-tSKiss^ssaa 



381 



745. , and Sterlln;?, W. L. 1973. Inferences on the dispersion of cotton 

 arthropods in Texas. Environ. Entomol. 2: 863-867. 



Discrete frequency distributions obtained from samples of various arthropods 

 commonly found in Tex^'m cotton fields were fitted to 6 discrete mathematical 

 frequency-distribution models. The observed frequency distributions most 

 frequently fit the ne;',ative binomial expectations, suggesting a clumped 

 pattern of dispersion. These same arthropods were seldom observed to be 

 randomly dispersed as indicated by poor agreement with the Poisson expec- 

 tations. 



746. , and Sterling, W. L. 1973. Sampling techniques for cotton arthropods 

 in Texas. Tex. Ajjric. Exp. Stn. Misc. Publ. MP-1110, 8 pp. 



The D-Vac® sampling nethod generally provided a better estimate of arthropod 

 abundance than visual In situ whole plant examination for mcst cotton arthropods 

 in studies in 1971-72. Exceptions included the sampling of Heliothis spp. larvae 

 and eggs, boll weevils and plant parts. For these units, the visual whole plant 

 examination method was preferred. Three sample unit sizes were compared during 

 1972: a) single plant, b) 1-meter row of plants and c) 13 feet of row of plants. 

 Arthropods collected by the D-Vac® sampler had similar relative variation values 

 for the single-plant and l-meter-row-of-pla"uts sample unit sizes. Larger rela- 

 t?rv5~ Veiriation values resulted for D-Vac® sampling with a 13-foot^row-sample 

 unit size. (The lower the relative variation the more acceptable the sampling 

 technique). The estlr-ite of the number per acre was consistently larger for 

 the single-plant sample unit size. For visual whole plant sampling of boll 

 weevils, Heliothis spp. larvae and eggs, and damaged fruit, the 1-meter-row- 

 of-plants sample unit size yielded the lowest relative variation values. The 

 D-Vac® sampling method was preferred over the sweep-net method for collecting 

 cotton fleahoppers. 





