N M s u u B R A R Y B-1 1 27 December 1972 INFLUENCE OF PLANT GROWTH STAGE AND ENVIRONMENTAL FACTORS ON THE RESPONSE 0F HONEY MESQUITE ' '»\\.. l’ .2, (T: v31 a ' ,_ t . -_ '_'. ._». to ‘M n Stapncedei _ 66v? ‘ -.j ‘n. v‘ i‘ ‘ _ X l ‘ z bl/wvfl - ~H§ I i," i Duplicated LAS CRUCES The Texas Agricultural Experiment Station J. E. Miller, Director, College Station, Texas Texas A&M University In cooperation with U. S. Department of Agriculture, Agricultural Research Service Mention of a trademark name or a proprietary product not constitute a guarantee or warranty of the product by U.S. Department of Agriculture or The Texas Agricult Experiment Station and does not imply its approval to exclusion of other products that also may be suitable. Contents Summary ........................................................................... .. 4 Introduction .................................................................... -- 5 Materials and Methods ................................................. .. 6 Experimental Site and Plot Layout ..................... .. 6 Chemical Applications and Control Ratings .... .. 6 Plant Characteristics ............................................... .. 6 Environmental Variables ......................................... .. 7 Statistical Analyses ................................................... .. 7 Results.-. . l ............... .................. .. 8 Plant Response to Herbicides ............................... -. 8 Plant Characteristics ............................................... .. 9 Environmental Variables ........................................ ..ll Statistical Analyses . ......... "l3 Discussion .... .. .. .-.l6 Acknowledgment ........................................................... ..l8 Literature Cited ............................................................. ..l8 l7. Summary Honey mesquite [Prosopis juliflora (Swartz) DC. var. glandulosa (Torr.) Cockerell] plants 4 to 6 feet tall grown on an upland, clay loam site near Bryan, Texas, were sprayed with three herbicides at 14 dates during 1969 and 1970. Herbicide treatments con- sisted of 0.5- and l-pound per acre rates of the potas- sium salt of 4-amino-3,5,6-trichloropicolinic acid (picloram) and the butoxyethanol ester of 2,4,5-tri- chlorophenoxy)acetic acid (2,4,5-T) alone and in 1:1 mixtures. Mixtures of the two herbicides as the tri- ethylamine salts were applied at six dates at appli- cation rate of 0.25 + 0.25 and 0.5 + 0.5 pound per acre. Plant factors measured were new stem elongation growth, transectional stem dimensions, upward dye movement rate in the xylem, total available stem carbohydrates and leaf moisture stress. Environmental factors included were air temperature, percent rela- tive humidity, soil temperature, soil moisture, rainfall and total daily solar radiation. Picloram and a mixture of picloram + 2,4,5-T lwcaused 5s and 54, and s2 and ea percent defoliation l year after treatment at the 0.5- and l-pound per acre rates, respectively, whereas 2,4,5-T caused 44 and 48 percent defoliation. Picloram at 0.5 and 1 pound per acre and picloram + 2,4,5-T at the 0.25 + 0.25- and 0.5 + 0.5-pound per acre rates killed l1 and 12 percent and 23 and 27 percent of the plants, whereas 2,4,5-T alone killed only 2 percent of the plants at both rates. Control by the triethylamine salts of picloram + 2,4,5-T was equal to that of the mixture of potassium salt of picloram + the 2,4,5-T ester. ' Most effective control of honey mesquite occurred from treatments applied between April 30 and July 6. Plant characteristics most closely associated with con- trol were widest translocating phloem thickness, most rapid rate of new xylem ring radial growth and lowest predawn leaf moisture stress. Environmental variables most closely associated with honey mesquite control were lower maximum air temperatures of 77° to 96° F l week before treatment, maximum soil temperatures at 63° to 79° F at a depth of 3 feet l week before treatment and decreasing percent soil moisture from 4 25 to 18 percent at a depth of 2—3 feet l week befor treatment. . The higher percent defoliation and percent dea plant ratings were most closely associated in ord of effectiveness with herbicidal treatments of piclora n '1 mixture of picloram + 2,4,5-T, and 2,4,5-T. The d; foliation correlations were higher than those for th percent of dead plants. Generally, thickness of tran’ locating phloem, rate of upward dye movement =1, the xylem, lower minimum relative humidity, if soil moisture and higher rainfall ‘before spraying we directly correlated with higher plant control, whil . measurements of new xylem thickness, air tempera ture, maximum relative humidity before sprayin= maximum soil temperature, rainfall after sprayin and leaf moisture stress were inversely correlated wi high percent control by herbicide treatments. i Simple correlations of environmental data show ti‘ two groups of variables. Air and soil temperatu Q were directly correlated. Percent relative humidity percent soil moisture and rainfall were directly cor related, but they were inversely correlated with ai and soil temperature. Regression equations w g developed for estimating percent defoliation and per cent dead plants for both rates of picloram, piclor w‘? salt + 2,4,5-T ester and 2,4,5-T at all 14 dates of appli cation. Moisture stress of the honey mesquite leav varied from a low level in the morning to the high p level at midday to a low level again at night. Th stress values at night increased slightly, and the max‘ mum level during the day remained longer as th summer progressed. Within 2 or 3 days after herbicide application leaves of treated plants had a lower moisture str , level during the day than the untreated leaves. ~' the fourth day, treated leaves began dying, so th became more stressed than untreated leaves. Leav sprayed with picloram and a mixture of picloram 2,4,5-T tended to turn brown and remain on plants; those sprayed with 2,4,5-T tended to tu :- yellow and lose their leaflets from the rachises befo *- either dying on the plant or abscising. i Plant Growth Stage ental Factors s: HONEY MESQUITE R. E. MEYER R. W. BOVEY W. T. MCKELVY T. E. RILEY‘ HONEY MESQUITE [Prosopis juliflora (Swartz) DC. var. glandulosa (Torr.) Cockerell] varies widely in its response to herbicides. In some cases almost all plants are killed by a given treatment; at other times very few plants are killed. In West Texas, when growing con- ditions are favorable, 0.5 pound per acre of (2,4,5- trichlorophenoxy)acetic acid (2,4,5-T) generally de- stroys most top growth and kills about 20 to 30 per- cent of the plants (5). The most effective treatments occurred 5O to 80 days after the first leaves appeared in the spring and leaves were fully formed and dark green. Treatments were ineffective when applied be- fore this time or during summer and fall when the plant was not actively growing. Dahl et al. (2) reviewed the literature on varia- tions of plant response to herbicides. They found that soil temperature of 80° F and above at the 18-inch depth was the most important factor affecting the response of honey mesquite to 2,4,5-T. No plants were killed when soil temperature was in the low 70's or below. The easiest plants to kill were those having mature, dark green foliage and mature legumes. Trees on upland and sandy soils were apparently more susceptible to 2,4,5-T than those on bottomland and clay sites because of the difference in soil temperature. Robison, Fisher and Cross (l5) and Fisher et al. (6) have shown that honey mesquite is more suscep- tible to mixtures of 2,4,5-T + 4-amino-3,5,6-trichloro- picolinic acid (picloram) than to 2,4,5-T. In six ranch tests, picloram + 2,4,5-T at 0.25 + 0.25 pound per acre killed an average of 52 percent of the plants as com- pared to 21 percent for 0.5 pound per acre of 2,4,5-T alone. Thus, factors that prevented the death of all plants were present. Meyer et al. (ll) showed that the toxic agent from 2,4,5-T, picloram and picloram + 2,4,5-T sprays was translocated from the leaves to the stem of honey mesquite within 4 days after application. Brady (1) "Respectively, plant physiologist, research agronomist, biological laboratory technician and agricultural research technician, Agri- cultural Research Service, U.S. Department of Agriculture; The Texas Agricultural Experiment Station (Department of Range Science). 5 found herbicide applications in May to be more effec- tive than those made later in the growing season on sweetgum (Liquidambar styraciflua L.), green ash (Fraxinus pennsylvanica Marsh.) and water oak (Quercus nigra L.). Effective killing of plant tops was attributed to the high rates of absorption and translocation of the herbicide. A 4-day absorption period was closely correlated with the control of top growth l year later. Davis et al. (3) sprayed honey mesquite with 0.5- and l-pound per acre rates of 2,4,5-T, picloram and a mixture of picloram + 2,4,5-T. After 48 hours, highest concentrations of herbicides occurred in the phloem of those plants sprayed in June and lowest in those sprayed in August. Similar levels of 2,4,5-T occurred in the phloem from applications of either 0.5 or 1 pound per acre, but more than three times as much picloram occurred in plants sprayed with 1 pound per acre as in those sprayed with 0.5 pound per acre. At College Station, Texas, honey mesquite leaves begin emerging about the end of March (14). Emer- gence is probably controlled by environmental condi- tions of high temperatures (10). The new stems elongate during approximately a month beginning about the first of April until tip abortion occurs (14). Then the plant enlarges radially in May and June, producing new translocating phloem and a new xylem growth ring. Fisher, Fults and Hopp (4) and Meyer, Haas and Morton (l2) showed that upward transloca- tion of dye occurred in the new, outermost xylem ring of honey mesquite; the dye streak widened from about 0.16 inches at the point of injection to about a 2-inch width at 4 feet and moved about 1.7 feet per hour. Meyer, Haas and Wendt (l3) using greenhouse plants under controlled conditions showed that either low soil temperature (55° F) or cool aerial environ- ihent (52° to 77° F) retarded shoot growth; however, maximum shoot growth occurred as a result of an interaction of optimum soil temperature (84° F) and an aerial environment of 20.9 millimeters of mercury (mm Hg) vapor pressure deficit (74° to 105° F). Soil- moisture regimes from —0.5 to ~15 bars soil water potential did not impede growth. Fisher et al. (5) and Robison et al. (16) studied the total available carbohydrate level in honey mes- quite stems. That carbohydrate level was low in May could reflect the partial drain of food reserves for foliage production and radial enlargement of stems. Haas and Dodd (8) studied water stress of the honey mesquite leaf petiole. They showed a diurnal pattern of low stress at predawn and postsunset periods and high stress during the day. A gradual increase in stress occurred at all three periods through- out the season when leaves were present. The objectives of this study were to develop a reliable means of estimating the ultimate response of honey mesquite to herbicides, to determine the 6 I interrelationships of various plant and environmen -r variables and to observe the seasonal response of hon ' mesquite to various herbicide treatments. ~ Materials and Methods Experimental Site and Plot Layout A A 30-acre site near Bryan, Texas, with a den stand of honey mesquite plantsv4~fito 6 feet tall selected as the experimental site. Most honey g quite plants had three to five stems that had emerg near the base of the stem. The plants had r5 mowed about 3 years before the initiation of p study. The area was an upland site with about a l-l cent slope. The soil was a poorly drained clay loa In 1969, about half the area was divided in plots containing five plants each. The plots =5 arranged in rows 20 to 40 feet long depending = plant density and width sufficient to contain at five plants with at least a 2-foot space between pla from the adjoining plot. In 1970, the remaining :1 of the plants were tagged in groups of six for ~_ last five sprayings. Four replications were used _ each treatment at each of 14 dates. Four untreat plots were also included at each date, but the res 5 were not included in statistical analyses. " Chemical Applications and Control Ratings (A The chemicals were sprayed at 0.5- and l-pou per acre acid equivalent rates at 14 dates. The p0 v sium salt of picloram, the butoxy ethanol ester - 2,4,5-T and equal amounts of two herbicides (piclo - + 2,4,5-T) were sprayed at all dates. A mixture _ equal amounts of triethylene salts of picloram 2,4,5-T (picloram + 2,4,5~T amine) at equal rates applied at six dates as a comparison with the po ~ sium salt of picloram and the ester of 2,4,5-T. The herbicides were applied either in the even‘ or early morning with a hand-carried, compressed - j S-nozzle boom sprayer. The herbicides were applied to the plots or individual plants at a spray vol a of 20 gallons per acre. a Visual ratings of percent defoliation, whiff measures the amount of stem tissue killed, and perc dead plants were made the season following spra ' October 3, 1970, and May 17, 1971. 5 Plant Characteristics Plant characteristics measured at each date ’ treatment included new stem length, transectio dimensions of stern tissue, rate of upward move |-_~ of dye in the xylem, total available carbohydrates the stem and leaf moisture stress. Data, except f some moisture stress readings, were collected on i sprayed plants at the site. 3 New stem length growth was recorded by ta‘ five new stems on each of five plants. The w- 9 were measured weekly until growth ceased and the , in May. They were observed periodically ,- to determine whether further elongation ,_i“-~ transactional dimensions were measured ig-u samples collected from untreated trees the l spraying. A piece of stem 6 to 12 inches ground was collected from each of l0 trees. " piece was cut into smaller cylinders, fixed i‘ ' killing solution (l4) and ultimately. stored ent ethanol. Sections, l5 to 25 microns ~ cut either from blocks embedded in Para- "from une_mbedded blocks mounted directly crotome. The transections were stained with and fast green. Tissue measurements were V f: three places equidistant around the circum- ,; f each stem transection. Measurements were gperideim, translocating and nontranslocating xylem ring thickness. Pith diameter was frded. - o movement of methylene blue dye in the - s followed by infusing the dye into unsprayed i‘ ~- were about 5 feet tall with reasonably ~ unforked stems. The tip of a 1-inch long 5‘ needle, attached to a medical transfusion _ , was inserted just under the stem bark about ‘fiabove- the ground. A 0.1 percent aqueous f} of the dye was allowed to infuse into the fr“ 9 to 9:30 or l0 a.m. After the allotted time 'on and translocation, the bark was peeled e length of dye streak measured. p. ~r f: w on stem samples collected the morning spraying. A 6-inch long stem sample was beginning 6 inches above the ground from ' ffive plants. The bark was peeled off. The "-0 were dried at 158° F and subsequently pass through a 60-mesh screen. The samples f} run at each date. cation of methods described by Hall and, (9) and Wildman and Hansen (17). Dupli- olyzed at 131° F for 2 hours with Taka- The mixture was cleared with lead sub- l hich was subsequently precipitated with “xalate and then neutralized. Carbohydrates 'dized- with the concomitant reduction of lfate to copper oxide which was titrated with m permanganate after the addition of ferric ‘Mm sulfate. g > ‘lture stress in the leaves was determined with der pressure apparatus (8). Ten trees were dates. Moisture stress was determined in from each of five plants in each treatment, (4-5 a.m.), 8-10 a.m., 12-2 p.m., 3-5 p.m. and L Central Standard Time. Pressure readings percent total available carbohydrates was. i?‘ ed, and at least one air of du licate sam- i‘ .. P P available carbohydrates were analyzed by‘ samples were refluxed in distilled water - J l?» each treatment at each date, except the" were made 1 to 7 days after spraying. Leaves were removed from half the trees at each time period to minimize the effect of leaf removal on plant response. Data presented show the predawn and daily maximum for each spraying date, the daily stress cycles at several dates, the seasonal stress cycles at several times of day and the cycle of stress in leaves of plants treated with herbicides in the 4-day period following spraying. Environmental Variables Environmental variables measured included max- imum and minimum air temperatures; maximum and minimum percent relative humidities; maximum soil temperatures at l-. and 3-foot depths; percent soil moisture at depths of 0-1, l-2 and 2-5 feet; rainfall; and total solar radiation. All except solar radiation were recorded as the mean for the 7-day period prior to spraying. Maximum and minimum air temperatures and percent relative humidity were recorded at the site with a hygrothermograph. Soil temperature was recorded at the site with a recording thennograph having probes set at 1- and 3-foot depths. The mean daily maximum and minimum mil temperatures were computed. Soil moisture was determined weekly using gravi- metric analysis. Samples were taken from soil depths of 0-1, l—2 and 2—3 feet either with a bucket auger or a screw-type auger (7). Most samples were collected on Mondays. Five cores were dug at each date. Rain- fall data were collected weekly at the site with a rain gauge. Total solar radiation on the day of spraying was extrapolated from weather records taken by the Department of Meteorology at Texas A8cM University at College Station, Texas, about 5 miles from the experimental site. The data are presented as Langleys. Statistical Analyses Data on percent defoliation and percent dead plants were analyzed as a split plot analysis with sub- plots as a factorial of three chemicals at two rates. Four replications were used. Untreated plots also were included at each date but not analyzed with the other treatments. " An attempt was made to develop a predictive indicator for correlating plant response to herbicides at any given date with plant characters and environ- mental factors. Two multiple regression models were used to predict percent defoliation and percent dead plants. The independent plant variables included new stem length, translocating phloem thickness, total and rate of new xylem ring development, rate of upward dye movement, total available carbohydrates and predawn and maximum leaf moisture stress. Environmental variables during the week prior to spraying included maximum and minimum air tem- peratures and percent relative humidity; soil tempera- tures at l and 3 feet; soil moisture at depths of 0-1, 7 1-2 and 2-3 feet; and rainfall. Also, total radiation the day of spraying was included. Several data analyses were made with different combinations to derive the best equations since all variables could not be run in a single analysis. All regression equations pre- sented had F values significant at the l-percent level with independent variables significant at least at the 5-percent level. The first regression developed the best equation for the combination of all treatments over all l4 dates. All treatments were fixed into the model which required the line to be drawn through the origin. Time and time? factors were fixed into the equation to fit the parabolic curve of the data with time; time counting was begun January l. Ten independent variables were placed in a pool. The best eight equa- tions for one to nine independent variables were calculated. The second regression model was used to develop the best equation for each treatment individually over all l4 dates. The time and time? variables were placed in a pool with l0 other independent variables. The eight best equations were then calculated having one to six independent variables. Simple correlations of all dependent and inde- pendent variables were calculated for the best ll dates. March 24, April l4 and October 9, 1969, we omitted. Selected correlations are presented to sho ' the interrelationships of these variables. Results Plant Response to Herbicides The day following spraying, the elongating an tips curled in all treatments. '~ By the second da" different plant responses among treatments were ii servable. Leaves of plants sprayed with picloram i; picloram + 2,4,5-T had darkened water-soaked spo by the end of the second day. Within 4 days m leaves remained intact, turned brown and died, l‘: by 7 days all leaves were dead. Leaves of pla F“ sprayed with 2,4,5-T turned yellowish-green and i came detached from the plants by the end of I third day. Defoliation most frequently began wi the leaflets detaching from the rachises. By the - ‘ of 7 days the remaining dead portions of the leav either remained on the plant or became detach Death of the foliage generally occurred a day orlt i. sooner when the plant was sprayed in late su u~ rather than in April. Differences in percent defoliation occurred amo f. treatments averaged for all l4 dates (Table 1). Pic ram and a mixture of picloram + 2,4,5-T were equa TABLE l. PERCENT DEFOLIATION AND PERCENT DEAD HONEY MESQUlTE PLANTS AT A FIELD SITE NEAR BRYAN, TEXAS, SPRAYED IN AND l97O AND RATED ABOUT l YEAR AFTER TREATMENT Year sprayed l969 l97O ,, Rate 3/24 4/l4 4/30 5/20 6/9 7/l 8/25 l0/9 5/4 5/20 6/l9 7/6 7/30 8/l0 lb./acre — — — * — — — — — — -— —- Defoliation, % — — — — — — — — — — —— -— _‘ Picloram 0.5 2 l4 82 78 92 74 32 4 65 76 62 66 48 4l 5 Picloram l l l8 96 86 96 88 34 2 84 94 85 77 76 45 6 Picloram salt —|- 0.25+ 2,4,5-2 ester 0 25 0 l2 86 73 69 86 27 4 62 79 88 75 48 47 Picloram salt + 0 5 + ‘ 2,4,5-T ester 0.5 l l6 80 77 9l 96 44 4 62 86 96 94 60 64 6 2,4,5-T 0.5 0 8 7l 62 46 72 20 2 49 72 66 63 48 32 2,4,5-T l 0 25 7l 50 65 78 24 2 56 70 6l 6l 59 5l 4 Untreated l2 l2 l2 l3 l3 l4 l8 l6 32 2i l0 0 ll 0 j Mean‘ 2" l5‘ 71‘ as“ s7" 73‘ 2a’ s" s9" 71' 67"’ s2“ so‘ 40' ‘f Picloram + 2,4,5-T 0.25+‘ ‘ amine salts 0.25 68 74 8O l7 86 78 Picloram -l- 2,4,5-T 0.5 +2 amine salts 0.5 75 88 84 l3 9l 93 : _ _ _ _ — — — — — — -——-Dead plants-—-—---—--—--—---—---- Picloram 0.5 0 0 . 30 l8 60 5 5 0 5 l0 4 l7 5 0 l; Picloram l 0 0 75 22 75 35 0 0 l5 55 46 33 20 0 i‘ Picloram salt + 0 25+ ¢' 2,4,5-T ester 0.25 0 0 30 l5 l5 25 O 0 5 20 29 l2 l0 8 l Picloram salt + 0.5 + 2,4,5-T ester 0.5 0 0 35 l5 4O 65 5 0 5 30 66 54 0 8 2,4,5-T 0 5 0 0 5 0 0 l2 0 0 0 0 0 4 0 0 2,4,5-T l 0 0 0 0 l0 l5 0 0 5 0 0 O 0 0 Untreated 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Mean‘ o‘ o’ 2s" to“ 29' 22"“ i‘ o’ s" l6“ 21"" 17° s" 2' Picloram + 2,4,5-T 0.25+’ amine salts 0.25 l0 3O l0 O 3O l7 Picloram —l— 2,4,5-T 0.5 —l-’ amine salts 0.5 2O 50 35 O 45 58 ‘Values on the same line or column followed by the same letter are not significantly different at the 5% level using Duncan's Multiple Raf, Test. ‘Values for picloram + 2,4,5-T as the triethylamine salts were not subiected to statistical analysis. ;more than 50- and 60-percent defoli- 'j.-and l-pound per acre rates, respec- ' "T treatments were less effective than picloram + 2,4,5-T. aried markedly among the l4 dates. u 'cals show that the most effective s» from April 30 to July 1 in 1969 l to July 6 in 1970. At some dates, imixture of picloram + 2,4,5-T treat- " than 90-percent defoliation. No ces occurred between the mixture potassium salt of picloram and the _ and the mixture containing triethyl- cloram and 2,4,5-T. percent dead plants were lower than liation but showed the same trends ged over l4 dates, picloram and a oram + 2,4,5-T were about equally about ll and 25 percent of the plants, the 0.5- and 1-pound per acre rates, treatments killed only 2 percent of (‘some dates, 1 pound per acre of piclo- ture of picloram + 2,4,5-T killed as 66 percent of the plants, respectively, est percentage killed b-y 2,4,5-T was tics were initiated about March l5 to elongated rapidly until about April Then the stem tips died and aborted. f‘ made into August showed no further _ April 30. These and older stems T1 few inflorescences and no legumes tions measured at each spraying date illimeters in diameter. Annual radial largely in the phloem and xylem. translocating phloem and total and 2-week growth (rate of growth) of the new xylem ring are presented in Table 2. The translocating phloem was thickest between mid-May and. mid-June. The new translocating phloem was initiated in late March. After the rapid growth period, some of the trans- locating phloem sieve elements were crushed against the phloem fibers. This crushing action reduced the translocating phloem thickness by July. The new xylem ring began radial enlargement the early part of April, and growth was minimal after July (Table 2). The other tissue thicknesses varied somewhat in size; however, they fell in the following ranges: peri- derm—0.l3 to 0.20 millimeter; nontranslocating phloem—0.26 to 0.72 millimeter; and means of xylem rings produced in 1969, 1968, 1967 and 1966-115, 2.26, 1.86 and 1.53 millimeters, respectively. Pith diameter was 0.82 to 1.12 millimeters. Upward movement of moisture and dye occurred almost entirely in the outermost xylem ring. Although the rate varied widely, a trend showed that the most rapid movement occurred between April 30 and June 9 (Table 2). Warm mornings with full sunlight and adequate soil moisture seemed to promote the most rapid dye translocation. Changes in total available carbohydrates were small (Table 2). However, the lowest level for the year occurred largely before the best time for spraying in April, May and June. Moisture stress levels in leaves have a daily cycle. Stress ratings of untreated plant leaves at several dates in 1969 are shown in Figure 1. Moisture stress was at a minimum during the night when moisture uptake surpassed transpiration and the plant cells became turgid. Moisture stress levels in the leaves increased to a high of about 30 atmospheres in the afternoon and subsequently receded toward evening. The April 30 stress levels were lower than those later in the growing season. Differences in moisture stress among > CHARACTERISTICS AT 14 DATES OF SPRAYING HONEY MESQUITE AT BRYAN, TEXAS Xylem ring thickness R919 °f T9831 Leaf moisture stress Translocating upward available I New stem phloem Total for Rate of dye stem car- Midday length thickness year growth movement bohydrates Predawn maximum cm mm mm mm/2 wk cm/hr % atm atm 1969 1 0.08 0 00 0.00 8 9.0 6.2 6.2 27 .11 11 .07 85 2.0 6.2 6.2 31 .13 23 .10 114 4.6 7.1 13.4 31 .18 76 .36 110 6.0 6.4 25.1 31 .18 1 21 .32 101 5.5 8.0 27.6 31 .08 1 59 .24 59 6.9 9.7 29.7 33 .07 1 77 .04 78 6.5 11.2 26.1 29 .06 1 72 .00 78 7.4 15.8 28.7 1970 28 .16 23 .14 102 4.0 7.1 20.7 28 .15 46 .20 153 4.0 7.2 29.1 28 .18 1.09 .29 60 5.0 8.4 25.9 27 .13 1.30 .17 28 6.1 10.1 29.8 27 .07 1.24 .00 28 6.1 11.7 25.8 27 .03 1.31 .01 18 6.0 13.8 29.5 TIME OF DAY dates from June 9 through October 9, 1969, were small except for the daily extremes. -""'"'"'7-_ 25- / °____ (D E m 20- I O. 3 E l5‘ '4 lo‘- SEPT 2 »—-@ JULY l A"""'A 9 JUNE 9 ‘l’ '" ‘l’ 5- MAY 2| <>—<> APRlL 3° "_" T2321’? r'.'r..fi“v°li'°.i? h‘ mesquite plants 0t five l‘ irel°llrli§llsmri iigloi“. r ' r | | 1 ore means for l to 4 -~i 4T0 e e TO I0 NOON TO a TO 5 e "r0 a beg""""go""*edov AM AM Z PM PM PM ' treated plants had much lower moisture stress untreated plants. At night, leaves of no treatmen were under moisture stress. The seasonal stress levels of untreated leaves at Mvarious times of the day in 1969 and 1970 are pre- sented in Figure 2. The predawn stress level increased during the year and probably reflected the progressive decrease in soil moisture through the season. The readings made at 8-10 a.m., noon-Z p.m. and 3-5 p.m. were similar and are averaged in Figure 2. They were all much higher than the predawn reading. Moisture stress readings made at 6-8 p.m. were intermediate between the predawn and the other readings; the readings at 6-8 p.m. were low on April 30, reached a maximum in June and gradually de- creased through the remainder of the growing season. The decreasing values may have resulted from the shortening of the days in the fall. Predawn, the minimum, and the midday maximum stress level data from untreated plant leaves used for regression analysis are presented in Table 2. 30- ATMOSPHERES Moisture stress levels in leaves of untreated plants and those sprayed with 0.5 pound per acre of picloram //\"N"x\,rx//x . j/ “\A\A mA \* AJ / U J\. / /\/ .»°—' 8AM TO 5PM x-x 6PM TO 8PM A—A 4 AM .-ro 6AM -—-- or 2,4,5-T were studied in the 4-day period following spraying on June 9, 1969. At dawn all leaves had the same low value (Figure 3). Maximum stress values occurred at noon. However, by 6 p.m. the IO APRIUMAY mum; mum mus 'SEPT'OCT' MONTH Figure 2. Seasonal moisture stress levels in leaves of honey quite at three intervals during the day in I969 and 1970. A __1 o UNTREATED °-—' Figure 3. Moisture stress levels in leaves of honey e second and third day, however, the l’ nts treated with herbicides failed to _ tressed during the midday as the un- A The difference between herbicide v -~ probably not significant at this stage. i fourth day, the treatments differed untreated leaves had about the usual i‘ cycle. Almost all leaves sprayed with wever, had died and had begun drying ntially infinite values recorded on the aratus were indicative of the dried state at 4 days. _ A of leaves sprayed with 2,4,5-T began to e third day after spraying. By the fourth ; only petioles and rachises remained. The e1 in the petioles was regained by the '}'ng. The petioles yellowed and abscised th through the seventh day. The leaves _' the mixture of picloram + 2,4,5-T re- tially like those sprayed with picloram hi‘ levels in untreated plant leaves and ed with the l-pound per acre rate of either if: 2,4,5-T are compared in Figure 4. At ileaves in the plants treated with herbicides me as stressed as the untreated leaves the 2,4,5-T X——X Tzsquite plczlnts fllulruingh 2F. olio bee?’ eilflliatraln? unlsllbrltilyed cbr sprayed on June 9, ‘I969, i ' ~ ' ‘ * 1Z1‘: s‘? gfls-p-‘fllil ‘iflofiffi’ DAYS AFTER SPRAYING first 3 days after spraying. The depression in stress at this rate occurred earlier than at the 0.5-pound per acre rate. The leaves were not stressed on the night they were sprayed nor on the following two nights. As in the 0.5-pound per acre rate, the ratings in the sprayed leaves were abnormally high the fourth day. The leaves sprayed with picloram remained intact and turned brown when they died. The leaf- lets of the leaves sprayed with 2,4,5-T abscised. The remaining petioles and rachises had yellowed some- what and either died on the plant or abscised by the seventh day. The leaves sprayed with the 0.5 + 0.5-' pound per acre rate of picloram + 2,4,5-T reacted essentially the same as those sprayed with picloram alone. Pressure readings of treated plants taken at other times of the year gave similar results to those of June 9, 1969. However, the leaves seemed to die about a day earlier as the season progressed into periods of lower soil moisture. Environmental Variables The mean maximum and minimum air tempera- tures for the 1-week period prior to spraying are presented in Table 3. The range of temperatures during spraying dates of April 30 to July 1, 1969, and ATMOSPHERES 2,4,5-T PICLORAM O UNTREATED 9-0 Figure 4. Moisture st levels in leaves of ho ‘ mesquite plants during .\ 4-day period which ' been either unsprayed i sprayed on June 9, l9 __ with a l-pound per xl-X 0-—-O Q | | I l AM PM AM PM I 2 AM DAYS AFTER SPRAYING May 20 to July 6, 1970, was wide. Maximum tem- perature during these periods ranged from 77° to 96° F, and the minimum temperature ranged from 58° to 80° F. As expected, the temperatures in- Mcreased progressively from March until reaching a maximum in July. Temperatures remained high into September but were lower in October. TABLE 3. NEAR BRYAN, TEXAS I 3 AVERAGE ENVIRONMENTAL CONDITIONS ON A CLAY LOAM SOIL DURING THE WEEK PRIOR TO SPRAYING HONEY MESQU AM4 rate of 2,4,5-T or piclorcl PM PM Percent relative humidity varied widely duli-i the day, frequently as much as 50 percent. Diff ences among dates, however, were small (data presented). Maximum percent relative humidi: generally reached 9O percent or more at night. t‘ minimum levels of 30 to 50 percent generally occu A in the afternoon. “ I Air temperature Maximum soil temperature Soil moisture 2-3 } Date sprayed Maximum Minimum l ft 3 ft 0-1 ft. 1-2 ft _ _ _ _ _ —-—°F——————-- -----___-%-_____> l9s9 _. March 24 s9 sl ss s2 24 29 2s April l4 7s s2 7o 5s 24 27 2s u April so 77 ss 72 ss so 27 2s ‘ May 2o 79 ss 7s s4 2s 2s 24 1 June 9 8O 6i 84 74 21 24 ' 22 1 July l 9s 7s 9o 79 l4 l7 ls August 2s ss 7l ss ss 9 l2 ls October 9 s7 s5 ss 7s 2o 22 2l ' 1970 1 May 4 79 sl 7s ss ls 2s 24 g May 2o s4 ss 7s ss l2 2l 2s June l9 9l 74 s2 72 ls 22 22 < July s 9s so ss 77 s ls l9 » July so 92 s9 ss so s ls ls ” August lo 97 s9 9s ss s ls ‘l2 MLE 4. SIMPLE CORRELATION COEFFICIENTS BETWEEN EIGHT DESIGNATED VARIABLES AND PERCENT DEFOLlATlON OF HONEY MESQUlTE KANTS CAUSED BY EACH OF SIX TREATMENTS AT 11 DATES* Picloram Picloram + 2,4,5-T 2,4,5-T 0.5 l 0.25 —l— 0.25 0.5 + 0.5 0.5 l 1.. Translocating phloem thickness, p. 0.65 0.67 0.54 0.46 0.39 0.20 Rate of xylem radial growtht‘ .66 .76 .60 .27 .25 .20 Predawn leaf moisture stress, atm — .70 .73 —— .59 — .36 ——- .49 -— .28 Maximum soil temperature, °F, 1 ftIlI — .49 .50 —— .37 — .10 — .35 — .06 Maximum soil temperature, °F, 3 ftIlI — .59 —. 63 — .50 — .23 —— .44 — .20 Soil moisture, %, 0-1 ffl‘ .64 .58 .47 .25 .34 .20 Soil moisture, %, 1-2 ftI .73 .73 .61 .40 .45 .31 Soil moisture, %, 2-3 ftili .70 .76 .62 .40 .52 .34 spectively. an values during the 1-week period before spraying. Soil temperature at both 1- and 3-foot depths i generally increased from the earliest date recorded to early July when the best spraying period ended “in 1969 (Table 3). In 1970, however, the soil tem- perature increased continuously from May 4 to August Consequently, soil temperature in 1970 was not well correlated with response to herbicides. The maximum soil temperatures the day of spraying at ;_ the onset of the best spraying period were 72° and i1i76° F at 1 foot and 63° and 66° F at 3 feet in 1969 ‘sland in 1970, respectively. The cessation of good honey mesquite response to herbicides occurred at maximum soil temperatures of 90° and 86° F at 1 foot i. and 79° and 77° F at 3 feet in 1969 and 1970, respec- tively. Daily minimum soil temperatures (data not maximum. Soil moisture 1 week before spraying varied from 6 to 30 percent at the 0-l-foot depth, 12 to 29 per- scent at the 1-2-foot depth and 14 to 26 percent at _ the 2-3-foot depth (Table 3). The moisture level was {Li-usually highest in March and April and lowest in (Zas-"July, August and September. The best spraying period ates AT n DATES* ‘March 24, April 14 and October 9, 1969, dates were not included. iladial enlargement occurring in the 2-week period before spraying. ggpresented) were normally 4° to 6° F lower than the Correlation of i 0.60 and 0.74 are significant at 5 and 1%, re- generally occurred during an overall period of de- creasing soil moisture. In 1969, rainfall was fairly well distributed except for the July 1 spraying, where little rainfall occurred. Most rain fell in early April. In 1970, abundant rain- fall occurred until May, but little occurred thereafter. Total daily solar radiation was markedly affected by the amount of cloud cover and varied from 235 to 656 Langleys. The radiation varied widely during the year; the only consistently low values were re- corded in September and October. Statistical Analyses The simple correlations along with the sign for percent defoliation data for all herbicide treatments with the eight most important independent variables are presented in Table 4. The 0.5- and 1-pound per acre picloram treatments were directly correlated with translocating phloem thickness, rate of xylem radial growth and percent soil moisture at l—2 and 2-3 feet depths. They were inversely correlated with predawn moisture stress. Most temperature variables were not significantly correlated with plant control because ITTABLE 5. SIMPLE CORRELATIONS OF SELECTED HONEY MESQUlTE PLANT VARIABLES WITH 11 OTHER PLANT AND ENVIRONMENTAL VARIA- _ New xylem T"°"5l°¢°'""9 ring thickness Moisture stress F phloem thickness Total Rate Predawn Maximum q Variable ; p. mm mm / 2wk Atm Atm ; 1. “Translocating phloem thickness, p. —0.54 0.80 —0.86 —O.37 2. Total new xylem ring thickness —-0.54 .74 ~ .87 s 3. Predawn leaf moisture stress, atm — .86 .74 -- .65 .68 j 4. Maximum leaf moisture stress, atm — .37 .87 .69 .68 1" 5. Maximum air temperature? — .68 .73 —— .35 .84 .75 a 6. Minimum air temperqturel‘ — .41 .69 — .15 .57 .67 7. Maximum soil temperpture, 1 ft, °F'l' — .71 .75 — .50 .86 .65 8. Maximum soil temperature, 3 ft, ° F1‘ -— .76 .83 — .55 .91 .69 9. Soil moisture, 0-1 ft, %‘l‘ .62 .57 .55 — .82 -— .53 I O. Soil moisture, l—2 ft, "/°'l' .86 — .78 .65 — .93 — .59 - l. SOll moisture, 2-3 ft, %'l‘ .86 — .84 .61 -—- .97 — .71 respectively. iafMean values during the 1-week period before spraying. ‘March 24, April 14 and October 9, 1969, dates were not included. Correlations of i" 0.60 or i 0.74 were significant at 5 and 1%, ‘l3 TABLE 6. SIMPLE CORRELATIONS OF SELECTED ENVIRONMENTAL VARIABLES DURING THE 1-WEEK PERIOD BEFORE SPRAYING AT 11 Maximum soil _ _ _ temperature Soil moisture Minimum air ; Variable temperature 1 ft 3 ft O-1 ft 1-2 ft 2-3 f '—'----—'°F———- — — — — ——%-——-—-—-, 1. Maximum air temperature, ° F 0.87 0.82 0.81 —0.80 -— .82 — .8 2. Minimum air temperature, ° F .70 — .72 —_= .65 — 1f 3. Maximum soil temperature, 1 ft, ° F .97 — .82 '-'~f .83 -— .8 4. Maximum soil temperature, 3 ft, ° F -— .83 — .91 — .9‘ 5. Soil moisture, 7., 0-1 ft .86 .7 '_ 6. Soil moisture, 7., 1-2 ft .9 ‘March 24, April i4 and October 9, 1969, dates were not included. respectively. temperature was directly correlated in May and June and inversely correlated thereafter. Of the mixtures, only the 0.25 + 0.25-pound per acre rate of picloram + 2,4,5-T defoliation data were significantly correlated with the rate of xylem growth and soil moisture at l-2 feet and at 2-3 feet. Other chemical treatments were not significantly correlated with the factors measured. None of the variables were significantly related to the percentage of plants killed in any of the six treatments (data not presented). The simple correlations of five plant characters with ll plant and environmental variables are shown Correlations of i 0.60 or i 0.74 are significant at 5 and '* in Table 5. Translocating phloem thickness _ inversely correlated with predawn leaf moisture st s», and maximum air and soil temperatures and directl correlated with soil moisture. ‘ Total new xylem thickness was directly correlati with predawn and maximum leaf moisture st a7 maximum and minimum air temperatures and u. j imum soil temperature at both depths; it was invers correlated with soil moisture at the two greater dep 1 f Rate of new xylem radial ring growth was direc correlated with translocating phloem thickness, 11. g imum leaf moisture stress and soil moisture at 1! two greater depths; it was inversely correlated wi predawn leaf moisture stress. TABLE 7. REGRESSION EQUATIONS SHOWING THE RESPONSE OF HONEY MESQUITE NEAR BRYAN, TEXAS, TO HERBICIDE SPRAYS APPLIED 14 DATES IN 1969 AND 1970 Standard Chemical Rate Equation‘ error lb./acre % 1. All chemicals, both rates (percent defoliation) A 9 = 4.2air)-o.oo97s(r’1+4.a2(x11 15.2 Picloram 0.5 —-465 P‘; Picloram 1 —454 ‘ Picloram + 0.25 + 2,4,5-T 0.25 —464 Picloram -)- 0.5 + 2,4,5-T 0.5 —455 2,4,5-T 9.5 —474 2,4,5-T l —470 2, Pidoi-qm 0,5 (percent defoliation) ‘y 9 = 27.s+1so.six2i 20.7 o. ' 9 =_4as.o+4.2o(r1-o.oo9¢4it'1+5.31(x11 14.3 .7 (percent dead plants) I 9 = -9.a7+1.a1ixs1 18.8 .1) IY‘ = —1233+1.64(T)—-0.00395(T2)-—-2.4B(X4)+2.31(X5)+'I.6Q(X6) 15.7 .4“ 3. Picloram 1 (percent defoliation) If ' 9 = ss.s+19s.o(x21 25.7 9 = —502.8+4.74(T)—0.0i1ti(T')+4.48iX1) 15.7 9 = —586.i+3.79(T)—0.0093i(T’)+2.i0(X5)+6.7i(Xi) 14.4 .a, a (percent dead plants) p 9 = -12.7+s.4s(x31 27.0 .2 ,Y\ = —545.6-|—l.63(T)——0.044l9(T:)—)-3.l9(X5)-l—9.56(Xl) 21.7 .50; 14 (courmuso) Standard , Rate Equation‘ error R’ lb./acre ‘)4 + 0.25 + (percent defoliation) 0.25 '9 = 2a.a+1e1.o1x2) 22.9 .49 '9 = -4as.1+4.34m-o.o1oo211’)+4.771x1) 14.7 .80 (percent dead plants) v '9 = 4.e7+52.41x21 14.3 .17 '9 = -133.o+1.a21x41+1.o61x61-3.o31x71 13.1 .32 -|- 0.5 + (percent defoliation) 0.5 '9 = 35.4+192.3lX2l 24.2 .50 '9 = s33.3+4.97(r1-o.o114o11'1+4.321x1) l 11.6~ .39 (percent dead plants) '9 = 4.77+132.71x2) 24.1 .32 '9 = -94.o+1.161x41+132.s1x2i 22.3 .43 0.5 (Percent defoliation) '9 = 24.2+141.21x2) 23.0 .37 '9 = -444.s+3.ao1r1-o.oos671r’1+4.741x11 16.2 .70 9 = -74o.2+s.9s1r1-o.o13s2(r*)-4.o4(x3)+1o.s1x11 14.2 .77 (percent dead plants) '9 = s.37-o.67(x31+27.6(x21 6.2 .09 '9 --— -ss.3+o.s1(x41+o.7s1xa1-1.os1x71 6.0 .17 I (percent defoliation) '9 == 31.a+117.61x2) 23.3 .23 '9 = -33o.9+a.34(n-o.oo7as1r’)+2.2s1x1) 15.9 .68 '9 = -601.a+4.42111+o.o1oss11")-3.23(x31+1.241xs1+7.931x11 14.0 .76 (percent dead plants) '9 = -o.o633+1s.91x21 7,1 ,0; . '9 = 4.9o-o.72(x3)+39.41x2) 6,9 _'|5 , 2X's = .42I.57, 3X's = .53/.66, 4X's = .62/.74. wn leaf moisture stress was inversely cor- éiwith translocating phloem thickness, rate of (,3 growth and soil moisture, and it was di- n-related with total new xylem ring thickness, l 11 leaf moisture stress and air and soil tem- . Maximum leaf moisture stress was corre- ‘v9.11 ‘larly to predawn leaf moisture stress. ple correlations of seven environmental vari- J a shown in Table 6. Maximum and minimum fperatures are q directly correlated with each ild. with soil temperature at both depths, and l‘ inversely correlated with soil moisture. Maxi- temperature at depths of 1 and 3 feet are 1 correlated with each other and inversely cor- (with soil moisture. Soil moisture at the three of are directly correlated. abbreviations are the following: T = time in days from January ‘l to spraying date; T’ = time period (T) squared; Xi = per- i’ fmoisture at the depth of 2-3 feet l week before spraying; X2 = rate of new xylem ring radial growth in the 2-week period before l X3 = translocating phloem thickness; X4 = average maximum air temperature during l week before spraying; X5 = average '9 soil temperature at a depth of 3 feet during i week before spraying; X6 = percent soil moisture at the depth of O-l foot l week ' ying; X7 = predawn leaf moisture stress; and X8 = percent available carbohydrates. (1)4 dates were analyzed. Df = i4—(l for Y +1 for each X variable). Significant R’ values are the following for 5% Ii %: IX ppropriate negative value of the specific herbicide treatment to the above equation. The other variables measured (data not presented) were not significantly correlated either with plant con- trol or other independent plant or environmental variables. Regression equations were developed from the data to compute the best estimated value (l?) for the observed percent defoliation and percent dead plants at any period during the growing season. Hopefully, these equations will closely fit data from other similar treatments showing seasonal honey mesquite responses to picloram and 2,4,5-T throughout the growing sea- son. If reliable equations can be developed, they can be used commercially to predict the amount of honey mesquite control that can be expected from the treat- ment under any given seasonal and plant growth conditions. iNew stem length In the first regression analysis, an equation was developed to compare percent defoliation by all treat- ments at the same time (Table 7). Percent soil mois- ture 2-3 feet deep was the most important inde- pendent variable. The predicted defoliation (Y) of any treatment can be calculated by adding the appro- priate negative treatment constant to the products of the coefficients and their appropriate measurements. The standard error was 15.2 and was not improved appreciably by adding more variables. This equation was approximately as good as any individual treat- ment equation. The equation for percent dead plants was appreciably less useful because of the large differ- ences among treatments and is therefore not presented. Regressions for percent defoliation and percent dead plants for each treatment are presented in Table 7. Defoliation caused by both rates of picloram could best be estimated with rate of new xylem ring growth alone. Soil moisture at 2-3 feet was appreciably better upon adding the time-time? factor. Percent dead plants was best estimated by translocating phloem thickness as a single variable. The precision was some- what improved by the four- or five-variable equations including the time variables, maximum air tempera- ture (for 0.5 pound per acre only), maximum soil temperature 3 feet deep and percent soil moisture either at depths of 0-1 or 2-3 feet. On the basis of a single variable, percent defolia- tion and percent dead plants at both levels of pic- TABLE 8. A SUMMARY OF HONEY MESQUlTE AND ENVIRONMENTAL FACTOR RANGES DURING THE BEST PERIODS FOR SPRAYING IN 1969 AND 1970 NEAR BRYAN, TEXAS Factor measured Range 27-31 cm Stem tissue thickness Translocating phloem 0.08-0.18 mm Total xylem ring 0-23-1-59 mm Rate of radial xylem growth 0.10-0.36 mm/ 2 weeks Upward xylem methylene dye movement 28-153 cm/hour Total available carbohydrates 4.0-6.9% Leaf moisture stress in untreated leaves the day after spraying Predawn 6.4-l0.1 atm Maximum 13.4-29.8 Maximum soil temperature, 1 week before spraying 1 foot deep 72°-90° F 3 foot deep 62°-82° F Soil moisture, 1 week before spraying O-1 foot 8-301» 1-2 feet 17-27% 2-3 feet 18-25% Air temperature, 1 week before spraying Maximum 77°-96° F Minimum 58°-80° F Relative humidity, 1 week before spraying Maximum 70-1001, Minimum 36-50% Rainfall, 1 week before spraying 0-3.49 inches Total solar radiation 238-528 Langleys 16 loram + 2,4,5-T were best estimated using the y of new xylem ring growth. For percent defoliatii the best equation was obtained using the time-ti i.» factor with percent soil moisture at a depth of 2-3 f g Percent dead plants was best estimated at the 0.25 l, 0.25-pound per acre rate with maximum air tem ' ture, soil moisture at a depth of 0-l foot and preda leaf moisture stress. Percent dead ‘plants at the 0.5 . 0.5-pound per acre rate was best “estimated by m ; mum air temperature and rate of new xylem _l growth. i On the basis of a single variable, rate of xylem ring growth gave the best estimate of u percent defoliation and percent dead plants by ei i‘ rate of 2,4,5-T. Percent defoliation with the 0i pound per acre rate was enhanced using the time-ti i’ factor plus translocating phloem thickness and v cent soil moisture at the depth of 2-3 feet. Perc defoliation at the l-pound per acre rate was enhan using the same four variables plus maximum 5' temperature at the 3-foot depth. Only a few plants were killed by either rate 2,4,5-T. Consequently, none of the variables incl - ing the time-time? factor contributed appreciably l the equation. Maximum air temperature, per total available carbohydrates and predawn leaf ture stress added slightly to the equation for the pound per acre rate of 2,4,5-T, and transloca phloem thickness and rate of new xylem growth added to the l-pound per acre rate. Representative variables during the most ef y tive spraying period from April 30 to July 6 are s l’ marized in Table 8. Discussion q Little difference occurred in control of l." mesquite among picloram and picloram + 2,4, treatments (Table 1). The mixture is preferred ; cause 2,4,5-T is less persistent, and the mixture ‘ presently less costly than picloram. The formulati of the mixture was not important. Thus, either a a mix or the formulated herbicides can be used equa well. Both picloram and picloram + 2,4,5-T superior to 2,4,5-T alone. The latter is presently ommended for the control of honey mesquite. 2,4,5-T treatments killed very few plants since th plants had regrown after mowing and possessed l leaf area to intercept and translocate herbicide to A lower stem and roots than comparable undistur plants. Also, chemical control of plants growing q heavy soils has generally been poorer than for pl i! growing on lighter soils, even in West Texas (2). f In almost every case, the l-pound per acre A l ment was more effective than the 0.5-pound per ‘ treatment. Other unpublished results from ex ments near Bryan, Texas, have indicated increasif effectiveness with rates up to 2 pounds per acre {A 2,4,5-T and / or picloram. ' i-r differences in plant response occurred Ies of treatment using the same chemical (Table 1). This finding shows the impor- pplying the spray at the date which will give tive control. Generally the best time to lrcides is between April 30 and July 6, which W- roughly to the recommended time period days after bud break. In years with normal {best results will be obtained by spraying dur- However, the amount of control ob- j. year to year varies markedly. Therefore, freliable measurement is needed to quantify amount of control that will be obtained ' g at any given date. f“ elongation growth did not occur in the presented (Table 7). The end of the stem K period essentially marks the beginning of I time to spray (Table 2). Thus, recognition ition of stem elongation growth at spraying ‘uld be useful to show the onset of the spray- f. but would be of little value either during ying season or to mark the end of the spray- n. A more meaningful measurement might elongation growth per unit time during a Zsuch as 2 or 3 weeks prior to spraying. slocating phloem thickness (Table 2) occur- Qfour regression equations (Table 7). Phloem If was directly correlated with percent defolia- able 4). The range in thickness during the ying season was 0.08 mm to a maximum of kt size stems (14) and thus would need to be A f! ized to be a useful measurement. However, l translocating phloem is the tissue that con- ' herbicide to the lower stem and roots, the l’ of tissue should have a significant effect on esquite response to herbicides. _-. - new xylem ring (Table 2) is produced dur- f‘ best spraying season. The total thickness is Qcorrelated during the period of best plant con- to 1.59 mm) but is little correlated at the e summer when plant control decreases. The new xylem ring radial growth (Table 2) fol- (‘period of best control more closely. With- f- use of a recording dendrograph, measurements f ed at a minimum of 2 weeks. Rate of new g radial growth occurred as the most impor- _( 'able where the time-time” factor was not _te of upward methylene dye movement in the (Table 2) failed to occur in any useful regres- uation (Table '7»).".The rate varied from 28 to timeters (cm) per hour during the best spray- "od. Two factors handicap the importance of '. - urement. First, only l0 trees were infused f. date. Consequently, the variation from tree reduced the validity of the measurement. Sec- i’ was difficult to infuse trees under exactly the y. This measurement is small and varies on s same environmental conditions each time. Changes in day length during the season and cloud cover can affect the radiation and resulting transpiration at time of infusion. Further refinement of the injection pro- cedure may increase the value of this measurement. Total available carbohydrates varied from 4.0 to 6.9 percent during the best spraying season (Table 2). This variable failed to be included in any equations (Table 7). It generally was lowest just prior to the best spraying season and then increased until about the middle of the spraying season. Determination of total available carbohydrates is expensive and difficult to reproduce. Also, the entire sapwood was sampled. Perhaps only the outermost one or two growth rings would have been a better measure because they are more responsive to seasonal changes in carbohydrates (14). The leaf moisture stress readings (Table 2 and Figures 1 and 2) were similar to those of Haas and Dodd (8). The values were low in the morning and evening and high during the day. There was a gen- eral increase in the predawn value of 6.4 to 10.1 atmospheres during the best spraying period, and a maximum of 13.4 to 29.8 atmospheres at midday. The predawn rating occurred in one regression equa- tion. It might be a useful measurement if it can be correlated with control more closely during the late summer period. The maximum leaf stress value was less well correlated. It rose essentially to a high level early in the season and changed very little thereafter (Table 2 and Figure l). The herbicides reduced the moisture stress in honey mesquite plant leaves for several days after treatment (Figures 3 and 4). This reduction in stress may be due to closing of stomata or other changes in the plant's metabolism. The importance of this reduction in leaf moisture stress by herbicides needs to be studied more fully. Maximum and minimum air temperatures dur- ing the l-week period before spraying were 77° and 58° F, respectively, at the beginning of the best part of the spraying period and increased about 20° F by July 6 (Table 3). air temperature appeared in one regression equation (Table 7 ). Air temperature was directly correlated with total xylem thickness, leaf moisture stress (Table 5) and soil temperature (Table 6) and inversely correlated with soil moisture and translocating phloem thickness. As expected, the temperature during the best spraying period was rather warm because honey mesquite was found to grow readily at a minimum average temperature of 65° to 67° F. Maximum soil temperature at a depth of 3 feet occurred in four regression equations (Table 7); max- imum soil temperature at a depth of l foot was less correlated with plant control. Maximum soil tem- peratures at depths of 1 and 3 feet were 58°-93° F and 52°—83° F, respectively (Table 3). Apparently ‘l7 a minimum temperature is required for rapid plant growth. Dahl et al. (2) found that soil tempera- ture above a threshold of 80° F was best for killing honey mesquite. These findings also support their work in the early part of the spraying season. How- ever, soil temperature remains high late in the sum- mer when plant control is not obtained; this gives a final negative correlation (Table 4). Maximum soil temperature was an easy value to measure with a recording thermograph and seems to be a useful one to predict plant control early in the spraying season. Percent soil moisture was measured during the season at depths of 0-1, 1-2 and 2-3 feet. Percent soil moisture at a depth of 2-3 feet occurred in 11 regres- sion equations (Table 7) and was the most important variable measured particularly where the time-time” factor was included. Soil moisture at a depth of 0-1 foot was less correlated with honey mesquite responses than the 2-3-foot measurement. The 1-2-foot depth measurement was not included in the regression equa- tions (Table 7) but presumably could be used almost interchangeably with the 2-3-foot measurement be- cause of the 0.94 correlation (Table 6). The influence of soil moisture needs to be investigated further, par- ticularly at the end of the spraying season. Numerous regression equations were developed, which are not presented here, that included both a soil temperature and a soil moisture variable. As expected, low levels of soil moisture (Table 3) were directly correlated with limited moisture in the plant (high leaf moisture stress) (Table 5). Maximum and minimum percent relative humid- ity measurements taken during the week prior to spraying failed to be highly correlated with honey mesquite control. Generally maximum and minimum percent relative humidity ranged between 90 and M100 percent and 36 and 50 percent, respectively. High daily fluctuations and small seasonal trends seem to limit the use of humidity measurements for predicting plant control. Rainfall the week prior to spraying and total daily solar radiation the day of spraying do not seem to be extremely useful for predicting honey mesquite control because of erratic response to them. The final decision on the factors to be measured in developing a predicting indicator for the control of honey mesquite depends on (1) the ‘high correlation of measurement with control, (2) the reproducibility of measurement and (8) the cost of measurement. At present, measurements of soil temperature and soil moisture seem to be the most useful. However, a variable that measures the general status of the plant would seem to be the most useful. Phloem thickness, rate of new xylem ring radial growth and predawn leaf moisture stress seem to be the best measured. Measurement of short term daily variables such as rate of dye movement seems to be highly influenced by numerous environmental variables. Simple correlations of the environmental tors recorded 1 week before spraying can be divid roughly into two groups. Maximum air temperat ._ minimum air temperature and maximum soil i peratures at 1 and 3 feet were directly correlated one group. Maximum and minimum relative hul ity, percent soil moisture at 0-1, 1-2 and 2-3 feet rainfall were directly correlated,‘ but inversely n related with the first group. Generally, the high direct correlations of the factors recorded 1 w prior to spraying were either the same variables corded at other time periods or similar variables s as other extremes or depths at the same time peri‘ A number of these variables are correlated cl‘ enough to eliminate the need for collecting data more than one variable in future studies. The sim correlations with the herbicide treatments decr as the effectiveness of the treatments decreased. l Acknowledgment _ This study was a cooperative investigation of Agricultural Research Service, U.S. Department Agriculture, and The Texas Agricultural Experi ; Station. ‘i Literature Cited . 1. Brady, H. A. 1971. Spray date effects on behavior of t: cides on brush. Weed Sci. 19:200-202. 2. Dahl, B. E., R. B. Wadley, M. R. George and j. L. r t 1971. Influence of site on mesquite mortality from 2,4 I . ]. Range Mgt. 24:210-215. 3. Davis, F. S., R. E. Meyer, j. R. Baur, and R. W. hi) 1972. Herbicide concentrations in honey mesquite phl Weed Sci. 20:264-267. . 4. Fisher, c. E., J. 1.. Fults and Henry Hopp. 1946. 1" -' affecting action of oils and water-soluble chemicals in quite eradication. Ecol. Monogr. 16:100-126. i 5. Fisher, C. E., C. H. Meadors, R. Behrens, E. D. Robi 1 T. Marion and H. L. Morton. 1959. Control of mesq on grazing lands. Tex. Agr. Expt. Sta. Bull. 935. 24 6. Fisher, C. E., E. D. Robison, G. O. Hoffman, C. H. w 5 and B. T. Cross. 1970. Aerial application of chemicals; control of brush on rangelands. Tex. Agr. Expt. Sta. i Rept. 2801. pp 5-11. 7. Flynt, T. 0., T. E. Riley, R. W. Bovey and R. E. 1971. Auger soil sampler for herbicide residues. 19:583-584. ' s. Haas, R. H., and J. n. Dodd. 1970. Seasonal water .-f pattern in honey mesquite. Tex. Agr. Expt. Sta. Prog. ' 2815. pp 59-62. 9. Hall, W. C., and j. Hacskalyo. 1963. Methods and cedures for plant biochemical and physiological The Exchange Store, College Station, Texas. pp 39-44.: 1o. McMillan, Calvin and ]. T. Peacock. 1964. Bud-bu ‘ - in diverse populations of mesquite (Prosopiszlegumin _, under uniform conditions. Southwestern Naturalist 9; 188. ll. Meyer, R. E., R. W. Bovey, T. E. Riley and W. T. Mc 1972. Leaf removal interval effect after sprays to V plants. Weed Sci. 20:498-501. E 12. Meyer, R. E., R. H. Haas and H. L. Morton. 1965. . quite stem, its structure, seasonal growth characte ' and area of active xylem dye movement. Proc. Sou -. Weed Sci. Soc. Conf. l8z632 (Abstr.). '. H. Haas and C. W. Wendt. 1972. Inter- mental variables and growth and develop- mesquite. Botanical Gazette (In press). ‘PH. L. Morton, R. H. Haas, E. D. Robison . 1971. Morphology and anatomy of honey A Tech. Bull. 1423. 186 pp. ‘C. E. Fisher, and B. T. Cross. 1970. Con- it: and associated West Texas brush with 16. 2,4,5-T/picloram ‘combinations. Proc. Southern Weed Sci. Soc. 232219 (Abstr.). Robison, E. D., R. E. Meyer, B. T. Cross and H. L. Morton. 1970. Influence of preconditioning defoliations on honey mesquite control. Tex. Agr. Expt. Sta. Prog. Rept. 2808. pp 31-34. Wildman, S. G., and E. Hansen. 1940. A semi-micro meth- ed for determination of reducing sugars. Plant Physiol. 15:719-729. The Texas Agricultural Experiment Station Texas A8cM University College Station, Texas 77843 J. E. Miller, Director Publication Penalty For Private Use, $300 POSTAGE AND FEES PAID U.S. DEPARTMENT OF AGRICULTURE AGR 101