Gray, L. J.; Mitchell, B. A.; Milner-Bradford, I. L.; Bell, G.; Revelly, K.; Keller, L.; Travers, C.

Zoo data collected by keepers while looking after endangered species are increasingly recognised as important scientific resources. In New Zealand, kiwi chicks (Apteryx spp.) are subject to the protective conservation programme Operation Nest Egg (ONE), during which growth, developmental and life history data are recorded. We have conducted comparative analyses on hand rearing records from 306 chicks, from Coromandel, Eastern and Western populations of Brown kiwi (A. mantelli), and rowi (A. rowi) and Haast tokoeka (A. australis Haast). We analysed chick responses to an old diet vs. a new kiwi diet, introduced in 2017. Both diets are fixed nutrient ratio mixtures. The old diet was high-protein, low-energy, while the new diet is high-fat, high-energy, with better micronutrient levels. We found Coromandel chicks, who originate from the environmentally variable K[u]aotunu Peninsula, grow the most efficiently overall on either diet, indicating they may be nutritional generalists. Western and Eastern chick growth efficiency was intermediate, while the South Island species grew the least efficiently on either diet. Rowi chicks developed the fastest overall on either diet, while Haast tokoeka chicks developed the most slowly, especially on the new diet. Rowi chicks therefore had to eat large volumes of either diet over a short time to maintain their rapid development, while Haast chicks were required to eat large volumes, especially of the new diet, over a protracted developmental period. This situation may have led South Island chicks to over-consume one or another diet component, with likely health consequences. Neither diet was obviously superior for chick wellbeing overall, though the new diet better supported chicks that needed hand feeding. This work demonstrates genetic populations of kiwi differ in their physiological responses to nutrition. As ONE is ongoing, tailored diets for chicks from each genetic group should be developed, and we present methods to achieve this. In our life history trait analyses, we found chick starting size (hatch mass) did not significantly influence growth efficiency across kiwi genetic groups, nor did chick sex. We identified that chicks malpositioned as embryos were more likely to require extended periods of hand feeding, and that Eastern males produce more malpositioned embryos than other populations. Our study shows that effective zoo records can be used to improve captive care, to stimulate future research to refine species management practices, and to explore fundamental questions of life history evolution in wild and captive populations.