RESUMEN
The aim of this paper is to examine the statistical relevance of bird species on the endpoints of avian long-term toxicity studies (eggs laid, eggs set, eggs hatching, embryo survivor, 14-day old survivors and eggshell thickness). Data from 561 animals of three different species (Colinus virginianus, Anas platyrynchos and Coturnix coturnix japonica) tested with five different pesticides were analyzed in this study. The substances considered were: Thiamethoxam (EZ-3-(2-chloro-1,3-thiazol-5-ylmethyl)-5-methyl-1,3,5-oxadiazinan-4-ylidene(nitro)amine), Thiacloprid ((Z)-3-(6-chloro-3-pyridylmethyl)-1,3-thiazolidin-2-ylidenecyanamide), Acetamiprid ((E)-N(1)-[(6-chloro-3-pyridyl)methyl]-N(2)-cyano-N(1)-methylacetamidine), Phosmet (O,O-dimethyl S-phthalimidomethyl phosphorodithioate) and Dicofol (2,2,2-trichloro-1,1-bis(4-chlorophenyl)ethanol). Several general lineal mixed models were conducted to evaluate the factors affecting variables used in long-term reproductive toxicity tests. Test significance was p < 0.01 in all models tested. Model R(2) value was high (0.80) for all variables except for eggs laid (R(2) = 0.42) for the three species studied. Tukey studentized range test showed significant differences among species and pesticides. For pre-hatching period the differences were significant for eggs laid and eggs set among species. C. japonica showed statistical differences for egg hatching. With respect to embryo survivor and 14 days old survival, significant different were found for C.virginianus and A. platyrynchos, respectively. These results indicate that the selected species have an influence in the endpoints to be used for risk assessment.