Predator Diet and Trophic Position Modified with Altered Habitat Morphology

Abstract

Empirical patterns that emerge from an examination of food webs over gradients of environmental variation can help to predict the implications of anthropogenic disturbance on ecosystems. This "dynamic food web approach" is rarely applied at the coastal margin where aquatic and terrestrial systems are coupled and human development activities are often concentrated. We propose a simple model of ghost crab (Ocypode quadrata) feeding that predicts changing dominant prey (Emerita talpoida, Talorchestia sp., Donax variablis) along a gradient of beach morphology and test this model using a suite of 16 beaches along the Florida, USA coast. Assessment of beaches included quantification of morphological features (width, sediments, slope), macrophyte wrack, macro-invertebrate prey and active ghost crab burrows. Stable isotope analysis of carbon ((13)C/(12)C) and nitrogen ((15)N/(14)N) and the SIAR mixing model were used to determine dietary composition of ghost crabs at each beach. The variation in habitat conditions displayed with increasing beach width was accompanied by quantifiable shifts in ghost crab diet and trophic position. Patterns of ghost crab diet were consistent with differences recorded across the beach width gradient with respect to the availability of preferred micro-habitats of principal macro-invertebrate prey. Values obtained for trophic position also suggests that the generalist ghost crab assembles and augments its diet in fundamentally different ways as habitat morphology varies across a highly dynamic ecosystem. Our results offer support for a functional response in the trophic architecture of a common food web compartment (ghost crabs, macro-invertebrate prey) across well-known beach morphologies. More importantly, our "dynamic food web approach" serves as a basis for evaluating how globally wide-spread sandy beach ecosystems should respond to a variety of anthropogenic impacts including beach grooming, beach re-nourishment, introduction of non-native or feral predators and human traffic on beaches.

Fig 1. Schematic of relative changes in beach morphology with increasing width (x-axis, one order of magnitude) and decreasing slope (y-axis, one order of magnitude) including the distribution of micro-habitat compartments (sub-tidal, swash, inter-tidal, supra-tidal zones), macrophytes (wrackline), principal macro-invertebrate primary consumers and relative abundance (mole crab—Emerita talpoida, amphipod—Talorchestia spp., coquina calm—Donax variablis) and associated relative swash sediment size (coarse, medium, fine) and water levels (MHW = mean high water, MLW = mean low water).

Mean individual prey sizes range from 1.0 (amphipod) to 2.5 (mole crab) cm in length. The mole crab (E. talpoida) passively filters particulate organics from receding waves in the swash zone and maximizes feeding efficiency at exposed beaches characterized by large waves, steep slopes and strong wave surge [57, 79, 80]. The coquina clam (D. variabilis) is also found in the swash zone but appears to prefer lower wave energy sites while actively filtering organic particulates [57, 79, 80]. Semi-terrestrial amphipods (Talorchestia spp.) occupy the low supra-tidal areas of sandy beaches in strong association with the spray zone and damp stranded macrophyte materials where they can feed and shelter [57, 84].

Fig 2. A working model reflecting relative measures of resource availability/contribution to diets (y-axis) of a secondary consumer (ghost crab, Ocypode quadrata) with increasing beach width (x-axis).

As beach width increases the preferred habitat conditions for the principle macro-infaunal prey (A = mole crab—Emerita talpoida, B = amphipod—Talorchestia spp., C = coquina calm—Donax variablis) shift from high energy, coarse sediment beaches preferred by mole crabs (A) to low energy, finer sediment beaches preferred by coquina clams (C). At intermediate beach width the accumulation of damp macrophyte wrack (D) provides both shelter and food for semi-terrestrial amphipods (B), thereby supporting a food subsidy for ghost crabs. Abundance of Ocypode quadrata (E) may be highest at intermediate width beaches if multiple resources are highly accessible for the consumer. Arrows indicate energy flow direction only; relative uses of prey resources (A, B, C) are expected to fluctuate across beach width. Aquatic-derived particulate organic material provides primary resources to mole crab and coquina clams. TL = trophic level.

Fig 3. Geographic distribution of barrier island beach sites (N = 16, S1 Table) sampled during this study (2010–2012).

S.P. = State Park, N.S. = National Seashore.

Fig 4. Mean density (+/- se) for A) Mole crab (Emerita talpoida); B) macrophyte wrack; C) Coquina clam (Donax variablis), and D) Ghost crab (Ocypode quadrata) in relation to beach width (Fig 2 and S1 Table).

Fig 5. Variation in proportion of three principal components of ghost crab diets calculated from Stable Isotope Analysis in R (SIAR) across the range of beach width (see S1 Table).

Prey component coding as follows: Coquina clam (Donax variablis) white;, Amphipod (Talorchestia spp.) grey;, Mole crab (Emerita talpoida) black. Ghost crab isotopic signatures were assessed from muscle tissue collections (N = 3–7 crabs/site) between May 2010 and July 2012.

Fig 6. Relationship between beach width (S1 Table) and proportions of dietary components (Fig 5 and S2 Table) for Ghost crabs (Ocypode quadrata) determined using Stable Isotope Analysis in R (SIAR).

(A) Mole crab (Emerita talpoida); (B) Amphipod (Talorchestia spp.); and (C) Coquina clam (Donax variablis),. (D) Log ₁₀-transformed mean density of ghost crab (O. quadrata) in relation to beach width. The relationships between amphipods as a dietary component and Ghost crab density with beach width were modeled using a Gaussian distribution, y = [-0.5 ((x—x_a)/(b))²] (Table 1). Regression coefficient (R²) and level of significance (P-value) are indicated for each relationship.

Fig 7

Relationship between: (A) ghost crab trophic position and beach width and B-D) ghost crab trophic position and dietary proportion of three principal prey items (S2 Table). Principal prey are: (B) Mole crab, Emerita talpoida; (C) Amphipod, Talorchestia spp. (data point for SGSP excluded from regression, see results and discussion); and (D) Coquina clam, Donax variablis. The dietary component of E. talpoida was modeled using an exponential saturating hyperbola, y = a (1-e^-bx). Regression coefficient (R²) and level of significance (P-value) are indicated for each relationship.