Publications Archive

Abstract: House mice (Mus musculus) have proven to be the most difficult introduced mammal to eradicate from (and keep out of) New Zealand reserves and sanctuaries. Partly as a consequence of this, little is known about how bird communities respond to mouse eradication. Mice were successfully eradicated from 217 ha Mana Island Scientific Reserve, near Wellington, in 1989–90. Five-minute bird count surveys undertaken in spring and autumn before and after mouse eradication revealed that 13 of 22 species were recorded significantly more often after mouse eradication, and just two species were recorded significantly less often following the eradication (and each of these in one only of the two seasons that were compared). Four species had no significant change, and three species showed mixed responses between the two seasons. While the overall pattern was of increased relative bird abundance after mouse eradication, there is limited information on why individual bird species increased during the study period, and whether this was a consequence of mouse eradication. Bird count data revealed that insectivorous passerines may have benefited the most from mouse eradication on Mana Island, suggesting that competition for invertebrate prey was the main impact that mice had on the birds of the island. The use of anticoagulant rodenticides to eradicate mice from Mana Island had little detectable impact on populations of the island’s birds.

Abstract: Male and female adult Floreana mockingbird (Mimus trifasciatus) have monomorphic plumage features that make them impossible to sex in the field. In this study, we use discriminant function analysis (DFA), a widely used technique, to assess the best measures to determine sex. We measured six morphological characteristics (mass, beak depth, beak width, tarsus length, wing length, and head-beak length) for birds of known sex (determined by molecular techniques) from the two extant populations of M. trifasciatus on Champion and Gardner islets, within the Galápagos archipelago. Using a coefficient of sexual dimorphism, we found that males are significantly larger than females in three of the variables. Discriminant functions using wing length and a combination of wing length + mass, and wing length + tarsus length could classify birds with a 98% level of accuracy. Furthermore, we were able to estimate a robust cut-off point to determine the sex of individuals in the field through a decision tree, using only wing length as morphological variable. Fast and accurate sexing of the bird based on one variable will reduce handling times and minimise stress for captured birds.