From the agouti’s perspective, the whole point of burying a seed is to store it away so it can be eaten later in the year, when less food is available.  The problem is that all the local seed eating animals know this, and everyone is searching for these buried treasures.  By spreading the seeds out in scattered caches each agouti does their best to hide seeds in places unlikely to be  found by other animals.

Biologists studying this phenomenon have assumed that there is a cacher’s advantage – that the individual who buried the seed is more likely to find it again than some other animal because they remember the location.  However, this had never actually been tested, and our recent results on dramatically high cache removal made us second guess this assumption.

Michiel Veldhuis, a MSc student at the University of Groningen, tackled this problem by pretending he was an agouti.  Michiel went out to the forest and buried seeds, 700 in total.  But Michiel is not your average agouti, he’s got technology on his side (not to mention opposable thumbs and mad dance moves).  Michiel followed the fate of 200 of these seeds with tiny radio-transmitters to see how long it took for other animals to find them, and what they would do with them next.

Michiel embedded this experiment within the framework of our larger projects by creating these cache mimics in 19 known agouti territories.  In these areas we knew not only how much food was available, but also the local densities of agoutis, and other animals. This design allowed him to test for factors that affect seed pilferage.  Clearly, these other factors must be important, as the rates varied greatly across the territories (100-15% pilferage).  Michiel found that cache pilferage increased in areas with fewer fruits and more agoutis, i.e. where competition for food was higher, hidden seeds were more likely to be stolen.

Interestingly, not all stolen caches were eaten, rather, many were moved and reburied.  These aren’t hungry thieves, but individuals raiding each others food stores to prepare for the upcoming lean season.  Areas with fewer available fruits had seeds pilfered and re-cached multiple times, and carried further away.

Finally, the key test for the existence of a cacher’s advantage required comparing pilferage between his artificial caches (i.e., without owner) and natural agouti caches (i.e., with owner).  For this he teamed up with Veronica Zamora who was using the same methodology to track natural agouti caches.  Caches with an owner did have a significantly higher recovery rate than the mimicked caches without an owner. This difference reflects the a recovery advantage of the original cacher remembering the location of seeds over naïve foragers.

Michiel returned from his life mimicking agoutis in Panama to share these secrets of agoutis in his thesis defense last month in the Netherlands.  Congratulations Michiel!