Do similar restoration treatments produce similar communities?

Background/Question/Methods

The objective of restoration treatments is to produce a specific community type.   To achieve this objective we must know that a set of restoration treatments will consistently produce the same outcome.  The restoration of wetland ecosystems often involves re-establishment of hydrologic function.  Two ditched and farmed wetlands in central Wisconsin were restored in 2009 (Lost Creek, 72.5 ha)  and 2010 (Moses Creek, 6.5 ha) with the same set of treatments: clearing, removal of topsoil, lowering  the soil level, re-channelization, re-application of ~30cm of topsoil, and sowing of a wetland mixture (sedges, grasses, forbs).  The two sites were inventoried in September,  2013 (Lost Crk) and 2012 (Moses Crk) with transects.  Five spatially separate areas were sampled at Lost Crk (Areas A-E), whereas the transects at Moses Crk were systematically arranged throughout the floodplain.   Due to previously documented patterns, the transect at Moses Crk were split into a northern and southern group.  The research hypothesis was that difference among areas [in composition and structure] would be found, and the magnitude of the differences would be greater at Lost Crk.  This hypothesis was explored and tested with cluster analysis, ordination and MRPP.  The three  dominant species in each area was determined to assess structure.

Results/Conclusions

The cluster analysis and ordination indicated at least 3 distinct groups in the different areas,  but Moses Crk North and South were not different.  Four of the five areas at Lost Crk clustered, and all Lost Crk communities separated from the community at Moses Crk.  The MRPP test found a significant difference among areas (p < 0.001).  Analyses revealed that spatial proximity in the large wetland did not consistently results in greater similarity, and I conclude that the same set of treatments does not result in the same community within or between sites.    The dominant species clearly indicate why Moses Crk stands out from Lost Crk; none of the dominants at Moses Crk showed up as dominant in any area at Lost Crk.  Though Lost Crk has more internal variation in composition and structure than Moses Crk, the  magnitude is small give that Lost Crk is ~ 10x larger than Moses Crk.  An observational assessment suggests that the surrounding land use, through its effect on propagule input, partially explains the differences between sites.