Abstract
This study of customary harvests of sooty shearwater Puffinus griseus chicks by Rakiura Maori compares the utility of Traditional Ecological Knowledge (TEK) and ecological science for understanding patterns in prey availability. We recorded TEK of 28 muttonbirders about emergence patterns and variation in chick size at different aspects of 14 breeding islands and in their coastal fringe compared to inland areas. Spatial and temporal variation of chick availability were measured using the methods of ecological science in the 2001 harvest season across Putauhinu Island, south west of Rakiura, New Zealand. As predicted by TEK, titi emerged earlier from west than east coast locations on Putauhinu. Scientific measures were also consistent with an earlier emergence in coastal compared to inland areas as asserted by TEK, but conclusive inference is potentially confounded by movement of chicks between burrows just before fledging. A TEK construct predicting heavier chicks on the west coast was not supported by scientific measurements. We also measured the characteristics of areas preferred for harvesting so that we could gauge representativeness of the areas “sampled” by the muttonbirders to accumulate their TEK. Within forested habitat, areas harvested by muttonbirders had 62–65% higher chick density than unharvested areas. The muttonbirders concentrated harvesting where there was less ground cover and taller canopy cover and only hunted on nights and times of the season when harvesting was most profitable. Therefore, TEK may be less able to detect wider‐scale variation and harvest impacts on prey in particular. Short runs of scientific information from spatially and temporally stratified sampling will complement and assist inference from longer term TEK. As shown in this case study, TEK and science often agree on pattern, but are likely to disagree on why a pattern exists.