-501, containing six elements was sidedressed with the fertil' at the 180 N-60 P205 level. In 12 other plots" soluble spray material MULTI-TRACIN1, T containing all six elements, was applied as a i. age spray three times during the season at same fertility level. None of the plots recei ' trace elements in any of the tests showed ;_ visibe response. Also, there was no response 1 yields to any of these treatments which could _ attributed to the trace elements. To determine if any response could be i tained from trace element applications on =1 coarse-textured soils, a 7 x 7 Latin square fi test was conducted on the D. H. Brewster at Van Horn in 1958. This test, on deep Ree loamy sand, included five formulations of tr minerals, four applied to the soil and one foliar spray, each with and without nitrogen phosphate. Here again, no response in cot growth or yields was observed from any of applied materials. It must be concluded f these tests that no benefit would be obtai from applications of any trace elements to co v _. on any Trans-Pecos soils at this time. Acknowledgments The authors wish to acknowledge the ex ' lent assistance of Fred Buckland, D. H. Brews and Chandler Farms, Inc., for their fine coo tion in the establishment and carrying out of i operative field tests which provided much of information contained in this bulletin. i ‘Distributed by Dril-Kem, Inc. In 12 plots, _