Ecosystems are knit together by interactions among species, such as competition, herbivory, and predation. One of Darwin’s many interesting ideas was that these interactions are stronger and therefore more important in tropical and lowland ecosystems. This idea is at the heart of some key predictions in ecology and evolution. More intense interactions are thought to accelerate adaptive diversification in the topics and to frequently set the low-elevation and low-latitude edges of species’ distributions. For plants, some biologists think that more intense attacks on foliage (herbivory) explains the impressive array of chemical defenses in tropical leaves, while high consumption of seeds (seed predation) near parent trees helps maintain high tree diversity in tropical forests.
Despite long-standing interest in these predictions, evidence for large-scale patterns in the intensity of species interactions was, until recently, cobbled together from many small-scale studies using different methods. Perhaps unsurprisingly, these studies often arrived at different conclusions. A better way to test for latitudinal and elevation patterns is to conduct the same experiment up many mountains that cover the full latitudinal scale of interest: from the Arctic to the equator.
The goal of our study, which we affectionately call the B.I.G. (Biotic Interactions Gradients) experiment, was to test for geographic patterns in the intensity of species interactions using a simple, standardized experiment conducted at a continent-wide scale.
Video: nocturnal rodent eats seeds at tropical site, Colombia. J Muñoz
But which interaction to test? We needed one that is biologically important and standardizable. For example: if herbivory is chronically intense in an ecosystem, we would expect plants to evolve strategies to combat it, like thorns or toxins. To detect an underlying gradient in herbivory, we would need to use the same leaves at all sites to control for evolutionary responses. But imagine mailing identical spinach leaves to collaborators from the Arctic to equator: they would be slime before they arrived. Happily, there is one interaction that is life-or-death, that shapes plant communities and thus whole ecosystems, that lends itself to standardized experiments, and that has been central to speculation regarding stronger interactions in the tropics: seed predation.
To test whether interactions are more intense toward low latitudes and elevations, our team of 17 biologists from seven countries set out more than 50, 000 seeds along 18 mountains from Alaska to Ecuador. We used sunflower and oat seeds, agricultural species with few defenses and high nutritional value that should tempt any wild seed eater. We placed seeds in 30 small piles at each site, with 4 to 5 sites up each mountain, and checked them 24 hours later to see how many were eaten. As we knew temporal variation would probably be high (and it was), we ran the experiment multiple times at each latitude.
We found that after only 24 hours, seed predation was significantly higher in the tropics than the temperate zone. Overall, seed predation increased by 17% from the arctic to equator and by 17% from 4000 meters elevation (high in the Andes) to sea level. For seeds that take a while to germinate, we estimate that protection by snow at high latitudes could as much as double the steepness of the latitudinal pattern.
In addition to testing patterns in overall seed predation, we were interested in which groups of animals might be driving high predation at some sites. By excluding vertebrates from seeds using small cages, we also showed that high predation in the tropics and lowlands is caused mostly by insects and other invertebrates. Thus, a relatively underappreciated group of animals likely plays an outsized role in the ecology and evolution of our most biodiverse ecosystems.
These findings are described in the article entitled Seed predation increases from the Arctic to the Equator and from high to low elevations (DOI: 10.1126/sciadv.aau4403), recently published in the journal Science Advances.