Multiple levels of ecological organization determine susceptibility to critical transitions
Regime shifts or critical transitions are sudden changes between distinct community and ecosystem states caused by gradual environmental changes and/or stochastic perturbations. These abrupt, large-scale changes occur in nearly all habitats and have wide-ranging effects on ecosystem services; therefore, it is essential to identify the characteristics that make ecological systems prone to regime shifts. Multiple levels of ecological organization (species traits, community composition, and ecosystem properties) likely interact to determine the presence of alternative regimes. Freshwater lakes are well-known for their susceptibility to regime shifts, which are often defined by the dominant primary producers and driven by eutrophication. Shifts between submerged aquatic vegetation (SAV) and unrooted, floating plant (FP) regimes are driven by changes in nutrient levels and are an ideal system for studying critical transitions. To determine how multiple levels of ecological organization determine the presence of this critical transition, I analyzed a publically available dataset of 193 Connecticut (CT) lakes, I re-surveyed 14 of those CT lakes and 14 lakes on Long Island, NY (LI), and I conducted laboratory experiments to quantify trait differences (redundancy vs. complimentarity) of FP species, with respect to nutrient stoichiometry (nitrogen: phosphorus) and temperature.
The floating plant regime was uncommon (8.6 and 42.8% of waterbodies with FP present, in CT and LI respectively). Ecosystem size appeared to limit the occurrence of this regime shift (FP regimes did not occur in ponds > 3.5 ha.) and large waterbodies remained in a SAV regime, even at high nutrient levels. Transitions showed reversibility in both directions across years (shifts from FP to SAV regimes, and vice versa). In laboratory experiments, FP species had complementary growth rates across nutrient stoichiometry and temperature. At high nutrient levels and 30 °C, Spirodela polyrhiza had the fastest growth rate, while Wolffia brasiliensis had the highest growth rates at warmer temperatures. Other species (Lemna minor and L. valdiviana) maintained relatively high growth rates at low nutrients. Based on observed trait differences, I expected to find a positive relationship between FP species richness and the occurrence of a FP regime in surveyed ponds. In CT, more species-rich FP assemblages were positively associated with FP regimes, but this pattern was not seen in LI. Although FP species show trait complimentarity in the laboratory, factors on other levels of ecological organization (e.g., ecosystem size) may outweigh its importance for determining the presence of critical transitions.