LIBRARY February 1971 NOV 5 1975 Texas A&M University 3 \<§\\ m“ i\\v\‘\\\@‘$\‘q w“ A w” BIOLOGICAL C OIVT R OL OF RHODESGRASS SCALE IN TEXAS B Y NEODUSME T IA SANG WANI (RA Effectiveness and Colonization Studies TEXAS AGM UNIVERSITY TEXAS AGRICULTURAL EXPERIMENT STATION H. O. Kankel, Acting Director, College Station, Texas [Blank Page in Original Bulletin] CONTENTS ..................................................................... __ 4 a ____________________________________________________________________ __ 5 ies .............................................................. .- 5 ..... 7 k ..................................................................... -- 8 i” ion Regulation of esgrass Scale ............................................ __ 8 i trol and Parasite Activity .................... --l0 ntrol and Yield Response ...................... __l0 V Experiments on Ranges ........................ __1l ‘l; ass Pasture Longevity ............................ --l1 N. sangwani Longevity ............................ ..l1 i; of N. sangwani Females ...................... ._l2 i Spread in Grasslands .............................. __l2 ; Establishment as Influenced by ber of Females Released Month of Release .................................... __l3 "wani Colony Establishment as uenced by Lift; §tages of ‘sites at TimesifofRelease ........................ __l3 - rea Distribution ........................................ “l3 i Coalescence ................................................ ..1?> ....................................................................... __1l} ent ............................................................ _l5 ited .............................................................. __ 1 5 SUMMAR Y Investigations from 1961 to 1968 demonstrated the effectiveness of the parasite, Neodusmetia san- gwani (Rao), as a controlling agent of rhodesgrass scale, Antonina graminis (Maskell). This internal parasite reduced scale populations 68.0 percent dur- ing the year. The two normal scale population peaks were reduced by 50 percent. The yearly mean para- sitism varied from 28.1 to 34.6 percent. Parasitized scales produced 93.7 percent fewer crawlers on rhodes- grass and 90.1 percent fewer crawlers on paragrass. Ant control did not affect the effectiveness of the parasite under range conditions. N. sangwani was successful in eliminating yield losses due to scale damage on a paspalum-fringed sig- nalgrass site. This range control was substantiated with model experiments in greenhouses with rhodes- grass, fringed signalgrass, Texasgrass and cane sour- grass. Longevity studies at three biological control locations showed that rhodesgrass was not killed by scales even under severe grazing conditions. Females of N. sangwani were found to have a short lifespan even under the most favorable condi- tions. Experiments showed that release of adults or distribution of parasitized scales (allowing adults to emerge) gave about equal results in establishment of colonies. The best month for release was August when 76.9 percent of the colonies became established. From 100 to 200 females were required per release site to insure 64 percent establishment. Colony spread was found to occur at the rate of one-half mile per year in grasslands with normal scale populations. The female moved 6 feet or less by means of hopping and crawling in her life-time; however, wind transport was found to be important during periods of peak parasite activity in July- August and October. Females were caught at heights of 6 feet with sticky traps. Techniques developed for distribution of N. sangwani on a large scale from naturally infested rhodesgrass scale were as follows. The grass was cut- off just below the ground with hoes and placed in boxes. The whole grass plant was then dropped from a pickup truck at points along predetermined lines. The potential number of parasites per stem. was esti- mated by scale dissection. Release of one colony per mile was made, and colony spread and coalescence was evaluated in a pasture near Encino. The study showed that under the conditions of this test, 4 years would be required for the parasites to invade 83 percent of the range. The spread was restricted in saline areas which supported only ‘Spartina sp., a plant that is not a host of rhodesgrass scale. 4 and Colonization Studio’! scale, Antonina graminis (Mas- found in Texas in 1942 attacking ' gayana Kunth (Chada and Wood time this scale insect has been found species of grasses in North America 00d 1960 and Schuster 1967). Intro- scale parasite, Neodusmetia sangwani India was made in 1959 (Dean, et al. ' of the parasite as a control agent scale was begun in 1961. This paper l data and parasite-colony ecology 'tions. Previous Studies scale was first described from grass China, as Sphaeroco-ccus graminis It was later described from India by ‘as Antonina indica. The scale has been Africa, Australia, Brazil, Bermuda Is- Island, Ceylon, Colombia, Cuba, East or, Formosa, Guatemala, Hawaii, Johnston Island, Kwajalein (Marshalls), infested by rhodesgrass scale in Texas. I ICAL CONTROL 0F RHODESGRASS SCALE B Y NEOD USMETIA SANG WANI (RA o).- Michael F. Schuster and C. Boling* Madagascar, Mariana Islands, Puerto Rico, South China, Sumatra, United States, Venezuela and West Pakistan. The distribution within the United States was given by Chada and Wood (1960). These authors reported 62 counties infested by the scale in Texas. Areas in Texas known to be infested are shown in Figure 1. The number of recorded hosts of the scale is in- creasing each year. Chada and Wood (1960) recorded 69 host species in the United States; Brimbelcrombe (1966), 14 in Australia; Guagliumi (1963), 22 in Venezuela; and Williams and Schuster (in press), 86 in Brazil. The large host range and wide distribution indicate that forage losses on ranges might be great under climatic conditions favorable for the scale. Chada and Wood (1960) reported that berrnudagrass, Cynodon‘ dactylon Pers., and rhodesgrass, C. gay- ana, were the only forage grasses affected by these scales in Texas. Schuster (1967) found that yields of 38 range grasses were significantly reduced by scales in greenhouse tests (Table 1). Interviews with ranchers in Brooks, Kenedy, Wil- lacy, Kleberg and Duval counties indicated that the grazing capacity of native ranges had been reduced approximately 30 percent since the introduction of rhodesgrass scale into Texas in the early 194015. This heavy loss has not been regained, presumably due to scale infestation of native grasses. The life history of the scale was described by Chada and Wood (1960). The adult scale is parthen- ogenetic and reproduces ovoviviparously. The crawl- ers are positively thigmotropic and attach themselves to the plant nodes under the leaf sheath. The legs are lost at molting and the second and third instar larvae are saclike and resemble the adult. There are five generations annually in southern Texas. About 85.4 percent of the scales are found below the first plant node on the crown node. Adults lived up to 6 weeks without food. Dissimination of individuals was mainly by crawling or wind. qlespectively, assistant professor and technician I, Texas A8¢M University Agricultural Research and Extension Center at Weslaco. TABLE 1. REDUCTION IN YIELD AND PERCENT PLANT MORTALITY RESULTING FROM RHODESGRASS SCALE INFESTATION ON GRASSES IN A GREENHOUSE TEST‘ Yield Plants loss, killed, Grass percent percent Cane Sourgrass 26.1 42.1 Hybrid sourgrass 32.4 42.5 Longspike silver bluestem 28.5 23.6 Silver bluestem 18.2 85.0 Wright Threeawn 8.4 0.0 Red gramma 25.4 74.4 Fringed signalgrass 44.0 Buffel sandbur 18.8 55.0 Coast sandbur 37.8 42.5 Big sandbur 10.2 52.0 Fringed windmillgrass 86.9 Hooded windmillgrass 48.8 Rhodesgrass 18.0 49.3 Nash windmillgrass 12.7 0.0 Shortspike windmillgrass 49.4 87.5 Berrnudagrass 59.1 82.5 Arizona Cottontop (glabrous sp.) 88.3 85.0 Arizona Cottontop (pilos sp.) 28.6 0.0 Texas Cottontop 36.4 83.3 Plains lovegrass 36.4 0.0 Mourning lovegrass 17.4 0.0 Stinkgrass 20.1 0.0 Red lovegrass 28.7 6.2 Tumble lovegrass 48.7 65.6 Sand lovegrass 24.3 0.0 Green Sprangletop 29.0 55.0 Filly panicum 51.6 18.4 Halls panicgrass 34.7 26.5 Natalgrass 26.4 0.0 Knotroot bristlegrass 63.4 85.0 HBK bristlegrass 0.0 0.0 Southwestern bristlegrass 15.6 0.0 Texas bristlegrass 12.4 0.0 Hooked bristlegrass 18.6 90.3 Johnsongrass 38.0 0.0 Sand dropseed 29.1 0.0 Two flowered trichloris 20.6 17.5 Four flowered trichloris 32.4 0.0 Texasgrass 25.9 0.0 ‘ From Schuster, 1967. Early control measures were attempted with chemicals (Wene and Riherd 1950; Richardson 1953; Chada and Wood 1960); however, Chada and Wood (1960) pointed out that insecticides were too costly and are useless under range conditions. Attempts to control the scale biologically were begun in 1949, when the parasite, Anagyrus antoninae Timberlake, was introduced from Hawaii (Riherd 1950). In 1954 and 1955 several parasites were intro~ duced from France (Dean and Schuster 1958), but no establishment was obtained with Xanthoencyrtus phragmitis Ferr., Boucekiella antoninae (Ferr.), Tim- berlakia europaea (Mercet), and Anagyrus diversi- comis Mercet. Dean and Schuster (1958) concluded that the effect of A. antoninae on rhodesgrass scale populations was of little value. The parasite was not able to withstand high temperatures and low humidi- ties of Texas, except in certain ecologically modified areas around lakes and canals in the Lower Rio 6 Grande Valley. The lowest parasite activity occu during periods of highest scale populations. t j The parasite, N. sangwani, was first found at ing rhodesgrass scale near Delhi and Bangalore, I 1 by Narayanan, et al. (1957) and was originally scribed as Dusmetia sangwani by Rao (1957). I duction and establishment of N. sangwani in T was made in 1959 (Dean, et._al. 1961). Shipmen the parasite from India to tlhe Insect Identific and Parasite Introduction Laboratories at n town, N. ]., was made by Rao (1965). Dean, e (1961) described the method of rearing the pa for releases on range sites. Colonies establishe Dean, et al. (1961) were the source of parasiteséi duced into Arizona, California, Bermuda Isl Mexico and Brazil. Methods of laboratory re and distribution of individuals are described by’ chado da Costa, et al. (in press). Schuster (1965) studied the biology of the pa ‘i under laboratory conditions. At 30° C, 17 to 20: were required to complete a life cycle; at 20° 6 to 56 days were required. Later studies indicated some individuals would complete a life cycle in ' 35 days at 20° C, indicating that 20° C is ne threshold temperature of arrested development. y pupal period was extended by low temperatur the laboratory. Apparently the parasite over ' as pupae near Delhi, India, in this mannerg( . yanan, et al. 1957). Preimaginal stages inclul caudate first instar, a hymenopteriform secon star, a similar but nonfeeding prepupa, and a v Reproduction by unfertilized females resulted in progeny. The short-lived brachypterous femal an average of 6.2 eggs in each of 5.7 scales an duced an average of 35.3 offspring. The sex ra field-collected parasites ranged from 6.7 to 7.8 f q to 1 male. The male has functional wings. Fi, shows a female ovipositing in a scale. °°°¢°°*- 'son of biological control areas with check out parasites) proved to be the best meth- instrating scale control. The first 3 years showed a 44}_2percent greater yield from al control area than from the check area y Population regulation data showed that reduced 52.6 percent, with an average .1 of 73.9 percent (Table 9). Following _. sangwani into the check area in 1965, 23A t difference in yield was found between biological control and check plots for the next. 3 years of clipping. Similarly, slightly greater yields were obtained in check areas than in biological con- trol areas as was found in Texasgrass and cane sour- grass model experiments (Tables 7 and 8). Rhodesgrass Pasture Longevity Chada and Wood (1960) stated that when rhodes- grass infested with rhodesgrass scale were grazed they died within 3 years. Similar observations were re- corded in Queensland, Australia, and fertilization and controlled grazing were recommended to restore vigor (Anon. 1940). In general, pure grass stands were conducive to wide fluctuations of scale populations. As scale popu- lations increased, severely damaged culms died, and new culms were produced. Finally the entire plant died. At certain times greater scale numbers were found on grass in the biological control area when compared to the check area. This resulted indirectly since scale control by the parasite resulted in more robust plants and greater areas on the plants for scale development. But parasite activity soon reduced scale peaks. The average scale populations for the Kleberg and Kenedy county test areas are shown in Table 2. In all cases, plants in the check areas began to show considerable damage by the end of the second year and were dead at the end of the third year. Plants in the bio-control areas were still alive and producing good yields at the end of 5 years when observations stopped. Female N. sangwani Longevity The number of female parasites surviving for 24 hours at l0, 15, 20 and 30 degrees centigrade was 50.7, 40.7, 43.2 and 40.9 percent respectively (Table 10). Survival of females was best at 10° C, at which 41.3 percent survived for 48 hours. Since summer TABLE l0. THE INFLUENCE OF TEMPERATURE ON THE LONGEVITY OF FEMALE N. SANGWANI Percent alive at end of time indicated in hr. Temperature °C 24 48 72 96 124 10 50.7 41.3 16.0 0 15 40.7 29.6 9.3 7.4 3.7 20 43 .2 18.5 1.0 0 30 40.9 0 ll TABLE ll. NUMBER OF FEMALE N. SANGWANI CAUGHT AT TWO HEIGHTS ON 18-INCH SQUARE TANGLEFOOT TRAPS, 1965 Females caught 0n trap Weslaco Donna Period 3 ft 4 ft 3 ft Total June 4-21 0 0 0 O July 2-30 ll 0 0 ll Aug 6-27 3 0 0 3 Sept 3-24 3 l 0 4 Oct 1-25 0 9 1 l0 Nov 12-26 0 1 0 1 Dec 3-17 0 0 0 0 Total l7 ll l 29 temperatures reach 35° to 40° C on Texas ranges, the life of the female parasite must be very short under range conditions. There was only slight mobility of the female at 15° C and none at 10° C. Feeding studies with l percent sucrose, royal jelly, honey and 50 percent royal jelly to 5O percent honey at 20° C showed no increased parasite survival over the nonfeeding check insects at 24, 48 and 72 hours. Neither food supplements nor temperature modification appeared t0 be practical in reducing parasite mortality sufficiently for a mass distribution program. Zl/Iobility of N. sangwani Females Weekly inspection of sticky traps showed that females were caught at 3 and 4 feet above the ground during 1965 (Table ll). During 1966 females were caught 6 feet above the ground (Table 12). Parasites were caught in greatest numbers during July and August, with a lesser peak in October. These peaks coincide with periods of greatest parasitism of scales by N. sangwani. Rhodesgrass plants support 2 to 4 times greater scale numbers than most grasses. Assuming that an equal percentage of scales were parasitized during a period of time, it can be seen that rhodesgrass pas- tures would produce greater numbers of individual parasites; hence, greater numbers were trapped in a rhodesgrass pasture at Elsa. Equal size plates, placed TABLE 12. FEMALE N. SANGWANI CAUGHT ON l8-INCH SQUARE STICKY TRAPS AT VARIOUS HEIGHTS i. FOUR LOCATIONS IN SOUTH TEXAS, 1966 on the ground during the same period of tim August, often trapped 150 to 200 individuals. Females coated with fluorescent powder an leased in a lawn were found trapped 2 days later the sticky traps at distances up to 6 feet from n of release. This lawn grass was 2 inches in he and was almost free of rhodesgrass scale. The A was repeated in a rhodesgrassgppasture which aver six scales per node and 65 percent parasitizag None of the 1,500 females marked and released recovered although several hundred unmarked o sites were captured on the traps. Apparently f_- on the rhodesgrass did not have to search as aéti because 0f the existing scale population and j failed to reach the traps. Activity of marked females was closely ob A with the aid of ultra-violet light. Females releas the grass crown moved rapidly upward in sear suitable scales. Females were found at the ape meter high) of rhodesgrass plants within a few utes of release; then individuals reversed the se V ing pattern and proceeded down the plant. A hopping by the female was infrequent and appe to be an avoidance reaction. Wind caused som males to become dislodged from plants. These data suggest that wind dispersion of brachypterous female may be greater than by i ing and hopping. The greatest colony increase‘. during the summer months and airborne move was greatest in grasslands with large scale populat Colony Spread in Grasslands At Armstrong, N. sangwani colony spread the rate of about one-half mile per year in a past ; rhodesgrass and sandburgrass. However, the sp through a saline area in which Spartina sp. (a host species) was growing was less than one-half in 4 years and in the direction with the preva wind. At Encino in a paspalum-fringed signal site, the parasite spread 0.4 miles each year f years. Near Three Rivers, colony spread in berm grass and sandburgrass along a highway was - one-half mile each year for 3 years; however, the a site was found 21 miles downstream on the N Weslaco 3 ft Donna Period 1 ft 6 ft l ft 3 ft 6ft Encino Elsa Sh lfi 6fi O9 7-91 r-r 1 ft 6 ft Tota Jan 1-_]uly 1 July 4-27 Aug 2-29 Sept 6-27 Oct 7-24 Nov l-l4 Nov 14-Dec 31 Total OOOCGQQO OOOOQOOQ OOGQQOGO l-‘OOCWQOO QOCQOOCDO QOOQOOQO NJOQGOIQOO >—'©O©©©'—'© OOOCCOQQ EOOUI©IOUIO L>9©©N>©~$© IOOQGCOIQC l2 V e same period of time. Apparently, flood- grass clones with scales and parasites down 3- annel. Thus, parasite spread downstream i important means of distribution. In a ‘_ ~: pasture near Kingsville the rate of spread ‘rd mile per year for 21/2 years. These f indicate that colony spread is greatest gds which provide the most contiguous scale ablishment as Influenced by 10f Females Released, p- h of Release fdata in Table 13 show that as the number . released per site increased, the percentage establishment increased; however, 100 per- lishment was never attained. The release .200 females established colonies 64 percent e. The best month for release was August, i’ percent of the colonies were established. ttime was the November-December period percent of the released colonies became An average of 53.4 percent of the total v leased became established. i - sites for release of parasites were chosen '; rd to vegetation, scale population, tem- l; precipitation. These factors obviously n to low colony establishment in the above ites were collected each morning between 1». .m. and delivery time to range sites varied i’ 8 hours. Many females were probably un- iposit before death. These factors would be in any large-scale release program. Dis- adults does not appear practical unless in- "can be collected periodically during the day mo quickly. i COLONY ESTABLISHMENT AS INFLUENCED _." OF FEMALE N. SANGWANI AND PERIOD E ON RANGE SITES IN SOUTH TEXAS No. of females ‘Ales released Colony establishment, ._point, range ‘Z, , 100 41.8 I 200 64.0 + 90.0 ’ ge (133 points) 53.4 Period of release Colony establishment % 76.9 55.7 53.1 35.7 ‘rage (126 points) 51.6 TABLE 14. COLONY ESTABLISHMENT OF N. SANGWAN1 FOLLOWING RELEASE OF 150 FEMALES OR PREIMAGI- NAL STAGES ON RANGE SITES AND FREQUENCY OF COLONY DETECTION FOLLOWING RELEASE IN AUG- UST, 1966, ENCINO Established colonies detected, % Sample date (interval) preimaginal adult November 2, 1966 (3 mo) 21.7 18.2 January 12, 1967 5 mo) 39.1 31.8 May 19, 1967 (9 mo) 47.8 50.0 N. sangwani Colony Establishment as Influenced by Life Stages of Parasites At Time of Release The data in Table 14 show that there were only slight differences in the number of established col- onies resulting from the release of adults or pre- imaginal stages while still within rhodesgrass scale. There was a definite increase in the number of col- onies detected with time. The results of this test did not differ greatly from data on the number of sites established in the above test. The results indicate that the distribution of para- sites, while still within the scale bodies, would be as effective as the release of adults. Furthermore, the expense would be far less than if a laboratory culture had to be maintained to produce the adult numbers required for mass distribution. Also, the problem of the short life of the female parasite would be over- come as the parasite would be transported in the rela- tive hearty preimaginal stage and emergence of the adult would be under more natural conditions. Large Area Distribution A test was conducted in October, 1963, to deter- mine procedures and approximate cost involved in collecting and distributing parasites from a natural population. Scales and parasites were collected from a rhodesgrass pasture near Armstrong by cutting off the grass just below the ground with a hoe and transporting the vegetation in boxes to desired re- lease sites. Previous dissections indicated that l0 stems would be required at each release point to produce 100 female parasites. It required 16 hours of labor and 124 miles of driving with a %-ton truck to infest the 43-square mile pasture. The cost was about $1.28 per square mile. Grass stems ready for distri- butio-n are shown in Figure 5. Colony Coalescence The frequency of colony encounter for periods of time following release near Encino are shown in Table 15. There was a rapid increase in colony encounter for the first 2 years. Thereafter, the fre- quency of encounter increased slowly and at the end of 41/2 years 83 percent of the sample sites were in- vaded by the parasite. About 10 percent of the sam- ple sites were in saline areas, predominated by Spar- 13 logical control of insects attacking the 10w val grasslands offer.s opportunities in biological con that are not fully realized. As yet there is little qu titative information on the damage done by ins under normal or outbreak conditions. Sparse att tio-n has been paid to the damage of grasshop (Anon. 1962). App (1962) discussed various did and indirect results o-f insectaattack on pastures, Anderson (1961) evaluated iflosses caused b-y ; i‘ hoppers on grasslands. It is possible that range e ogists are not fully cognizant of the damage to ra a lands resulting from insects; hence, little effort a been directed toward their control. i" q Grasslands provide a relative stable environm for the operatio-n of biological control organisms. Ll (1960) suggested that biological agents have been ces.sful mainly where host plants are trees or o perennials maintaining their geographical locati over long periods of time. Turnbull and Ch (1961), however, questioned Lloyd’s hypothesis proposed that suitability for biological agents W, determined by the type of damage inflicted by an_ sect, the amount of such damage that can be r ated and pest populatio-n density required to intolerable damage. In general, perennials can I stand certain types of damage, such as defoliat better than annuals. The amount of damage to lands can only be measured as pounds of grass vested from the final animal end~product, suc meat or woo-l. Thus, the category of direct pest they proposed, does not exist in grasslands. The nition, which by directly attacking produce, destr significant part of its value, cannot occur in grassl _ fo-r even a blade of grass 20 percent eaten by a p, piller still retains 8O percent feeding value to Figure 5. Grass stems separated and prepared for distribution - . . . , _ animal. t to release points as descrlbed 1n the text. Parasites W111 emerge from scale on the grass and oviposit in scale at the release site. The 10w value 0f grasslands also increases attractive atmosphere for the utilization of biolo agents. Rhodesgrass scale fitted the category o; indirect pest which required large numbers 0v’ protracted period of time to cause damage. Che The daia iiidieaied iiiai ‘i 91' mere Years ma)’ be control was economically feasible only on lawns iediiiied idi ediiiPieie edveiage di ranges Where N- golf greens. Clearly, it lent itself well for biol sangwani were released at a density of one colony Control effort ~ per square mile. Greater scale numbers. and a more contiguous scale population through the year would probably reduce the time required for complete tina sp.; this accounted for the incomplete parasite coverage of the area. The rapid control experienced with Neodus i sangwani, while several other parasites failed to i lish themselves even with repeated introductions, Coviirage ports C1ause-n’s (1951) view that a fully Discussion parasite or predator is always easily and quickly a, This is believed to be the first successful intro- lished and thus control is usual in three host ge duction of a predator or parasite for the control of tions, or 3 years. Substantial host reductions an insect attacking grass plants on ranges. The bio- obtained in the first year. TABLE l5. FREQUENCY OF N. SANGWANI COLONY ENCOUNTER AS DETERMINED BY LINE TRANSECT SAMP) OF AN AREA NEAR ENCINO RELEASED AT A DENSITY OF ONE COLONY PER SQUARE MILE, OCTOBER 1962 - Time from release, mo 12 21 23 29 31 39 43 47‘ Colony frequency,% 4 20 33 40 ’ 52 62 68 76 I ‘Based on incomplete sample. 14 p; ani reduced its host primarily by pre- roduction of its parasitized host. The . ction of tot.al scale population was always the percent parasitism. For this reason, i-parasitism as an index of scale contro-l is Yield comparison between treatments was dicative, and laboratory data proved as data. Acknowled ent iation is expressed to the King Ranch, eir support of this project by grants-in-aid ‘r e. a Literature Cited . N. L. 1961. Seasonal losses in rangeland iondue to grasshoppers. Econ. Entomol. " 78. 1940. The felted grass-co-ccid. Queens- gr. 54:398. k; 196-2. Basic problems and techniques in research. National Academy of Science- j:- Research Council. Publication n. 890. '_wT1$-J-n€-F€I3"1 1962. Pasture insects. In Pa.sture and research techniques. Cornell Univ. Press. 191. R. 1961. The influence of ants upon es, predators and scale insects. Entomol. J er. Ann. 54:543-51. Vi be, A. R. 1966. The occurrence of the ilAntonina (Homoptera: Coccoidae) in gland. Entomol. Soc. Queensland. 5:5-6. 1957. A new parasite of the rhodes-grass Hymenoptera Encyrtidae). Bul. Broo-klyn _1. Soc. 52:l24-27. and E. A. Wood, Jr. 1960. Biology and of the rhodes-grass scale. U.S.D.A. Tech. 221. P. 1951. The time factor in biological l. Econ. Entomol. 44: 1-9. y. and M. F. Schuster. 1958. Biological of rhodes-grass scale in Texas. Econ. ‘o1. 51 :363-66. ., M. F. Schuster and C. Bailey. 1961. itroduction and establishment of Dusmetia ni on Antonina. graminis in So-uth Texas. . Entom.ol. 54:952-54. F. 1963. Predation by parasitic Hyme- - , the basis of ant-induced outbreaks of a Species. Econ. Entomol. 56:116. 7E. 1908. Remarks on Indian scale insects dae) Pt. 3. DefitfAgr. Mem., Indian Ent. j, pp. 15-46. , P. 1963. Insecto-s y Arachnidos de las comunes de Venezuela. DIPUVEN. i. 1964. On the European species of tia Mercet, and a new Oriental genus Hym., Chalcidoidae, Encyrtidae). Entomophaga. 9:75-9. Lloyd, C. C. 1960. Significance of the types of host plant crop in successful biological control o-f in- sect pests. Nature. 187:430-31. Machado da Costa, J., R. N. Williams and M. F. Schuster. In press. Cochonilha dos Capins, “An- tonina graminis” no Brasil — Introducae de Neo- dusmetia sangwani, inimigo natural da Cochoxil- ha. Pesquisa Agropecuaria Bra.silerina. 1969. Marrow, 1959. The relationship of soils, precipi- tation, phosphorus fertilization and livestock grazing to vegetational composition and forage production of native vegetation on {the Encino Division of the King Ranch. PhD Dissertation. Texas A8cM University. Mas-kell. W. M. 1897. On a collection of coccidae principally from China and Japan. Ent. Monthly Mag. 33z239-44. Narayanan, E. S., B. R. S. Rao, and H. S. Sangwan. 1957. New species of the parasites o-f the rhodes- gra.ss scale fro-m the Indian Union. Indian Entomo-l. 19:65-66. Nord, E. C. 1956. The influence of drought, fertiliza- tion and clipping on native range vegetation in south Texas. PhD Dissertation. Texas A8cM University. Rao, B. R. S. 1957. Some new species of Indian Hymenoptera. Pro-c. Indian Acad. Sci. 46z385-90. Rao, V. P. 1965. Shipment of Dusmetia. sangwani Subba Rao, a parasite of the rhodes grass scale to the U.S.A. Commonwealth Inst. of Bio. Control. Tec. Bul. 5. Richardson, B. H. 1953. Insecticidal control of rhodes grass scale on St. Augustine grass lawns. J. Econ. Entomol. 461426-30. Riherd, P. T. 1950. Biolo-gical notes on Anaygrus antoninae Timberlake (Hymenoptera-Encyrtidae) and its host Antonina gmminis (Maskell) (Ho- moptera-Coccidae). The Florida Entomol. 33:18- 22. Schuster, M. F. 1965. Studies on the biology of Dusmetia sangwani (Hymenoptera-Encyrtidae). Entomol. So-c. America. Ann. 58:272-75. Schuster, M. F. 1967. Response of forage grasses to rhodesgrass scale. Range Management. 20:307- 9. Turnbull, A. L. and D. A. Chant. 1961. The practice and theory of biological control on insects in Canada. Canad. Zool. 39:697-753. Wene, G. P. and P. T. Riherd. 1950. Oil emulsion to control rhodes grass scale. Econ. Entomol. 43:386. Williams, R. N. and M. F. Schuster. In press. Co- -chonilha dos capins (Antonina gmminis) no Brasil. Distribuicao E. Plantas Hospldeiras. O Biologico. 1969. 15 Texas Agricultural Experiment Station ‘it Texas A8cM University -————- College Station, Texas 77843 H. O. Kunkel, Acting Director— Publication ' v I POSTAGE ~= United Sums Dcp