Estimating natural resource harvests: conjectures?

Estimates of natural resource harvests often inform rural conservation and development strategies. Retrospective household surveys remain one of the most commonly employed methods for estimating harvests. Pair-wise comparisons of estimates from household surveys versus diary records were performed for household harvests in the Brazilian Amazon. Although diaries and surveys produce similar estimates of mean economic value for different product groups, 33% of product-level estimates showed a three-fold difference between methods with no consistent patterns in discrepancy direction. Significant differences in estimates for highly valued products (cash crops, game animals, and fish) together with higher respondent confidence in diaries may undermine household models based exclusively on surveys.