Restoration is one of the major issues in ecological science as highlighted by the UN Decade on Ecosystem Restoration. Alarming reports of wetlands loss over the past 50 years has made wetland conservation a public policy priority leading to a proliferation of restoration programs. However, post-restoration monitoring rarely exceeds 5-10 years limiting an in-depth understanding of the processes critical to restoration success.

We investigated the influence of hydrologic restoration and plant introductions on restored vegetation convergence toward the desired reference community, for a restoration program initiated in 1998 in southern Colorado, USA. This restoration focused on 3 plant communities in 12 wetlands monitored at N+5, N+15, and N+25 years post-restoration. Using an approach based on beta regression models, we assessed the role of hydrologic similarity of a restoration site to the desired reference community and target species planting density in achieving and predicting vegetation similarity of a restored vegetation to the desired reference community at N+5, N+15 and N+25 years after restoration.

Achieving a high hydrologic similarity to the desired reference produced better results with less stochasticity while increase planting density produced better results but did not reduce stochasticity. Predictions of vegetation similarity are possible for some plant communities using beta regression models with hydrologic similarity to the desired reference and target species planting density as explanatory variables.

In conclusion, restoring proper hydrologic conditions using reference-based standards is critical to ensure restored plant communities convergence toward desired reference. When planning wetland restoration programs, managers should: 1 - Define appropriate reference-based standards. 2 - Focus on hydrologic restoration. 3 - Supplement with appropriate target species planting.