Model ecosystems (e.g. microcosms, mesocosms, small scale field experiments) are critical for testing hypotheses in ecology. Indeed, certain ecological questions such as the consequences of invasion/extinction and ecosystem response to disturbance are often impractical or even unethical to pursue in large, natural ecosystems. Yet insights gained from model ecosystems are often criticized as being irrelevant for natural ecosystems due to their artificiality and lack of complexity. We argue that because model ecosystems can experimentally remove complexity - for example by reducing or eliminating diversity, immigration, predators or abiotic fluctuations - they can help identify critical factors influencing population, community, and ecosystem dynamics in nature. Here we compare plankton and ecosystem dynamics observed in natural ponds in Michigan to two different model pond ecosystems: 1) enclosures placed within a natural pond, and 2) mesocosms constructed from cattle tanks maintained outdoors. For both model and natural pond ecosystems, we monitored plankton and ecosystem properties repeatedly over a field season in order to capture community and ecosystem dynamics.
Results/Conclusions

Our results indicate that zooplankton diversity in model ecosystems can be experimentally controlled, and can be maintained at levels similar to natural pond ecosystems when immigration of zooplankton from a regional species pool is allowed. Zooplankton variability (CV) can vary significantly between model ecosystems, but the variability observed in both types of model ecosystem is comparable to that observed in natural ponds, indicating model ecosystems adequately capture the range observed in natural ponds. By comparing results obtained within a model ecosystem with results across model types and natural ponds, we are also able to identify possible factors in complex natural ecosystems that are influencing community and ecosystem dynamics. These factors include species diversity, species richness, and the degree of abiotic fluctuation. We argue that model ecosystems remain invaluable for providing insights about the complex dynamics observed in natural ecosystems.