D. Albrey Arrington

Getting to the fat of the matter: models, methods and assumptions for dealing with lipids in stable isotope analyses

Within an organism, lipids are depleted in 13C relative to proteins and carbohydrates (more negative δ13C), and variation in lipid content among organisms or among tissue types has the potential to introduce considerable bias into stable isotope analyses that use δ13C. Despite the potential for introduced error, there is no consensus on the need to account for lipids in stable isotope analyses. Here we address two questions: (1) If and when is it important to account for the effects of variation in lipid content on δ13C? (2) If it is important, which method(s) are reliable and robust for dealing with lipid variation? We evaluated the reliability of direct chemical extraction, which physically removes lipids from samples, and mathematical normalization, which uses the carbon-to-nitrogen (C:N) ratio of a sample to normalize δ13C after analysis by measuring the lipid content, the C:N ratio, and the effect of lipid content on δ13C (Δδ13C) of plants and animals with a wide range of lipid contents. For animals, we found strong relationships between C:N and lipid content, between lipid content and Δδ13C, and between C:N and Δδ13C. For plants, C:N was not a good predictor of lipid content or Δδ13C, but we found a strong relationship between carbon content and lipid content, lipid content and Δδ13C, and between and carbon content and Δδ13C. Our results indicate that lipid extraction or normalization is most important when lipid content is variable among consumers of interest or between consumers and end members, and when differences in δ13C between end members is <10–12‰. The vast majority of studies using natural variation in δ13C fall within these criteria. Both direct lipid extraction and mathematical normalization reduce biases in δ13C, but mathematical normalization simplifies sample preparation and better preserves the integrity of samples for δ15N analysis.

CAN STABLE ISOTOPE RATIOS PROVIDE FOR COMMUNITY‐WIDE MEASURES OF TROPHIC STRUCTURE?

Stable isotope ratios (typically of carbon and nitrogen) provide one representation of an organisms trophic niche and are widely used to examine aspects of food web structure. Yet stable isotopes have not been applied to quantitatively characterize community-wide aspects of trophic structure (i.e., at the level of an entire food web). We propose quantitative metrics that can be used to this end, drawing on similar approaches from ecomorphology research. For example, the convex hull area occupied by species in δ13C–δ15N niche space is a representation of the total extent of trophic diversity within a food web, whereas mean nearest neighbor distance among all species pairs is a measure of species packing within trophic niche space. To facilitate discussion of opportunities and limitations of the metrics, we provide empirical and conceptual examples drawn from Bahamian tidal creek food webs. These examples illustrate how this methodology can be used to quantify trophic diversity and trophic redundancy in food webs, as well as to link individual species to characteristics of the food web in which they are embedded. Building from extensive applications of stable isotope ratios by ecologists, the community-wide metrics may provide a new perspective on food web structure, function, and dynamics.

BODY SIZE AND TROPHIC POSITION IN A DIVERSE TROPICAL FOOD WEB

We use stomach contents and stable isotope ratios of predatory fishes, collected over a 10-year time span from a species-rich river in Venezuela, to examine potential body-size–trophic-position relationships. Mean body size of predator taxa and their prey (determined by stomach content analyses) were significantly correlated, but trophic position of predators (estimated by stable isotope ratios) was not correlated with body size. This reflects no apparent relationship between body size and trophic position among prey taxa. Primary consumer taxa (algivores and detritivores) in this system are characterized by diverse size and morphology, and thus predatory fish of all body sizes and feeding strategies are able to exploit taxa feeding low in the food web. Regardless of relative body size, predators exploit short, productive food chains. For any given food chain within a complex web where predators are larger than their prey, trophic position and body size are necessarily correlated. But in diverse food webs...

Preservation Effects on Stable Isotope Analysis of Fish Muscle

Abstract We evaluated the effect of salt and formalin-ethanol sample preservation on carbon and nitrogen isotopic signatures of fish muscle tissue. We found statistically significant effects of the tissue preservation technique on both δ13C and δ15N; however, the magnitude of change was small and directionally uniform. Isotopic shifts were similar to those observed in previous studies in which formalin was used to preserve samples of quail blood and muscle and sheep blood. Because salt preservation caused minimal isotopic shifts (+0.13‰ δ13C, +0.72‰ δ15N), we propose salt as an easy, inexpensive preservation technique for biological samples collected in remote field settings. Specimens preserved with formalin and ethanol were minimally affected by preservation (−1.12‰ δ13C, +0.62‰ δ15N) and therefore may be suitable for ecological applications of stable isotope analysis when carbon and nitrogen sources are differentiated by more than 2‰. Further research is required to evaluate potential long-term storage...

HOW OFTEN DO FISHES “RUN ON EMPTY”?

We used a large data set of African, Neotropical, and North American fishes to examine the frequency with which fishes have empty stomachs (nspecies 5 254; nindividuals 5 36 875). Mean percentage of empty stomachs was low across all fishes (16.26 1.2%) but varied from 0% to 79.4% among individual species. Nocturnal fishes had empty stom- achs more frequently than diurnal fishes. Trophic classification was strongly associated with the percentage of empty stomachs, a pattern also revealed from an intraspecific analysis. Fishes appear to adjust their feeding intervals relative to the energy density, conversion efficiency, and particle size of their food. Piscivorous fishes seem to be the only trophic group that regularly experience long periods of empty stomachs, with species that consume prey whole and those that provide extended parental care having the highest proportions of empty stomachs. Activity patterns and life histories of some piscivorous species probably have evolved in partial response to energetic benefits of large, energy-rich food resources.

Habitat affinity, the seasonal flood pulse, and community assembly in the littoral zone of a Neotropical floodplain river

Abstract The annual flood pulse in tropical lowland rivers creates a continually moving land–water margin that forces aquatic organisms of the littoral zone to relocate to new habitats at intervals ranging from days to weeks. Thus, species assemblages in patchy littoral-zone habitats have large potential to be influenced by stochastic colonization dynamics, whereby organisms may exhibit weak selection for mesohabitat features and arrive on a patch in no particular sequence. We examined species assemblages of fishes and macroinvertebrates occupying habitats of varying structural complexity through an annual flood cycle of the Cinaruco River, a species-rich floodplain river in the southern llanos of Venezuela. We collected 268 standardized samples that contained 54,596 individual fishes representing 156 species (23 families) and 6973 macroinvertebrates representing 8 families. Analysis of species-specific patterns of habitat occupancy based on a randomization procedure indicated greater species selectivity during low water (92% of taxa) relative to rising- and falling-water periods (60% and 63% of taxa, respectively). Community-wide correspondence analysis (CA) revealed similar patterns. Assemblages from similar habitat types clustered together in the CA ordination, and these nonrandom patterns were most apparent during low water. Results from partial canonical correspondence analysis revealed that assemblage structure was strongly influenced by habitat type (patches of different composition located within the shallow littoral zone of either the river main channel or lagoons), which accounted for 56% of the total explainable variation in the species × sample matrix. Mesohabitat parameters (e.g., water velocity, depth), sampling period (month), and location on the landscape explained 30%, 23%, and 17% of variation, respectively. As water level changes and the aquatic–terrestrial interface moves across the landscape, species assemblages in patchy littoral habitats, while highly variable, reconstitute with significant determinism. At the same time, a large proportion of unexplained variation in assemblage structure probably is associated with unmeasured components of landscape variation as well as a stochastic element in colonization dynamics.

Spatiotemporal variation in fish assemblage structure in tropical floodplain creeks

Biotic assemblages of aquatic floodplain systems have great potential to randomly reshuffle during annual flood periods, and have been described both as stochastically and deterministically assembled. However, only a limited number of studies have been conducted in relatively few habitat types. To evaluate large-bodied fish assemblage structure of floodplain creeks, we used experimental gill nets to sample fishes at sites spaced at even intervals within three creeks in consecutive dry seasons. A total of 60 species were collected, 41 of which were collected both years. The most frequently collected species were piscivores and algivores/detritivores. Multivariate analysis suggested non-random patterns of assemblage structure in both years. Correspondence analysis (CA) of the species abundance-by-site matrix for 2001 suggests species assemblages were most similar among sites within the same creek regardless of depth or longitudinal position. Discriminant function analysis (DFA) correctly predicted 100% of samples based on creek identity, and species ordination scores revealed creek-specific species subsets. In 2002, CA and DFA did not distinguish creeks based on species assemblages. Instead, we observed a significant positive relationship between assemblage composition and site depth and position along the creek longitudinal gradient. Assemblages were most similar among sites of comparable depth and longitudinal position, regardless of creek identity. Predators occurred almost exclusively at mouth and mid-reach sites. Flood duration prior to our 2002 sampling period was prolonged due to abnormally heavy rainfall in November and December 2001 (typically the falling-water period), and may account for the observed inter-annual variation in fish assemblage structure.

Fish assemblage structure in relation to environmental variation in a Texas Gulf coastal wetland

We described seasonal fish-assemblages in an estuarine marsh fringing Matagorda Bay, Gulf of Mexico. Habitat zones were identified by patterns of fish species abundance and indicator species optima along gradients in salinity, dissolved oxygen (DO), and depth in our samples. Indicators of the lower brackish zone (lower lake and tidal bayou closest to the bay) were gulf menhaden (Brevoortia patronus), bay anchovy (Anchoa mitchilli), silver perch (Bairdiella chrysoura), and spotted seatrout (Cynoscion nebulosus) at salinity >15‰, DO 7–10 mg l−1, and depth <0.5 m. Indicators of the upper brackish zone (lake and fringing salt marsh) were pinfish (Lagodon rhomboides) and spot (Leiostomus xanthurus) at salinity 10–20‰, DO >10 mg l−1, and depth <0.5 m. In the freshwater wetland zone (diked wetland, ephemeral pool, and perennial scour pool), indicators were sheepshed minnow (Cyprinod on variegatus), rainwater killifish (Lucania parva), mosquitofish (Gambusia affinis), and sailfin molly (Poecilia latipinna) at salinity <5‰, DO <5 mg l−1, and depth ≥1 m. In the freshwater channelized zone (slough and irrigation canal), indicators were three sunfish species (Lepomis), white crappie (Pomoxis annularis), and gizzard shad (Dorosoma cepedianum) at salinity <5‰, DO <5 mg l−1, and depth >1.5 m. In brackish zones, seasonal variation in species diversity among sites was positively correlated with temperature, but assemblage structure also was influenced by depth and DO. In the freshwater zones, seasonal variation in species diversity among sites was positively correlated with depth, DO, and salinity, but assemblage structure was weakly associated with temperature. Species diversity and assemblage structure were strongly affected by the connectivity between freshwater wetland and brackish zones. Uncommon species in diked wetlands, such as tarpon (Megalops atlanticus) and fat sleeper (Dormitator maculatus), indicated movement of fishes from the brackish zone as the water level rose during natural flooding and scheduled (July) releases from the diked wetland. From September to July, diversity in the freshwater wetland zone decreased as receding waters left small isolated pools, and fish movement became blocked by a water-control structure. Subsequently, diversity was reduced to a few species with opportunistic life histories and tolerance to anoxic conditions that developed as flooded vegetation decayed.

Hydrologic Manipulations of the Channelized Kissimmee River Implications for restoration

H istorically, much of 50uthcentral Florida was dominated by a contiguous wetland system that extended from the headwater lakes of the Kissimmee River basin to Florida Bay. During the past half century, this wetland landscape has been compartmentalized with a network of canals, levees, and water-control structures (gated spillways; Figure 1). This network is used to manage hydrologic regimes of the regional hydrosystem, primarily for flood-control purposes (Light and Dineen 1994, Toth and Aumen 1994). Modifications of the physical configuration and hydrology of the South Florida. landscape have affected the Everglades, Lake Okeechobee, and the 7800 km2 Kissimmee River basin, where an extensive flood-control project was constructed from 1962 to 1971. Lakes in the rivers headwater basin were connected by canals and partitioned into floodstorage reservoirs; a 90 km long, 9 m deep, and 100 m wide drainage canal

Water depth modifies relative predation risk for a motile fish taxon in Bahamian tidal creeks

We evaluated the influence of water depth on relative predation risk for mojarra (Eucinostomus spp.) in six tidal creeks on Andros Island, Bahamas. Relative predation risk was determined using a tethering protocol combined with underwater visual census. In one experiment, we found that relative predation risk increased predictably with water depth (r2 = 0.83), and survival of tethered mojarra decreased with water depth (r2 = 0.71). We identified three depth zones containing differing levels of predation threat: refugia (0–19 cm), transition (20–69 cm), and predation (> 70 cm). Predation on mojarra rarely occurred within the refugia zone (2% eaten) and always in the predation zone (100% eaten). Additional factors not examined in this study (e.g., mangrove complexity, predator density) likely drive variability of relative predation risk within the transition zone. In a second experiment, we directly examined influence of water depth on relative predation risk at fixed locations from high tide to low tide in a single creek. Mean relative predation risk was significantly higher during higher tides at deeper water depths. Results provide experimental evidence that exploitation of shallow water refugia by motile prey can significantly reduce predation risk. We expect the distribution of motile fishes is at least partially influenced by spatially-dynamic shallow water refugia.