ADULT ATTRACTIVENESS AND OVIPOSITION PREFERENCE OF Bemisia tabaci ( GENN . ) ( Homoptera : Aleyrodidae ) B-BIOTYPE IN COTTON GENOTYPES

The silverleaf whitefly Bemisia tabaci B-biotype is an important pest of cotton; it affects plant vigour, transmits geminivirus and reduces lint quality. In order to evaluate the resistance of cotton genotypes, Gossypium hirsutum (L.), to the whitefly Bemisia tabaci B-biotype, both free-choice and nochoice attractiveness and non-preference for oviposition tests were carried out in a shade house, at room temperature. Low attractiveness to adults of this whitefly was observed for plants of genotypes Fabrika, CNPA Ita 90, Makina, Coodetec 407, and IAC 01-639 CPA 02-24, which may represent a resistance component of these genetic materials to the insect. Genotypes BRS Aroeira, Coodetec 406, Fabrika, and Coodetec 401 presented the non-preference-for-oviposition type of resistance in the free-choice and nochoice tests. The numbers of trichomes and gossypol glands per cm were not suitable to evaluate nonpreference for oviposition of whitefly adults on cotton genotypes.


INTRODUCTION
For the control of the whitefly Bemisia tabaci (Genn.)B-biotype, plant resistance should be used as an additional control tactic within the integrated pest management strategy in order to minimize the damages caused by this insect (Norman Jr. et al., 1996).Such resistance can be characterized by the occurrence of plant morphological and/or physiological traits that make them less or more preferred for feeding and ovi-position by the insects.Among the morphological traits, trichome density (McAuslane, 1996) and plant architecture can be underlined as important (Sippel et al., 1987).Also leaf pubescence in cotton has been reported as a very important morphological trait for oviposition preference.This trait was studied by Butler Jr. & Henneberry (1984), who verified that the number of B. tabaci adults on glabrous-leaved cotton genotypes was smaller than in semi-glabrous or pubescent plants.With regard to the presence of chemical substances, the concentration of gossypol (Butter et al., 1990), sugars, tannin, o-dihydroxyphenol (Butter et al., 1990), as well as plant nutrition (Bentz et al., 1995) have been highlighted as the most important biochemical characteristics.Gossypol glands on the internodes favor B. tabaci populations, while their occurrence on the leaf area is considered detrimental (Butter & Vir, 1989).
Considering the lack of data about effective control methods and the importance of B. tabaci B-biotype in cotton, we attempted to study the attractiveness and non-preference for oviposition of this whitefly for cotton genotypes as mechanisms involved in the resistance.

MATERIAL AND METHODS
The experiments were carried out in a shade house at room temperature.Attractiveness/repellency and non-preference for oviposition of the whitefly were evaluated for the following cotton genotypes PR CNPA Ita 90,Coodetec 406,BRS Aroeira,MG 405,Fabrika,BRS Ipê,Fibermax 986,Coodetec 407,Makina,Shrow Grow 618,Coodetec 401,Delta Opal,CNPA Acala I,and BRS Itaúba.Whitefly rearing -Adults were acquired from colonies in Campinas, São Paulo State, Brazil, previously identified as B. tabaci B-biotype by silvering induction in pumpkin leaves.The insects were multiplied (stock rearing) in a shade house, on cabbage (Brassica oleraceae var.capitata) and drunkard's dream plants (Euphorbia pulcherrima Willd).New plants were introduced every fifteen days to replace old plants already weakened by the high whitefly population present during rearing.
Attractiveness test -A split-plot design was used, with plots arranged in randomized blocks, in which the 20 genotypes represented treatments, with six replicates.Twenty-day-old cotton plants were submitted to artificial infestation with 100 non-sexed B. tabaci B-biotype adults per plant.Adult whitefly counts per genotype were determined 24, 48, and 72 hours after infestation.These counts were performed at sunset, on the underside of the first completely developed leaf from the apex, when the adults showed little mobility.
Non-Preference for oviposition tests -A randomized block design with 20 treatments and six replicates was used in the free-choice and no-choice tests.Twenty-day-old cotton plants were submitted to arti-ficial infestation with 100 non-sexed B. tabaci B-biotype adults per plant.In the no-choice test, a cylindrical iron cage, measuring 0.40 m in diameter by 0.60 m in height was used in each pot; the cage was wrapped with voile fabric to protect the plant before the release of whitefly adults.The first completely developed leaf from the apex of each plant was collected three days after infestation.Upon identification, the leaves were taken to the laboratory to count the number of eggs per cm 2 .A 1cm 2 template and a stereoscopic microscope were used to count the eggs.The trichome and gossypol gland densities per cm 2 on the abaxial surface were also evaluated in the free-choice test.
The values obtained for each variable were submitted to analysis of variance and test F. When significant, the means were compared by Tukey test, at 5%.

RESULTS AND DISCUSSION
Attractiveness test -No differences were observed among the cotton genotypes evaluated with regard to the mean number of B. tabaci B-biotype adults during each evaluation period (F = 0.99 ns ); (Table 1).One day after infestation, the number of whiteflies in the genotypes varied from 4.5 to 55.8 adults per plant.In this period, the genotypes CNPA Acala I and IAC 20-233 had higher number of adults than 50, while Coodetec 407 attracted less than 5.The variation in number of whiteflies per plant among cotton genotypes (10 to 58.8 and 5.3 to 82.2 adults per plant), observed two and three days after infestation, respectively, indicates that factors of a morphological and/or chemical nature present in the plants accounted for greater or smaller whitefly attractiveness.
A higher number of adults was found in CNPA Acala I, while IAC 01-639 CPA 02-24, Coodetec 407, Shrow Grow 618, and Makina had less.Differences among cotton genotypes were observed for the mean number of B. tabaci B-biotype adults during the three evaluation periods (F=3.03**).Genotype CNPA Acala I proved to be the most attractive, and genotypes Fabrika, CNPA Ita 90, Makina, Coodetec 407, and IAC 01-639 CPA 02-24 were the least attractive.When the means for the adult populations at 24, 48, and 72 h are compared in the horizontal direction (Table 1), they underwent non-significant variation, indicating that the plant/adult whitefly relationships during the evaluations did not interfere with the attractiveness test results.
With reference to gossypol (Table 2), the number of glands ranged from 78.0 per cm 2 for genotype IAC 01-639 CPA 02-24 to 136.8 glands per cm 2 for genotype Shrow Grow.In general, the results do not allow a correlation to be established between number of eggs per cm 2 and number of trichomes per cm 2 (r = 0.09), number of eggs per cm 2 and gossypol glands per cm 2 (r = 0.19), and number of trichomes per cm 2 and gossypol glands per cm 2 (r = 0.11) in the free-choice test.Butter et al. (1990) found a negative correlation (r = -0.45) between gossypol content and the egg density of B. tabaci A-biotype, while Butter & Vir (1989) reported a positive correlation (r = 0.53) between number of B. tabaci A-biotype adults and number of gossypol glands on the internodes of the main stem of cotton plants.
SE = Standard error of the mean; Original and analysis data were transformed to (x + 0.50) 1/2 for ANOVA; Means followed by the same letter do not differ by Tukey test, (P = 0.05).**denotes the F value is significantly at P = 0.01.

Table 2 -
Mean (± SE) number of Bemisia tabaci B-biotype eggs, mean number of trichomes, and mean number of gossypol glands for 20 cotton (G.hirsutum) genotypes, in a free-choice test.Jaboticabal-SP, 2004.The low attractiveness of plants from genotypes Fabrika, CNPA Ita 90, Makina, Coodetec 407, and IAC 01-639 CPA 02-24 to B. tabaci B-biotype may represent a resistance component of these genetic materials to the B. tabaci B-biotype whitefly.Genotypes BRS Aroeira, Coodetec 406, Fabrika, and Coodetec 401 presented the non-preference-for-oviposition type of resistance to whitefly in the free-choice and no-choice tests.The numbers of trichomes and gossypol glands per cm 2 were not suitable to evaluate non-preference for oviposition of whitefly adults on cotton genotypes.
SE = Standard error of the mean; Original and statistical analysis data were transformed to (x + 0.50) 1/2 for ANOVA; Means followed by the same letter do not differ by Tukey test (P = 0.05).**denotes the F value is significantly at P = 0.01.found lower B. tabaci B-biotype oviposition preference for genotypes BRS Aroeira and Fabrika.

Table 3 -
Mean (± SE) number of Bemisia tabaci B-biotype eggs for 20 cotton (G.hirsutum) genotypes in a no-choice test.Jaboticabal-SP, 2004.SE = Standard error of the mean; Original and statistical analysis data were transformed to (x + 0.50) 1/2 for ANOVA; Means followed by the same letter do not differ by Tukey test (P = 0.05).**denotes the F value is significantly at P = 0.01.