Testing the high-tannin hypothesis with scatter-hoarding rodents: experimental and field evidence

With a combination of both experimental (caged, hours/daily) and field (free-ranging, several months) studies, we tested the high-tannin hypothesis that hoarding animals prefer to eat more low-tannin food items immediately but hoard more high-tannin items for later consumption. We studied two common rat species (Edward's long-tailed rat, Leopoldamys edwardsi; and chestnut rat, Niviventer fulvescens) and two nut species (Henry's chestnut, Castanea henryi; and cork oak, Quercus variabilis) that show varying tannin levels (0.6% versus 11.7%) but are similar in other traits. Based on the high-tannin hypothesis, we predicted that (1) both rat species would eat more low-tannin C. henryi nuts instantly but fewer high-tannin Q. variabilis nuts and (2) after harvesting a given nut, they would prefer to hoard more Q. variabilis nuts and fewer C. henryi nuts. The first prediction was firmly supported in our study: both rat species ate more C. henryi nuts than Q. variabilis nuts under all conditions. However, the second prediction was supported only in the field study, which lasted over several months, in contrast to the experiments conducted in seminatural enclosures (only 1 day). We found that high-tannin Q. variabilis nuts, in contrast to low-tannin C. henryi nuts, had a significantly higher probability of being hoarded and surviving as seedlings in the field. We conclude that experimental conditions used here are less likely to result in natural feeding preferences of tested animals and the high-tannin hypothesis is supported especially in the field setting.