Burrowing crabs and physical factors hasten marsh recovery at panne edges

Abstract

Salt marsh loss is projected to increase as sea-level rise accelerates with global climate change. Salt marsh loss occurs along both lateral creek and channel edges and in the marsh interior, when pannes expand and coalesce. Often, edge loss is attributed to erosive processes whereas dieback in the marsh interior is linked to excessive inundation or deposition of wrack, but remains poorly understood. We conducted a two-year field investigation in a central California estuary to identify key factors associated with panne contraction or expansion. Our study explored how an abundant burrowing crab, shown to have strong negative effects on marsh biomass near creek edges, affects panne dynamics. We also explored which physical panne attributes best predicted their dynamics. To our knowledge, ours is the first study of panne dynamics in a California marsh, despite how ubiquitous pannes are as a feature of marshes in the region and how often extensive marsh dieback occurs via panne expansion. Overall, we found that pannes contracted during the study period, but with variable rates of marsh recovery across pannes. Our model incorporating both physical and biological factors explained 86% of the variation in panne contraction. The model revealed a positive effect of crab activity, sediment accretion, and a composite of depth and elevation on panne contraction, and a negative effect of panne size and distance to nearest panne. The positive crab effects detected in pannes contrast with negative effects we detected near creek edges in a previous study, highlighting the context-dependence of top-down and bioturbation effects in marshes. As global change continues and the magnitude and frequency of disturbances increases, understanding the dynamics of marsh loss in the marsh interior as well as creek banks will be critical for the management of these coastal habitats.

Fig 1. Study design from the landscape to plot scale.

(A) Study map showing panne (yellow circles) and creek edge (red circles) studies courtesy of the U.S. Department of Agriculture, Farm Service Agency with topo-enhanced NAIP. Examples of the panne study and variation in panne size, pooling, and drainage, pre (B-D) and post (E) flashing installation (See Supplementary Information).

Fig 2. Partial leverage plots for all of the best-fit model effects.

Plotted is the movement of marsh into the panne area (panne contraction along marsh-panne boundary) in cm. The dotted red horizontal line represents the average marsh-panne boundary movement from 2016–2018 of 16.298 cm. Each partial leverage plot includes the 95% C.I. For a list of the parameters included in the plotted indices, see Table 1.

Fig 3. Conceptual model of panne contraction and possible mechanisms.

Bolded arrows indicate relationships or main model terms that correlated to panne contraction and were shown in our study. Narrow arrows indicate pathways that were not demonstrated by our study but are likely to play a role based on marsh dynamics literature. Red arrows and cells indicate pathways to slower panne contraction and green cells indicate pathways to relatively rapid panne contraction. Blue cells indicate potential mechanisms.

Fig 4. Comparison of crab CPUE and burrow density along panne versus creek edges.

(A) Crab CPUE along panne versus creek edges. (B) Burrow density (# per m²) along panne versus creek edges. The data are plotted as points overlaid on top of boxplots. Pink points and boxplots represent panne edges and blue points and boxplots represent creek edges. Different letters denote significant differences (α = 0.05) between panne and creek edges.