Hydrological‐niche models predict water plant functional group distributions in diverse wetland types: Fid-Bau Portal

Hydrological‐niche models predict water plant functional group distributions in diverse wetland types

Deane, David C. / Nicol, Jason M. / Gehrig, Susan L. / Harding, Claire / Aldridge, Kane T. / Goodman, Abigail M. / Brookes, Justin D.

Abstract

Human use of water resources threatens environmental water supplies. If resource managers are to develop policies that avoid unacceptable ecological impacts, some means to predict ecosystem response to changes in water availability is necessary. This is difficult to achieve at spatial scales relevant for water resource management because of the high natural variability in ecosystem hydrology and ecology. Water plant functional groups classify species with similar hydrological niche preferences together, allowing a qualitative means to generalize community responses to changes in hydrology. We tested the potential for functional groups in making quantitative prediction of water plant functional group distributions across diverse wetland types over a large geographical extent. We sampled wetlands covering a broad range of hydrogeomorphic and salinity conditions in South Australia, collecting both hydrological and floristic data from 687 quadrats across 28 wetland hydrological gradients. We built hydrological‐niche models for eight water plant functional groups using a range of candidate models combining different surface inundation metrics. We then tested the predictive performance of top‐ranked individual and averaged models for each functional group. Cross validation showed that models achieved acceptable predictive performance, with correct classification rates in the range 0.68–0.95. Model predictions can be made at any spatial scale that hydrological data are available and could be implemented in a geographical information system. We show the response of water plant functional groups to inundation is consistent enough across diverse wetland types to quantify the probability of hydrological impacts over regional spatial scales.