Isotopic Evidence for Spatial and Temporal Changes in Everglades Food Web Structure

Bryan E. Bemis, Carol Kendall, and Scott D. Wankel

U.S. Geological Survey, Menlo Park, California, USA

Ted Lange

Florida Fish and Wildlife Conservation Commission, Eustis, Florida, USA

David P. Krabbenhoft

U.S. Geological Survey, Madison, Wisconsin, USA

Trophic structure within a food web is often implicated as a control on how mercury is distributed and transferred throughout an aquatic ecosystem. Methyl mercury bioaccumulates in the food web, with higher concentrations typically found in tissues of organisms that occupy higher trophic positions. Reducing mercury contamination in the ecosystem is a major focus of restoration efforts in the Everglades. Therefore, knowledge about Everglades trophic structure is critical for making management decisions about how to effectively minimize mercury concentrations in biota.

The nitrogen (15N) and carbon (13C) isotopic composition of tissues is a tracer of diet that can be used in many aquatic ecosystems to distinguish the relative trophic positions of organisms. We use measured 15N and 13C values of Everglades biota to investigate spatial and temporal variability in food web structure. Plants, invertebrates, and fish were collected from 16 well-studied ACME (Aquatic Cycling of Mercury in the Everglades) sites during 1995-1999 as part of a collaboration between the USGS and the Florida Fish and Wildlife Conservation Commission (FFWCC). These collections provided more than 350 site-date sampling groups.

Within this massive data set, we focus on several sites with multiple collection periods during 1996-1998 that have a sufficient number of organisms to investigate spatial and temporal differences in food web structure. The organisms analyzed for tissue 15N and 13C represent a broad range of trophic positions within the food web, from primary producers to top-tier predator fish. 15N and 13C values of primary producers reflect those of the water chemistry, modified by isotopic fractionations associated with nutrient uptake and growth. The isotopic compositions of invertebrates and fish are integrated measures of diet, with an expected relative enrichment of the heavier isotope (i.e., higher 15N/14N and 13C/12C) at each trophic level.

We use 15N versus 13C plots to identify relative trophic positions of biota within the food web for each site and collection date. Based on laboratory and field studies, the expected pattern of trophic enrichment is increasing 15N and 13C toward higher trophic positions (e.g., site U3, September 1997) (fig. 1). The

Figure 1. 15N versus 13C at site U3 during September 1997. The 15N:13C slope for invertebrates and fish is +1.3 (n = 12; p = 0.02; R2adj = 0.37). Symbols are median values.

Figure 2. 15N versus 13C at site F1 during September 1997. The 15N:13C slope for invertebrates and fish is –1.5 (n = 6; p = 0.03; R2adj = 0.66). Symbols are median values.

magnitude of 15N and 13C ranges varies among sites and collection dates. This is not entirely surprising, given the large range of consumer-diet isotopic fractionations reported for different species (e.g., McCutchan, 1999). However, the statistically significant, yet different,15N:13C slopes for the biota analyzed in each site-date group suggest that processes influencing consumer-diet isotopic fractionations within a food web are spatially and temporally unique. The 15N:13C slope differences likely indicate spatial and temporal differences in the food web base and/or complexity of trophic interactions.

15N:13C slopes are typically positive, as expected (fig. 1). However, rare negative slopes (e.g., site F1, September 1997) suggest the unlikely possibility that consumers are relatively depleted in the heavier isotope of carbon than their diets (fig. 2). Although this pattern has been observed for individual species in controlled growth experiments (e.g., DeNiro and Epstein, 1978), an alternative explanation for the multiple-species pattern plotted here is that the biota isotopes reflect the presence of multiple food web bases with distinct 13C and 15N values. For example, if mosquitofish, bluefin killifish, least killifish, and shrimp are components of one food chain at site F1 and golden topminnow and gator flea utilize another food web base (fig. 2), then this could explain the appearance of a single, “reversed” trophic hierarchy in carbon isotope space.

References

DeNiro, M.J. and Epstein, S., 1978. Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta, 42: 495-506.

McCutchan, J.H., Jr., 1999. Carbon Sources for Macroinvertebrates in St. Vrain Creek, Colorado. Ph.D. Thesis, University of Colorado, Boulder, 300 pp.

Bryan, Bemis, U.S. Geological Survey, 345 Middlefield Road, MS 434, Menlo Park, CA, 94025, Phone: 650-329-5603, Fax: 650-329-5590,