Seasonal Abundance and Dispersal of the Qotton as Related to Host Plant Phenology F leahopper Contents FIS®®KIUIUI (Ila-hulk OJ N) r--- N) r-I A r-i U1 Summary The cotton fleahopper poses a double threat to cot- ton production. One is the direct loss in yield and de- layed maturity caused by the insects destroying early fruiting structures; the second is the possibility of a Heliothis spp. outbreak following chemical treatment aimed at the control of the cotton fleahopper. Methods of managing the cotton fleahopper without adverse side effects and sacrificing yield and earliness of the crop are needed. Conditions determining when cot- ton is threatened by fleahopper migrations were studied. The host plants of the cotton fleahopper, Pseudatomoscelis seriatus (Renter), infested first during the growing season were those adjoining the previous year’s croton (Croton capitatus Michx.) stalks, the pri- mary overwintering host plant in the study area. Cud- weed (Gnaphalium spp.), cutleaf eveningprimrose (Oenothera laciniata Hill var. laciniata), and croton for the current season were present. Cudweed served as a host for nymphs hatching from overwintering eggs and for subsequent generations, until plant senescence in late June. Cutleaf eveningprimrose, not adjacent to the previ- ous year’s croton, was not infested until first generation adults were present. Initial cotton fleahopper flight ac- tivity coincided with the time the first adults were de- tected on the host plants. Fluctuations in numbers of adult cotton fleahoppers on cutleaf eveningprimrose corresponded to the flowering pattern of the plant — decreased flowering and fewer adults present. Cutleaf 2 Summary Introduction Methods and Materials Site and Plant Selection Cotton Fleahopper Aerial Movement Results Seasonal Abundance Seasonal Flight Activity j Aerial Movement of Cotton Fleahopper Nym Host Plant Phenology Host Sequence ‘ {h .: Discussion Literature Cited Acknowledgments w cudweed. ? t eveningprimrose was infested about 20 days longer g Of two groupsof showy sundrops (O. speciosa A var. speciosa) were studied, one was subjected to lated mowing and the other served as a control. S sundrops was initially infested by adults. Apparf reproduction and population increase was low on? host. The manipulated plants remained a host abo days longer than the control but they were not infi as long in the season as cutleaf eveningprimrose} numbers of adults on showy sundrops were appar-i related to the flowering pattern and percent mois the plant. f Horsemint (Monarda punctata L.) was inf, after first generation adults were available, and at supported high numbers of cotton fleahoppers. bers of cotton fleahoppers sharply increased with bud initiation and declined prior to peak HOWGII; production. ' The numbers of cotton fleahoppers detected on ton did not increase until squaring. Substantial activity in late June coincided with decreasing p lence of cotton fleahoppers on cutleaf eveningpri and horsemint and increasing numbers on cotton. infestation level on cotton was low, and apparently reproduction and population increase occurred. Croton served as a host plant both for overwinte eggs and throughout the growing season. * (Seasonal Abundance and Dispersal of the Cotton Fleahopper as Related to Host Plant Phenology The cotton fleahopper poses a double threat to cot- ton production. One is the direct loss in yield and de- layed maturity caused by the insects destroying early fruiting structures; the second is the possibility of a Heliothis spp. outbreak following chemical treatment aimed at the control of the cotton fleahopper. Early in “the growing season, beneficial insects are particularly e effective in keeping H eliothis spp. in check. Frequently, insecticides applied for cotton fleahopper control lower the beneficial insect population, allowing H eliothis spp. to increase to damaging levels (Anonymous 1973; Gaines, 1942; Ewing and Ivy, 1943; Ridgway, et al., 1967). Methods of managing the cotton fleahopper without {adverse side effects and sacrifice of yield and earliness of {he crop are needed. Since plant species other than cot- i gton are also infested by the cotton fleahopper, the vari- ions alternate host plants probably influence the cotton Lileahopper infestation in cotton. The cotton fleahopper was present in the United fStates in the late 1800’s but did not damage cotton until £1920 (Reinhard, 1926a). Early workers soon determined that cotton was not the only and perhaps not even the preferred host of the cotton fleahopper. Wherever cro- ton (Croton spp.) occurred, Reinhard (1926b) found it to be the most important host plant. Fletcher (1940) ob- served that large numbers of the cotton fleahopper built up on the eveningprimroses (Oenothera laciniata Hill ind O. speciosa N utt.), and horsemint (Monarda spp.). Denothera laciniata was especially important because it was available early in spring before croton came up. During this time it grew in the rosette form and afforded excellent protection for nymphs, especially from rain. Fletclfer also reported O. speciosa as being a favorite Food and the first plant on which he found adults in the rpring. F Respectively, (formerly) research assistant, (currently) field research ‘i representative, Chemagro Agricultural Division,‘ Mobay Chemical Corporation; associate professor and technician, The Texas Agricul- tural Experiment Station (Department of Entomology). Mention of a trademark or a proprietary product does not constitute a guarantee or a warranty of the product by The Texas Agricultural Experiment Station and does not imply its approval to the exclusion of rther products that also may be suitable. *L. K. Almand W. L. Sterling C. L. Green During April and May at College Station, Reinhard (1927) found purple cudweed (Gnaphalium purpureum L.) and O. laciniata frequently infested with flea- hoppers. Hixon (1941) found cotton fleahoppers on 84 species of plants, with the most important species be- longing to the genera Oenothera, Monarda, Solanum, ' and Croton. Some confusion, however, exists as to when cotton fleahoppers transfer from alternate host plants to cotton. Reinhard (1926b) reported that during March, when overwintering eggs began to hatch, nymphs fed on prac- tically any succulent weeds or even grass. Soon after reaching maturity, the insects mated and began to depo- sit eggs primarily in croton and cotton. Reinhard (1928) stated that cotton fleahoppers hatch- ing from overwintering eggs comprised the early de- structive infestation. When climatic conditions delayed emergence of nymphs from overwintering eggs and cot- ton was planted at the average date, conditions were favorable for extensive injury to crop. However, in some sections of Texas, Reinhard (1927) detected emergence of cotton fleahoppers from overwintering before either cotton or croton was available as a food plant. Fletcher (1940) reported that horsemint differed from croton and the species of Oenothera in that it was more attractive to cotton ‘fleahoppers later in its life span. The adults left as these plants matured and became less attractive. Ewing (1927) observed a distinct migra- tion, probably from horsemint, to cotton during the first 10 days of Iune, which was about the time or im- mediately before horsemint began to mature. Thomas (1936) noted that as alternate host plants approached maturity, many adult cotton fleahoppers mi- grated to more favorable host plants. This transfer of hosts began about the time cotton came up and lasted about 1 month. Gaines and Ewing (1938) stated that adult cotton fleahoppers began to drift into cotton from native food plants during April, while most of the disper- sal usually occurred from about the middle of May until the first part of Iune. First generation adult cotton fleahoppers infested the late spring and early summer hosts, horsemint, silverleaf nightshade (Solanum elaeagnifolium Cav.), and in some cases O. lacinata, in observations by Hixon 3 (1941). He believedhthat cotton was attacked by adults leaving these host plants. There is apparently some disagreement on the im- portance of fleahopper populations developing on alter- nate host plants. Eddy (1927) observed that croton and species of Oenothera served as a reservoir and continu- ous source of infestations as long as the host plants were available. Cotton growing in clean fields surrounded by woods or other situations remote from alternate host plants was not infested until late in the season and then did not develop heavy infestations. In the Rio Crande Valley of Texas, Schuster et al. (1969) suggested that native host plants did not greatly increase cotton flea- hopper populations before cotton was growing. The numbers of fleahoppers increased simultaneously on both native and cultivated plants, and it appeared they were infested first by adults. There is also disagreement as to the destination of nymphs as they hatch from overwintering eggs. Reinhard (1927, 1928) considered it likely that young nymphs may be carried considerable distances by wind and spread over the surrounding territory as they hatch from overwintering eggs. Early instar nymphs were on cotton plants in fields isolated from any source of infesta- tions (Reinhard, 1926b) early in the season before the appearance of adults. However, in preliminary tests, Reinhard could not demonstrate that nymphs were blown in the air. Click (1939) exposed collecting screens from an airplane for more than 1,007 hours. He captured two adult fleahoppers and one nymph at 20 feet altitude, two adults at 200 feet, two adults at 1,000 feet, and one adult at 2,000 feet. Coad (1931) reported capturing cotton fleahoppers by airplane at altitudes of up to 5,000 feet. Hixon (1941) found that nymphs hatching from overwintering eggs in Croton capitatus Michx. fed ' mainly on plants adjacent to that species. No fleahoppers were present on the same plants growing 100 feet from old croton. Nymphs did not appear to migrate; however, the adults were distinctly migratory. Thomas and Owen (1937) stated that some of the insects newly hatched from overwintering eggs are forced to feed on almost any form of tender vegetation. When host plants mature, Fletcher (1940) theorized that the adults leave and may disperse by flight, assisted by the wind, over wide areas. Gaines and Ewing (1938) reported that during the winter of 1933-34 about 10,000 acres of croton were destroyed in Calhoun County, Texas, in an attempt to control the fleahopper. The in- itial infestation the next spring was considerably less, but there was a heavy reinfestation of adult cotton flea- hoppers in the area apparently from a distance of at least 20 miles. F leahoppers moving with the wind could thus be carried long distances. Gaines and Ewing (1938) also observed that cotton fleahoppers transferring from na- tive food plants to cotton dispersed from about mid-May until the first part of June. Balloons released in various areas of Texas during this period drifted primarily to the north and northwest. 4 After extensive flooding of the Mississippi ' _ 1927, Click (1939) found cotton fleahoppers extr scarce. The wild host plants were killed by the flo A until these plants were established again, the fleahopper did negligible damage to cotton in the‘ Click's data suggest that fleahoppers do not always Q great distances to cotton and thus tend to disagreg; the findings of Gaines and Ewing. Since weather conditions vary from year to‘ influencing cotton planting dates and plant gr calendar dates alone probably are insufficient for pr ing when cotton fleahoppers will leave their alt host plants. A better understanding of host plan cotton fleahopper phenology may provide a able basis for determining when cotton is threaten fleahopper migrations. 7 Methods and Materials Site and Plant Selection The study of conditions determining when co threatened by fleahopper migrations was conduct) the Ellis Unit farm of the Texas Department of Co; tions, located about 15 miles northeast of Hunt i" Texas. The soil types ranged from loamy sand to black clay. The primary host plants detected in the p were cudweed, cutleaf eveningprimrose, showy _drops, horsemint or spotted beebalm, woolly goatweed, and commercially planted cotton (Goss _i hirsutum L.). a f Hixon (1941) reported that cotton fleaho nymphs hatching from overwintering eggs feed on species’ of plants adjacent to the overwintering plants. In the study area, croton was apparently the? i mary overwintering host; therefore, plants grog under the previous year’s croton were selected for s l cudweed and croton were the predominant hosts, y, panied by a lesser amount of cutleaf eveningprim- For comparison, cutleaf eveningprimrose and c growing more than 200 feet from the previous y croton were selected. The remaining alternate host pl included in the study, horsemint, showy sundrops cotton, were not adjacent to a source of overwinte, cotton fleahoppers. I Being a common roadside and ditchbank inhabi i showy sundrops is subject to mowing during the gro season. Plants which are mowed often produce regro thus delaying maturity and senescence. To determined, effect of this development on the cotton fleahopperi tation on such host plants, two adjacent locations ofsh sundrops were used. The plants at one site were cli I just below the flower bud area to simulate mowing. at an adjacent site, used as a control, were not clipp In each of the study sites, 10 to 15 plantsof Q‘ species were labeled for data collection on the same pl each inspection period. An electric fence was pl " around the plots to prevent cattle from chewing f‘ labels. Plant phenology records of the tagged plants p taken at weekly intervals during the growing seas J. .- - pt during foul weather. The cotton fleahopper infes- l} on on the various host plant species was determined sampling 2O plant terminals (on tagged plants) twice with a D-Vac® hand vacuum sampler. As an indication of plant succulence, the percent isture of plants of each species was determined weekly collecting the terminal 6-8 inches from 1O plants. The 4 areas with all unfolded leaves removed and the leaf- stem comprised separate samples. The samples _e weighed before and after oven exposure to 200° C .24 hours. To anticipate fluctuation in plant moisture, the per- (if: soil moisture was determined weekly. A sample oftfie rrpperfiaflfkrrrzfftrwerfiaflfofa Ffliircfi (“p soil core was taken at each host plant area. The ent moisture was determined by weight before and oven exposure to 200° C for 24 hours. tton Fleahopper Aerial Movement Cotton suffers the heaviest damage from cotton flea- Iper attack during the early fruiting period of plant iwth. Therefore, the time of cotton fleahopper move- nt from the alternate host to cotton is important. In- f ation on the seasonal flight activities may also con- ute to a better understanding of cotton fleahopper sequence. The seasonal flight activity of the cotton fleahopper determined by using seven traps, each consisting of Egallon size plastic cans, placed 6 feet above ground and (bated with Tacky-Trap‘? The traps were located in the llmediate vicinity of various host plants and in between tes of host plant occurrence. Traps were inspected twice weekly and recoated with Tacky-Trap® as needed. The possibility of cotton fleahopper nymphs being ansported by the wind was investigated further. Aerial lmples for cotton fleahopper nymphs which may be air time were taken in the truck-mounted net described by lmand, et al. (1975). Samples were taken from early March until late Sep- ember on the Ellis farm. The pickup truck with net tached was not driven in the growing area of plants so as 0t to disturb the plants and insects present. The collec- 0n route was semi-circular with about a 1-mile radius. 'uring heavy emergence of cotton fleahopper nymphs in ifi spring at College Station, Texas, additional samples ere taken downwind from overwintered croton. Num- airs of adult cotton fleahoppers captured during the sea- a3 were used for determining seasonal flight activity. The response of newly hatched first instar cotton aahopper nymphs to various wind speeds also was de- rmined in a wind tunnel with a testing area of about 3 IlJiC feet. Nymphs iless than 24 hours old which had Ltched from overwintering eggs (procedure developed r Sterling and Plapp, 1972) were collected for testing 1d held in 1-gallon cartons. An overwintered stem of oton collected from the field was placed in the carton; hen nymphs had moved onto it, the stem was removed id placed in the wind tunnel test chamber. Wind ve- cities up to about 3O miles per hour (mph) could be obtained, but sudden gusts of wind could not be simu- lated. Results Seasonal Abundance Cudweed was the predominant host growing be- neath the previous year’s croton plants in the study area. Having germinated in the fall and being in the rosette growth stage during the winter, it was readily accessible to the nymphs as they hatched from overwintering eggs awfl earfi/ fr! tfie season (Figure 1). The first adult cotton fleahoppers in the area were found on cudweed. Apparently cudweed was a good host for the cotton fleahopper as 95 adults and 130 nymphs per 100 terminals were detected. At College Station, about 9O percent of the cotton fleahopper emergence from overwintering eggs had oc- curred by mid-April (Sterling, unpublished data). Assum- ing a similar occurrence in the study area, the predomi- nance of cotton fleahopper nymphs on cudweed in April likely represents the peak time of emergence from over- wintering eggs. The peaks in nymphal abundance in early May and early June are evidence of cotton fleahopper reproduction on cudweed. Cudweed evidently serves both as a host for cotton fleahoppers batching from over- wintering eggs, and as a reproduction site to further increase the cotton fleahopper population. Cutleaf eveningprimrose, growing beneath an over- wintering host plant of cotton fleahoppers, was infested first by nymphs followed by adults (Figure 1). Adults, apparently first generation, were detected a few days later than those found on cudweed. A short time after the first adults were present, the plants were inadvertently destroyed. Additional cutleaf eveningprimrose plants studied were not adjacent to an overwintering source of cotton fleahoppers and were not infested until first gen- eration adults were available. Once adult cotton fleahop- pers were present, they reproduced on cutleaf eve- ningprimrose as evidenced by the presence of nymphs. Cutleaf eveningprimrose, like cudweed, served as a host to both cotton fleahoppers hatching from overwintering eggs and to the next generation. ' Showy sundrops did not support as many cotton fleahoppers as cudweed or cutleaf eveningprimrose (Fig- ure 1). The showy sundrops plants, not near a source of overwintering eggs, were not infested until after first generation adults were present. The cotton fleahopper population increased little on showy sundrops; however, higher numbers occurred on the plants cut back (simu- lated mowing) than on the undisturbed plants. The peak in adult numbers in late May on the manipulated plants was followed by a peak in the abundance of nymphs in mid-June. The last peak in the number of adults, in late June, probably represents the mid-June nymphs reach- ing maturity. Plants with frequent bud removal were infested throughout June; uncut plants were not infested past early June — a difference of about 3 weeks. This difference will be discussed in a later section. Figure 1. Cotton fleahopper seasonal abundance on several host ‘ ‘ ' ‘ ' F“ ' ' ' ' ‘ ‘ ‘ ‘ ' ‘ ' ' i plants and seasonal flight activity. 5° _ CRgESETéWITH COMPETING VEGETATION) _ 3o _ 0000000 _ I I I I I I I I I I I I I I 1 I I I I I - p-~ d " a 1Q I ‘I 13o - HORSEMINT _ L l ‘ l ,...,...,....,....,. ' ADULTS '1——- N0 "' r I I I I I I I I I I I F I 2; I I I I I I H0 - NYMPHS TEéMlNALs -i - CROTON ('IN DISTURBED SOIL) I 1;; I 9o ' 7 '4° ' ADULTS - ' — - NYMPHS ‘ - HO l- 70 “ - l- “ - 90 - 5O - - - . _ 7Q . 3O " I ' l- i I- l0 1' - ' I I I I I"? I": q I I I I I I I I 3Q l- ‘ p l 0000 130 - t» CUTLEAF EVEN INGPRIMROSE ' 1 1 1 1 1 1 1 1 1 1""- P i " I I I I I I I I I I I I I I é no»- NYMPHS-"m- QEIQQISKLS I so TAMCOT SP 37 COTTON - Z " ADULTS-——- ' g 90" i! I 00000000 ‘ l‘ ' a “ 5 _ 1V1 l ‘ c 10- 5 J l0 LTREATEID I I _ ,_%_ _ : 5Q‘ - I I I I T I I I r 1 I I I I I I I I I I fi gal - ' 5Q- STONEVILLE 7 COTTON '1 e 3°" : 1 ADULTS——- ' QQ 1- . ... 0000000 l‘ "‘ E 10k ' -- 4 I D l l I I I I I I I4 1o_ l A .. Z I I I I I I I I I I I I I I I I - __ h T I I I I I I F I F I I I I I I I. I I m‘ g CUWEED ' $01 SHONY SUNDROPS (CONTROL) 1 ' ADULTS -- ' - ADULTS __ 1 no: NYMPHg J 3o: NYMPHS N0./l00 TERMINALS i 9O - I : " 1Q I- 4 I i 1, : ‘ l 1 1 1 1 1 1 1 1 1 1 1 1 7° '- i. - fif I I —_I_' I I I I I T I I I j I I I I I ' ' I 50- SHONY SUNDROPS (SIMULATED MOWING) ' 5° - - ADULTS —— - I- " _ 000 00000 _ so -1 ' - ' - I ‘ 1o » - IQ u I I I I g1 A I I I I I I L L l l n I l I I I l I I I 1 I i I I I I, I I I F I I I I I " TRAPS MEANNO/TRAP- b i‘ d 7 I- 0000000 A 7 ' _ NET 0000000 MEAN NOJMILE I - 6 - ' _ . .. _ 5 _ _ g; 5: _ .. .l u - * ' " 2 4 C - b c1 O _ 3 - - m 3" - - g 2- : P - Z I l- ‘ I I _ '0.’ 0 .. . O .0.“ .1 l *“‘ . " m A31 10201020309 l9 2991929 a 182a 71727 7 3] 12%|)?’ ‘Ragga 9118i}? 9111:9139 aAIsscsza 75's‘? 7 APRIL MAY JUNE JULY AUG SEPT Horsemint, also not near a source of nymphs hatch- from overwintering eggs, was not infested until after {appearance of first generation adults (Figure 1). This ‘twas a good host for cotton fleahoppers and a site for louction. Although large numbers of adults were ent on the host, low numbers of nymphs were pres- Qfollowing the peak abundance in late May. The low phal population could have resulted from the pres- of large numbers of Orius spp. Sampling error may ‘j have influenced the number of nymphs detected. on fleahopper nymphs predominantly inhabit the er bud area, which, on horsemint, is very tight and s numerous hiding places for the small insects. the low numbers of nymphs corresponded with the iéring pattern of horsemint, the vaccum sampler may been inefficient in extracting the nymphs from the er bud. Nevertheless, the presence of rather high jibers of nymphs at one time indicates the potential of Jon fleahoppers for increasing their numbers on emint. 5. On the two cotton varieties, the cotton fleahopper ~ tation was light and remained below damaging levels re I). The fast maturing, short season variety, Tam- SP 37, supported somewhat higher numbers than eville 7A. Tamcot SP 37 cotton seemingly was a host hich at least some reproduction and population in- lie occurred. The peaks in abundance of adults and phs possibly represent generations. It is doubtful ; much reproduction occurred on Stoneville 7A cot- The last species of host plant included in the study, Q on, not only provided overwintering quarters but also a host throughout the growing season. The croton grew up under the previous years’ croton stalks were F: ently infested first by nymphs hatching from over- itering eggs (Figure 1). A short time after the first larance of adults, the plants were inadvertently de- ifyed, necessitating tagging additional plants. The sec- group of plants did not grow under a source of over- itering eggs and were infested first by adults. The ts were in pasture land with competing vegetation maintained only slow growth. The highest numbers ults detected were in late April and early May and ably represented first generation adults. Thereafter fipopulation remained rather static and likely did not tribute to an increase in the overall cotton fleahopper i lation. It is also doubtful that many overwintering f. were deposited in these host plants because the € tcotton fleahopper population was below detectible T ls during the fall when overwintering eggs are depo- t, In another area of pasture, the soil was disturbed by "ng, and a dense stand of croton resulted. Although J croton came up later, it grew faster and developed f larger plants than did the croton with competing station. This larger, faster growing croton also sup- ‘ ed a much larger cotton fleahopper population (Fig- than that undisturbed. Seasonal Flight Activity While neither the exact source of the adults captured in flight nor their destination was determined, a compari- son of adult abundance on the various host plants with the peak times of flight may indicate which host plants cotton fleahoppers were leaving and the host plants to which they went. The aerial net and the sticky traps yielded similar results in detecting cotton fleahopper flight. Although more cotton fieahoppers were captured on the sticky traps, the peak times of flight as determined by the two methods were in accord (Figure 1). There were primarily five peaks of flight activity of cotton fleahoppers during the season. The first evidence of flight coincided with the detection of the first adults, first on cudweed, and a few days later on cutleaf evening- primrose and croton. These host plants, growing under a source of overwintering cotton fleahoppers, were infested first by nymphs. The adults likely were first generation. Flight activity during late April and early May coin- cided with detection of cotton fleahoppers on cutleaf eveningprimrose, showy sundrops, and horsemint not adjacent to an overwintering source. These adults un- doubtedly were first generation leaving the nymphal host plants at maturity. During mid and late May the number of adult cotton fleahoppers decreased on cudweed and cutleaf evening- primrose and increased on showy sundrops and horse- mint. Flight activity indicated that cotton fleahoppers were leaving cudweed and cutleaf eveningprimrose and going to showy sundrops and horsemint. The relative peak abundance of cotton fleahoppers on cutleaf eveningprimrose, showy sundrops, and horsemint in mid-June was not accompanied by flight activity and probably represents reproduction on these hosts. - In late June a substantial increase in flight resulted in the greatest number of fleahoppers caught for the season. This coincided with a decrease in adult numbers on the two remaining seasonal weed hosts, cutleaf eveningprim- rose and horsemint. The only known host plants remain- ing in July are cotton and croton. Following a lull in July, flight activity increased in mid-August. Some cotton on the farm received insec- ticide treatment for H eliothis spp. control about this time, reducing cotton fleahopper numbers; probably they were going to croton. Although cotton fleahopper reproduction and population increase on cotton were not substantial, migration of the insects from 800 acres of cotton to fewer acres of croton could result in a heavy concentration on croton. In early September much croton on the Ellis farm was mowed. The September peak in flight activity likely represents cotton fieahoppers leaving the mowed croton. F7 Aerial Movement of Cotton Fleahopper Nymphs Cotton fleahopper nymphs were not captured in the aerial net. From early March until mid-April about 260,000 cubic feet of air was sampled. -During that time about 90 percent of the nymphal emergence from over- wintering eggs occurred (Sterling, unpublished data). It seems likely that if many of these nymphs were air borne, they would have been detected. Additionally, about 10 million cubic feet of air was sampled during the remainder of the season. Although high nymphal popu- lations were in the area during the study, no air borne cotton fleahopper nymphs were detected. In the wind tunnel, cotton fleahopper nymphs were frequently near the top of the test plant when the wind speed was low and there was no visible movement of the plants. When the wind speed increased, the nymphs tended to move to the lower portions of the plant where there was less whipping motion. At wind speeds up to about 30 mph, the cotton fleahopper nymphs were able to remain on the plant and move about. Host Plant Phenology Cudweed: While there are numerous reports of the various host plants of the cotton fleahopper, the more important hosts include only a few genera of plants. Cudweed was not included among the more preferred host plants; therefore, phenological data for cudweed were not considered in the original planning, and the data collected are not as detailed as desired. Cudweed seeds germinate in the fall, and the plants are in the rosette stage until spring, when flower shoots originate from the rosette. Flower bud production on the cudweed plants was initiated in early April, and senescence occurred in late June. Cudweed was suitable as at host for feeding and reproduction during the rosette - stage and through maturity. Cutleaf Eveningprimrose: Coefficient of correlation and linear and curvilinear regression analyses were made of data pertaining to adult cotton fleahopper sea-_ sonal abundance on the various host plants, flowering pattern of the host plants, plant stem and bud moisture content, and soil moisture. Those having significance at the .05 level will be discussed. The fluctuations in the numbers of adult cotton fleahoppers on cutleaf eveningprimrose during the grow- ing season closely corresponded to the flowering pattern of the host plant (linear regression r = 0.85; parabolic regression r = 0.86). The adult cotton fleahopper popu- lation decreased as the number of flower buds decreased (Figure 2). The flowering pattern corresponded to the upper 6 inches of soil moisture with a lag of about 7 days (linear regression r = 0.96; parabolic regression r = 0.99). As the amount of soil moisture declined, both the number of flower buds and the number of adult cotton fleahoppers declined. When adequate soil moisture re- turned, flowering and the incidence of adult cotton flea- hoppers increased (Figure 2). Number of adult cotton fleahoppers and cutleaf eveningprimrose stem and bud moisture (maximum r 8 value obtained = 0.62) were not significantly corrp However, curvilinear (parabolic) and linear regr, were similar (r = 0.89 and 0.88 respectivel plant moisture and soil moisture. p S howy Sundrops: Although relatively low nu l, of cotton fleahoppers were detected on the sho I drops plants, a comparison of cotton fleahopper nu; and the plant phenology data onthe two groups of sundrops plants indicates some important occu (Figure 2). As mentioned earlier, showy sundropi not infested until first generation adults were p ‘l On the control plants, adult cotton fleahoppers -*” the highest level during the peak of flower bu duction. As the number of flower buds declined, the number of adult cotton fleahoppers and the? moisture content. The plants were infested until cence occurred in early Iune. I The flowering pattern on the manipulated sundrops plants and the control plants differe period of flower bud production was lengthened plant senescence was delayed about 3 weeks on thf nipulated plants (Figure 2). The time of first occu of cotton fleahoppers was about the same on the and manipulated plants, but the period of infestati‘ lengthened with the delay in plant senescence. The peak numbers of adult cotton fleahoppe the manipulated plants coincided with the time of est number of flower buds; however, the adult i 1 fleahopper numbers declined prior to a reducti flowering (Figure 2). Numbers of adult cotton hoppers declined as plant moisture decreased. For showy sundrops, both manipulated and " plants, results indicate that both flowering patte t. plant moisture influence the incidence of adult fleahoppers. On the manipulated plants, the numyf adult cotton fleahoppers decreased when flower‘; production was still high but plant moisture was de ing. Later, increased plant moisture content was P; panied by a reduction in flowering without a subseq rise to previous levels in the number of adult fleahoppers. it On the control plants, plant moisture and {f bud production decreased, coincidental with a decli i adult cotton fleahopper population. A subsequent s‘ term increase in plant moisture was accompanied i) continual decrease in flowering and continued low y) dence of adult cotton fleahoppers. Regression analysis showed a parabolic relatio 1 between plant moisture and the upper 6 inches moisture content (mean r = 0.89). Plant moisture delay response of about 14 days to decreasing moisture; in cutleaf eveningprimrose, the delay: sponse was about 7 days. The difference in the d; responses of the two Oenothera species was likely du soil type: cutleaf eveningprimrose was growing loamy sand and showy sundrops in a clay type soil. 03°, ' soil has a slower rate of penetration and percolation . T greater water-holding capacity than sandy soil. Horsemint: Cotton fleahopper numbers on ho a mint did not closely follow the flowering habits of i’ c. Comparison of host plant phenology, plant moisture, soil ‘A , and cotton fleahopper seasonal abundance. I I I I I I I I I I I GP RIMROS % SOIL MOISTURE I I I I I I I I I I I STEM -—- CUTLEAF EVENINGPR IMROSE % PLANT ' BUD MOISTURE ‘ CUTLEAF EVENINGPRIMROSE NO. FLOWER BUDS/PLANT 00000;‘... .“ O000OOO0000g........... 0 * ....~OOQ 311020301020309192991929 818 8 JUNE JULY CUTLEAF EVENINGPRIMROSE ADULTS———- NYMpHsunn, NO./100 TERMINALS ‘s 2 7172 AUG SEPT 77 suouv SUNDROPS (SIMULATED MONING) 0 ' % SOIL MOISTURE - 6 - 120000000 9O 80- 70" IIIITIIIITIIIIfiIIIII SHOWY SUNDROPS (SIMULATED MONING) 5TEM""" % PLANT ‘ BUD”“”' MOISTURE - IIIIIIIIIIIIIIIIIjfi SHONY SUNDROPS (SIMULATED MONING) " NO. FLOWER BUDS/PLANT ‘ I I I I I I I I I I I I I I I I I I SHOWY SUNDROPS (SIMULATED MONING) ' ADULTS - NYMPHS - _ snowv SUNDROPS (CONTROL) ' : 22 — 6 C ?2Q..’..' ‘yo q _ MOISTURE # 18- - 14 - - 1o- — 9o I I I I I I I I I I I I‘fi I I I I I - _ \, _,- suowv SUNDROPS (CONTROL) _ ao- STEM""' % PLANT - - ‘t BUD MOISTURE 7O " I I I I I I I I I I I I I I I I I1/3 eR0wN-—- ,-' BQLL5 ..... .. - f l I I I I I I I I - TAMCOT SP 37 COTTON ' ADULTS -— ,, "' Q o . Q 00o - ‘TREATED _ son 5 ‘ 12 MOISTURE L COTTON ST EM -- BUD .... % PLANT MOISTURE lllllllllllllllllll? I I I I I I " STONEVILLE 7A COTTON - FRUITING PATTERN-NO./PLANT I T I I I I I I I I I I - SQUARES <1/ 3 GROWN ---- ~- r SQUARES>1/3 GROWN k BQLLS l L I I l l l l _L I .. ‘STONEVILLE 7A COTTON " ADULTS —-— ~ NYMPHS I j‘ - n- ¥ - TREATED / A Q ‘r . v 31 10 203010 20309 1929 9 192981828 8 APRIL MAY JUNE JULY AUG SEPT - 5 1 6 I 12...‘... ICROTON (WITH COMPETING I vsesm 10m 2. (continued) I I I I I I I I I I I I I I I Izoron (WITH COMPETING VEGETATION) % SOIL MOISTURE I I I I I I I I I I I I I I I I . Jp"\v~.uu"~u°'. . , STEM 1- BUD ....... l l I l I l I l I l l l I I l l % PLANT MOISTURE: I I I I I I I I I I I I I I I I CROTON (wma compmue VEGETATION) no. FLOWER BUD TERMINAL/PLANT |u|IIIITIII1II MAY JUNE JULY. AUG I I CROTON (WITH COMPETING VEGETATION) ADULTS ——— 0000000 m 10203010 20309 1929 91929 a 1a 2a a 1a 2s 7 APRIL 5i" IO 9O - '- - ...00~... . T a0 - ' 4 _ CROTON (IN DISTURBED SOIL) _ STEM ——- 7Q - “m” % PLANT MOISTURE - 1 1 1 1 1 l 1 1 1 1 1 1 1 '1 1 1 1 1 1 18 » I J _ CROTON (IN DISTURBED SOIL) 14 NO. FLOWER BUD TERMINALS/PLANT I IO P - 6 - - 2 - d _ CROTON (IN DISTURBED SOIL) _ 2 I % son. MOISTURE l I I l I I l l L l I I l I I l l l l l I I I I I I I I I I T I I I I I I I I fiq L . _ q _ I I40 IIO 99 7O SO 3O IO I I I I I I I I I I - CROTON (IN DISTURBED SOIL) I‘ ADULTS —-i - "' 0000000 - '- f. =1 - I - 2' I ‘E i - .8: q I I . .5 - - '..' 0 0 0 "0 . 1 1 1 1 ‘ 1 I 1 1 1 . |."v'1.m"\'. ‘f9'..}u,I "1'! 1 1 1 i‘ OIO203O 9 I9 29 9 I9 298 I828 7 I7 27 7 3] ‘R9283 MAY JUNE JULY AUG sen ‘l 1 plant (Figure 2). The numbers of insects increased sharply at the time of flower bud initiation; the peak in numbers occurred prior to the peak in flower bud pro- duction. There was a significant inverse-correlation be- tween the prevalence of adult cotton fleahoppers and horsemint stem moisture, described by a parabolic re- gression (r = 0.75). As the stem moisture increased, the number of adults declined. A parabolic relationship also existed between stem moisture and soil moisture (r = 0.85) with about a 7-day delayed response. This is an expected relationship since the horsemint was growing in a loamy sand soil. Cotton: The number of cotton fleahoppers on cotton did not increase appreciably until squaring had begun (Figure 2). Although the cotton had emerged and there was substantial flight activity, the cotton fleahopper population remained low until squaring. Abundance of adult cotton fleahoppers, flowering pattern of cotton, and plant moisture content were not significantly corre- lated. Croton: Croton served as a season-long host of the cotton fleahopper. Growing under the previous year’s stalks, plants were infested by nymphs hatching from overwintering eggs. The flowering pattern of croton ap- parently did not influence the incidence of adult cotton fleahoppers on the host plant (Figure 2). The peak in croton flower bud production occurred in early Sep- tember at the same time cotton fleahoppers were depos- iting overwintering eggs. Both plant stem moisture and flower bud production began to decline about mid- September, indicating the initiation of plant senescence. With the advent of plant senescence, the cotton flea- hopper population drastically declined. There were large differences in the numbers of cot- ton fleahoppers infesting croton growing in disturbed soil and croton growing with competing vegetation (Fig- ‘ ure 2). The croton in disturbed soil grew faster and pro- duced larger plants. From late Iune until mid- September, croton in’ disturbed soil averaged about 89 percent stem moisture content and that with competing vegetation about 87 percent. Thus, the larger croton was somewhat more succulent. There were no significant statistical correlations among abundance of cotton flea- hoppers, plant moisture, and flower bud production on croton. Plant moisture was significantly related to the upper 6 inches of soil moisture (parabolic regression r = 0.76) with about a 7-day delayed response. The croton was growing in a sandy soil. ' Host Sequence The host plants infested first in the season were those in the immediate proximity to a source of over- wintering cotton fleahoppers; others were not infested until first generation adults were present. Cudweed apparently served as an acceptable host until after early Iune; then the cotton fleahopper popula- tion declined on this host. Cutleaf eveningprimrose remained succulent about 20 days longer than cudweed; then cotton fleahopper numbers declined and plant maturity increased. Showy sundrops subjected to simu- 12 lated mowing served as a host about 10 fewer dat cutleaf eveningprimrose; the control showy su ceased as a host about 2O days before the manii plants, serving as a host for the shortest time. ing only the seasonal weed host plants, horsem" tained cotton fleahoppers later in the season t‘ other spring weeds but only about 1 week long cutleaf eveningprimrose. Croton served as l’ throughout the season. Discussion The data indicate that host plants growif mediately adjacent to an overwintering source fleahoppers are infested first in the season. Ho not so situated are not infested until first ge adults are available. These differences in time of tion, coupled with absence of cotton fleahopper in the aerial net, suggest that aerial move nymphs was not a major means of cotton fleahop persal during the period of this study. Adults are?‘ ently responsible for most of the transfer of hoppers from one host plant to another. ‘i Cudweed, previously recorded as a cotto; hopper host plant, was not included among garded as more important for feeding and reprod p Cudweed was an important host plant in the area‘ study. Cudweed can be very important in that it i, in the same situations as croton; that is, disturb soils (Steere, 1969; Fernald, 1950; Correll and Io): i 1970; Fletcher, 1940). With the seeds germinating fall and producing a rosette growth until spring? plants could afford nymphs, newly hatched fro ' wintering eggs, with readily available food and j tion, the same conditions as Eddy (1927) descri, cutleaf eveningprimrose. Additionally, first genf adults produce on cudweed, thus increasing the fleahopper population. i?) Cutleaf eveningprimrose was another imp host plant. It was infested by nymphs hatchin overwintering eggs and served as a host to subs adults for reproduction. Like cudweed, cutleaf ev primrose germinates in the fall, producing a rosettf. growth, and occurs in the same situations as i’ (Fletcher, 1940; Schuster et al., 1969; Steere, 1 eves and Bain, 1947). Therefore, cutleaf evenin’ rose frequently would be found under overwinter ton stalks and would be immediately available to y fleahopper nymphs hatching from overwintering .3 The number of cotton fleahoppers present on leaf eveningprimrose apparently was influenced l?) flowering pattern of the plant. A decrease in flow f production was accompanied by a decrease in of adult cotton fleahoppers. Flowering in cu) j eveningprimrose was in turn influenced by the up inches of soil moisture. Being prevalent in sandy which have low waterholding capacities, cu, eveningprimrose is subjected to fluctuations ing? moisture, depending upon the rainfall pattern. l fluctuations in soil moisture, the flowering pattern which, in turn, produces oscillations in the adult fleahopper numbers. The soil moisture could pos- , be a prediction device for adult cotton fleahopper f) nment of cutleaf eveningprimrose until the plant es senescence in early summer. , e potential for increase of the cotton fleahopper tion was not as great on showy sundrops as on the y early season host plants. Like cutleaf eveningprim- and cudweed, showy sundrops germinates in the and produces a rosette-type growth until spring; § er, it does not abound in the same habitat as cut- ieveningprimrose, cudweed, and croton. Showy i5 ips occurs later in disturbed soils (Fletcher, 1940); 1 ore, it would not be readily available to nymphs ing from overwintering eggs. Data indicate that sundrops was apparently infested first by adults. ‘i; prevalent during the same time of the season as f’ ed and cutleaf eveningprimrose, yet not infested ‘first generation adults were present, showy sun- ?» has less likelihood of hosting cotton fleahopper lation increase. However, removal of flower buds 7“ lated mowing) delayed senescence and lengthened time showy sundrops was available as a host. Thus, Zmowing of roadsides may increase the numbers of ‘oppers in the area. fThe incidence of adult cotton fleahoppers on showy ops was related to plant moisture and the flowering l, of the plant. These in turn were influenced by oisture. A delay of about 14 days in the response of lant to changing soil moisture was apparently due to i) lay soil type. As with cutleaf eveningprimrose, it be possible to monitor soil moisture and plant con- I and predict when cotton fleahoppers will leave I sundrops. However, senescence of these plants in i summer even in the presence of adequate soil a re would complicate these predictions. ~iaApparently, horsemint was more suitable as a cot- feahopper host plant later in the season when com- iv to cudweed, cutleaf eveningprimrose, and showy rops. Although horsemint was infested prior to (ring, the prevalence of adult cotton fleahoppers ly increased as flowering increased. This is similar servations by Fletcher (1940) that horsemint was attractive to cotton fleahoppers later in its life Where croton is an important overwintering host, mint is unlikely to be an important host of nymphs ing from overwintering eggs. The two plant species (unlikely to be prevalent in the same immediate kity at the same time. Croton is more abundant in ly disturbed soil (Fletcher, 1940; Hixon, 1941), eas horsemint is more abundant in areas fallow for ral years, and fluctuations in abundance seem to be ed to abundance of rainfall (Fletcher, 1940). Adult cotton fleahoppers did not leave horsemint as Qering declined but prior to peak flower bud produc- Fletcher reported that adult cotton fleahoppers left emint as the plants matured and became less attrac- p. In the current study, the plants had not reached ,9 maturity when the adults left, but some seed had begun to mature at the sites of the first flowers on the lower portions of the plants, which possibly could be inter- preted as approaching maturity. Ewing’s (1927) findings that adult cotton fleahoppers left horsemint about the time or immediately before horsemint began to mature appear to be more in agreement with the current find- ings. Croton appeared in nearly pure stand in freshly dis- turbed soil and the plants, when compared with those in competing vegetation, grew quite large. In areas where grass was dominant, Fletcher (1940) reported few croton plants, and those small. The current findings were in agreement with, those of Fletcher; in addition, more cot- ton fleahoppers were present on the plants in disturbed soil than on those in competing vegetation. Although the adult cotton fleahopper population on croton (in disturbed soil) fluctuated considerably, the numbers of nymphs remained rather constant until early September when the numbers increased rapidly. Rain- fall may have been influential in this sharp increase. Gaines (1933), reporting a sharp increase in the adult cotton fleahopper population on croton during the fall of 1931 and 1932, attributed the increase to rainfall which produced added plant growth, affording an ideal place for feeding and breeding. Reinhard (l926b) reported that rainfall was required for nymphal emergence from overwintering eggs. In the present study, the eggs from which the nymphs hatched in September could have been deposited earlier in the growing season but did not hatch because of lack of rainfall. Such may have been the case with Gaines’ findings, although nymphal reccords were not presented. Rainfall could have induced a hatch of cotton fleahopper eggs subsequently reflected in the adult records (Gaines, 1933). Gaines’ (1933) data also suggest that croton is a doubtful source in contributing many cotton fleahoppers to the population on cotton. While croton is an important host plant of the cotton fleahoppeF, particularly for over- wintering purposes, current data also indicate that cotton fleahoppers did not move into cotton from croton. Being one of the two remaining host plants at the time of move- ment into cotton, croton likely was competing with cotton ’- .. ~/ 1 u‘ n’ >11 1".’ 1'.’ I .. Cotton fleahoppers that infested cotton probably came primarily from horsemint and the species of Oenothera. The source of cotton fleahoppers moving into cotton, however, can be influenced by several factors. Apparently, large cotton fleahopper numbers do not ap- pear on cotton until the initiation of squaring (initiate flower buds), which is also the time when cotton is very susceptible to fleahopper damage. Such was the case reported herein and reported by Gaines (1933). Gaines detected emergence of cotton fleahoppers from over- wintering eggs during March and April of 1931 before cotton was up to a stand, which occurred in May 1931. According to his data, the cotton fleahopper population in cotton did not increase until early June, which likely corresponded to the initiation of squaring. 13 Although cotton in the current study was planted later than normal, since it_ was delayed by cool, wet condi- tions, the other alternate host plants likely progress in the same proportions each year. Had the cotton been planted, for example, 30 days earlier, cotton fleahoppers could have moved from cudweed as well as from the eveningprimroses and horsemint into cotton. The move- ment into cotton would also have been more prolonged. The length of time a spring host plant is infested with cotton fleahoppers would be determined by existing weather conditions. Lack of adequate moisture would reduce the length of the infestation period by causing a decrease in plant moisture and flowering, forcing cotton fleahoppers to leave the host plant. The stage of develop- ment of cotton at this time would determine the source of cotton fleahoppers and the severity of the infestation. The time of cotton fleahopper emergence from overwintering eggs also could influence the source of cotton fleahoppers infesting cotton. With adequate soil moisture, the alternate host plants could develop, but without adequate precipitation, overwintered cotton fleahopper eggs would not hatch. Conditions favorable for overwintered egg hatch would determine the time of entry of cotton fleahoppers into the host sequence and in turn influence the infestation on cotton. Lack of rainfall could influence the abundance of alternate host plants and have a definite effect on the severity of the cotton fleahopper problem in cotton. Dur- ing the study, lack of rainfall was not a problem. There _ was an abundance of host plants; however, by the time cotton began squaring and was infested by cotton flea- hoppers, croton was the only wild host plant remaining attractive. Thus, after the initial movement into cotton, there was not a source of subsequent movement. Gaines (1933) reports similar circumstances in 1931. In early season, numbers of cotton fleahoppers infesting croton and caught on traps by Gaines remained rather steady until early Iune when they increased to a peak and then sharply declined. The numbers in cotton did not increase until shortly after the decline in numbers of cotton flea- hoppers on weed hosts, probably coinciding with the initiation of squaring. Gaines stated that damaging infes- tations on cotton did not develop. Although Gaines re- ported the numbers of cotton fleahoppers caught on traps around fields of croton, his records probably also reflect cotton fleahoppers associated with otheralternate host plants which occur in the same habitat as croton. Gaines’ data indicate that when there is a lack of alternate host plants, heavy infestations of cotton flea- hoppers in cotton can be expected, a conclusion sup- ported by Thomas and Owen (1937). The data of both the present and the Gaines’ studies indicate that when alter- nate host plants are available, the time of squaring of cotton determines from which host plants the cotton flea- hoppers will move into cotton and possibly the severity of the infestation. If initiation of squaring occurs before cotton fleahoppers leave the various alternate host plants, alternate host plants are competing for the cotton 14 fleahoppers, and a prolonged movement into c0 If be expected. When cotton initiates squaring at i; of the early season host sequence or when lack moisture forces the cotton fleahoppers to leave Q1 host plants, a sudden movement into cotton" f‘ likely, possibly resulting in heavy infestations in Literature Cited Almand, L. K., W. L. Sterling, and C. L. Green. 1975. A truck mounted aerial insect net. Tex. Agric. Exp. Stn. Anonymous. 1973. Suggestions for controlling cotton ins Texas, blacklands and Gulfcoast counties of Texas. Tex. 1 ,{_' Ser. L-218. " Coad, B. R. 1931. Insects captured by airplane are found at heights. USDA Yearb. 1931. p. 320-3. 1 - Correll, D. W., and M. C. Johnston. 1970. Manual of the; Plants of Texas. Texas Research Foundation. Renner, i Eddy, C. O. 1927. The cotton fleahopper. S. C. Agric. Exp. ‘_ 235. 21p. Ewing, K. P. 1927. Seasonal history and host plants of fleahopper. USDA mimeographed report. p. 22-29. Ewing, K. P., and E. E. Ivy. 1943. Some factors influencing? populations and damage. I. Econ. Entomol. 36:602-6. p]. F emald, Merritt Lyndon. 1950. Gray's Manual of Botany. Book Co. New York. 1632 p. . Fletcher, R. K. 1940. Certain host plants of the cotton -- M“ Econ. Entomol. 33z456-9. Gaines, I. C. 1933. A study of the cotton fleahopper with s -..i ence to the spring emergence, dispersal and populati 26963-71. — 1942. Effects of boll weevil control and cotton aphid control. Ibid. 351493-5. Gaines, I. C., and K. P. Ewing. 1938. The relation of wind indicated by balloon drifts, to cotton fleahopper dispe f; 31:674-7. Glick, P. A. 1939. The distribution of insects, spiders and mi air. USDA Tech. Bull. 673. 150 p. Hixon, Ephriam. 1941. The host relation of the cotton fleaho St. Coll. Iour. Sci. 17:66-8. Reeves, R. G., and D. C. Bain. 1947. Flora of south cen W. M. Welch Mfg. Co. Chicago. 298 p. Reinhard, H. I. 1926a. The cotton fleahopper. Tex. Agric. 339. 39p. —— 1926b. Control of the cotton fleahopper in Texas. Tex. y Stn. Cir. 40. 8 p. I — 1927. Control and spring emergence of the cotton fleaho -.f Agric. Exp. Stn. - 356. 32 p. ~ '$ . .4. . 6 ' i. .. Y» — 1928. Hibernation of the cotton fleahopper. Ibid. 377.1%, Ridgway, R. L., P. D. Lingren, C. B. Cowan, Ir., and I. W 1967. Populations of arthropod predators and H eliothis < _" applications of systemic insecticides on cotton. I. Econ. 601012-15. Schuster, Michael F., Clyde A. Richmond, I. C. Boling, and Graham. 1969. Host plants of the cotton fleahopper inj-f Grande Valley: phenology and hibernating quarto-Q 62.11269. ‘ Steere, William C. (ed.). 1969. Wild Flowers of the United S 3. Texas. McGraw-Hill Book Co. New York. 553 p. Sterling, Winfield, and F. W. Plapp, Ir. 1972. Insecticide‘- mortality studies on the cotton fleahopper. Tex. Agric. ' PR-3091. 9 p. I Thomas, F. L. 1936. Control of the cotton fleahopper. Tex. p sm. Circ. 77. s p. Thomas, F. L., and W. L. Owen, Ir. 1937. Cotton fleahi ecological problem. I. Econ. Entomol. 30:848-50. Acknowledgments This research was conducted in cooperation with Cotton Incorporated under Cooperative Agreement No. 170-75 and the Entomology Research Division, U.S.D.A. This publication was supported in part by the National Science Foundation and the Environmental Protection Agency, through a grant (NSF CB-34718) to the University of California. The findings, opinions, and recommendations expressed herein are those of the au- thor(s) and not necessarily those of the University of California, the National Science Foundation, or the Environmental Protection Agency. 15 All programs and information of The Texas Agricultural Experiment Station are available to everyone without regard to race, color, religion, sex, age, or national origin. The Texas Agricultural Experiment Station, J. E. Miller, Director, College Station, Texas 2.5M—6-76