This document is an initial attempt to characterize the saltmarsh vegetative community on the Rhode Island NWR Complex, based on 64 1-m2 plots spread across 11 transects in two refuge marshes. Only data representing 2015 growth are presented.

The objective of this document is to:

  1. Characterize and compare the saltmarsh vegetative communities of the Narrow River and Maidford Marshes.

Study area

Hover over a marker to obtain plot identification information. Use the mouse wheel to zoom more efficiently.

Community composition

We present a couple of approaches to visualize/characterize plant community composition. The first simply summarizes the proportion of occurrence of a given species on each plot (based on the 50 sample points within each plot). For simplicity, the data set was reduced to those species recorded on at least five (out of 64) plots. For similar reasons, we group living and dead Spartina alterniflora and Phragmites australis. The species are ordered in decreasing order of total cover on all plots.

This figure suggests, at least visually, that the Maidford marsh might have a more substantial high marsh/upland component and increased Phragmites encroachment.

We also calculate the Euclidean distance similarity index between the two marshes, the same metric used in the power analysis of James-Pirri et al. (2007, Wetlands Ecol Manage 15:335-345). This calculation suggests the two marshes are quite similar, or at least the sampled locations in each marsh are as variable within a marsh as they are between marshes. Average between-marsh distance measures (B0 = 0.973) were nearly identical to average within-marsh distance measures (W0 = 0.997; i.e., L0 = 0.976). Compare, for example, this similarity index to Figure 1 in James-Pirri et al. This analysis uses all documented cover types, not only those recorded on at least five plots. It does, however, group living and dead Spartina alterniflora and Phragmites australis.

We can also visualize how the number of plots sampled per marsh influences our estimate of this similarity index (based on 500 replications of each sample size; plots sampled without replacement).