Disrupting subsidies to multiple trophic levels shifts top-down and bottom-up control of food webs through time
With expanding human activities altering both food-web structure and energy flow within and between ecosystems, it’s increasingly important to understand how these influences affect food-web dynamics. Resource quality can be just as important as quantity, but the trophic level at which energy enters a food web may also be particularly influential. Moreover, the effects of resource removals/additions might change over time depending on their influence on behavioral and numerical responses, i.e. what consumers eat compared to how many consumers there are. To test the effects of terrestrial subsidies of primary-producers (terrestrial leaf detritus) compared to primary-consumers (terrestrial invertebrates) on food-web dynamics, we ran a long-term (532 day) mesocosm experiment crossing subsidy trophic level with predatory fish presence/absence to test for trophic cascades. We predicted that 1) subsidies that enter at primary-consumer level would have greater influence on trophic cascades shortly after manipulation due to a behavioral switch in top-predators and 2) subsidies that enter at primary-producer level would alter trophic cascades after multi-generational, numerical changes in consumers.
Subsidies of primary consumers to the predator trophic level altered trophic cascades seasonally, with positive top-down effects from predators to primary producers occurring in winter and early spring. The seasonal top-down control suggests subsidies that enter at primary-consumer levels exhibit short-term influence on food-web dynamics. The effect of subsidies of primary producers on trophic cascade strength increased through time: no trophic cascades were observed after 58 days, but strong cascades occurred after 532 days. This long-term trend indicates bottom-up, numerical changes in consumers through the food web, which created conditions for strong top-down effects. Interestingly, when food webs received subsidies to both predator and primary consumer trophic levels concurrently, there was a lack of strong top-down control through time, until day 532. Because the variable effects of subsidy trophic level make it difficult to predict food-web dynamics, such contingencies illustrate the importance of expanding our food-web concepts to include cross-ecosystem influences on resource supplies, such as invasive plant species or loss of native top-predators. In addition, examining longer temporal contexts to account for changes in both behavioral and numerical food-web responses will elucidate whether altering cross-ecosystem resources will have short- or long-term consequences for food webs.