Skip to main content
NCBI home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Jan 1.
Published in final edited form as: J Great Lakes Res. 2019;45(1):129–137. doi: 10.1016/j.jglr.2018.11.006

Relative contributions of nearshore and wetland habitats to coastal food webs in the Great Lakes

Michael E Sierszen a,*, Lee S Schoen b, Jessica M Kosiara b, Joel C Hoffman a, Matthew J Cooper c,d, Donald G Uzarski b
PMCID: PMC6855385  NIHMSID: NIHMS1521669  PMID: 31728077

Abstract

Hydrologic linkages among coastal wetland and nearshore areas allow coastal fish to move among the habitats, which has led to a variety of habitat use patterns. We determined nutritional support of coastal fishes from 12 wetland-nearshore habitat pairs using stable isotope analyses, which revealed differences among species and systems in multi-habitat use. Substantial (proportions > 0.30) nutrition often came from the habitat other than that in which fish were captured. Nearshore subsidies to coastal wetlands indicate wetlands are not exclusively exporters of energy and materials; rather, there is reciprocity in the mutual energetic support of nearshore and wetland food webs. Coastal wetland hydrogeomorphology influenced the amount of multi-habitat use by coastal fishes. Fishes from systems with relatively open interfaces between wetland and nearshore habitats exhibited less nutritional reliance on the habitat in which they were captured, and higher use of resources from the adjacent habitat. Comparisons of stable isotope analyses of nutrition with otolith analyses of occupancy indicated nutritional sources often corresponded with habitat occupancy; however, disparities among place of capture, otolith analyses, and nutritional analyses indicated differences in the types of support those analyses inform. Disparities between occupancy information and nutritional information can stem from movements for support functions other than foraging. Together, occupancy information from otolith microchemistry and nutritional information from stable isotope analyses provide complementary measures of the use of multiple habitats by mobile consumers. This work underscores the importance of protecting or restoring a diversity of coastal habitats and the hydrologic linkages among them.

Keywords: Great Lakes, food webs, coastal habitats, stable isotopes, otoliths

Introduction

Identifying the habitats and resources that support production in aquatic ecosystems is essential to understand system dynamics and to identify priorities for ecosystem protection. Lacustrine food webs have been shown to rely on resources from a combination of benthic, pelagic, and littoral habitats (Hecky and Hesslein 1995, Schindler and Scheuerell 2002, Vander Zanden and Vadeboncoeur 2002). Although coupling among habitats has been emphasized for smaller lakes (e.g., Schindler and Scheuerell 2002), multihabitat interactions have been shown to be integral to even the world’s largest lakes (Eshenroder et al. 1999, Sierszen et al. 2014, Stockwell et al. 2014). In large lakes, adjacent habitats can be linked by energy flux (i.e., a physical exchange, such as carbon export from a coastal wetland to the nearshore; Bouchard 2007), animal movements (i.e., a biological exchange, such as nearshore fish larvae settling in coastal wetlands; Hoffman et al. 2015), or both.

Coastal wetlands are especially valuable systems; they are highly productive habitats that are coupled with less-productive nearshore waters (Uzarski et al. 2005) and produce a wide array of ecosystem services, which include the support of Great Lakes fisheries (Sierszen et al. 2012b, Trebitz and Hoffman 2015). Hydrologic linkages among Great Lakes nearshore and coastal wetland habitats allow coastal fish to move among the habitats for various life-support functions such as feeding, spawning, and thermal and oxygen refuge (e.g., Parker et al. 2012), which has led to a variety of habitat use patterns (Jude and Pappas 1992). Fine-scale microchemical analyses of yellow perch otoliths revealed life-history strategies that rely upon using those habitat linkages, including annual migrations between nearshore and wetlands, wetland residence, nearshore residence, and combinations thereof (Schoen et al. 2016).

A variety of approaches have been applied to examine the ways various lake habitats support aquatic food webs. Significant contributions of benthic habitats to whole-lake food webs have been demonstrated using fish diets (Schindler and Scheuerell 2002), measures of primary, bacterial, and secondary productivity (Vadeboncoeur et al. 2002), stable isotope analyses (Hecky and Hesslein 1995), and combinations of diet and stable isotope methods (Vander Zanden and Vadeboncoeur 2002). To quantify the importance of benthic, pelagic, and littoral habitats to support higher consumers in large lakes, stable isotope analyses (Sierszen et al. 2014) and a combination of bioenergetics modeling and stable isotope analyses (Stockwell et al. 2014) have been brought to bear.

In the Laurentian Great Lakes, movement of fishes among habitats has been tracked using mark-recapture methods, radio telemetry, and bioacoustics (reviewed by Landsman et al. 2011). The roles of coastal wetland and nearshore habitats in the nutrition and life-history of fishes have been explored using fish sampling (Jude and Pappas 1992, Brazner et al. 2001, Meixler et al. 2005), stable isotope analyses (Sierszen et al. 2012a, Hoffman et al. 2015), and otolith microchemical analyses (Brazner et al. 2004, Pangle et al. 2010, Schoen et al. 2016). We focus on the latter three techniques, as each provides useful and potentially different types of information on the ways aquatic habitats support fishes. Location of capture provides a basic level of information on habitat occupancy, but the duration of habitat use and the life functions supported by the habitat may not be apparent based on those measures alone (Cunjak et al. 2005). Otolith analyses also provide occupancy information, at time scales ranging from days to lifetimes, but also without revealing the type of support provided by the habitats (Hogan et al. 2007). Stable isotope analyses provide information on nutritional support at seasonal or annual time scales depending on growth rate (Vander Zanden et al. 2015), but with less information on occupancy.

We recognized that the degree to which exchanges occur between coastal wetland and nearshore habitats can be influenced by the hydrology and morphology of their interface (Polis et al. 1997, Keough et al. 1999, Morrice et al. 2011). We therefore designed this study to include sites with a range of hydrogeomorphology in selected coastal systems of lakes Huron and Michigan. Our primary objectives were therefore to quantify the nutritional support of coastal fishes by coastal wetland and nearshore habitats; to contrast nutritional support with occupancy measures of habitat use; and to examine the influence of coastal hydrogeomorphology on multi-habitat nutrition.

Methods

We collected fish and primary consumer (zooplankton and benthic invertebrate) samples from 12 wetland-nearshore pairings in Lake Michigan, northern Lake Huron, and Saginaw Bay (Fig. 1). Each site consisted of a nearshore area (3-5 m depth) adjacent to a coastal wetland (1-2 m depth). The sites were selected to represent the most common hydrogeomorphic wetland types of lakes Michigan and Huron as described in Albert et al. (2005). Hydrogeomorphic categories used in this study include barred drowned river mouth wetland, protected (embayment) lacustrine wetland, and open (non-embayment) lacustrine wetland, in order of increasing wave exposure and potential mixing with nearshore water. We selected sites based upon the distribution of hydrogeomorphic types in the basin and chose three drowned river mouth systems in Lake Michigan, three protected (embayment) lacustrine wetlands in Les Cheneaux of northern Lake Huron, and six open (non-embayment) lacustrine wetlands distributed among Thunder Bay and Saginaw Bay on Lake Huron (Table 1). To further characterize the degree to which the systems were exposed to waves and susceptible to exchange with the lakes, we calculated modified effective fetch and maximum fetch according to Howes et al. (1999). Preliminary nutrient analyses of a subset of the systems (Table 2, Schoen unpublished) suggest a slight nutrient gradient from Les Cheneaux systems (lowest) to Lake Michigan systems (highest). However, based upon total phosphorus and total nitrogen concentrations, all wetlands appeared to be mesotrophic, and nearshore habitats approximated oligotrophy (cf. Wetzel 1983).

Figure 1.

Figure 1.

Open in a new tab

Location of sampling sites on lakes Michigan and Huron. northern Lake Huron (Les Cheneaux) sites enlarged separately.

Table 1.

Study sites, locations (cf. Fig. 1), wetland hydrogeomorphic type, and estimates of exposure to lake waves (Modified Effective Fetch and Maximum Fetch, km).

Site Location HGM Type Mod. Eff. Fetch Max Fetch
Grand River Lake Michigan Drowned River Mouth 0.0099 1.4
Kalamazoo River Lake Michigan Drowned River Mouth 0.0072 0.61
Pentwater River Lake Michigan Drowned River Mouth 0.0310 2.2
Government Bay Les Cheneaux Protected Lacustrine 1.87 3.78
Duck Bay Les Cheneaux Protected Lacustrine 0.42 2.4
Mismer Bay Les Cheneaux Protected Lacustrine 0.1105 6.27
Wildfowl Bay Saginaw Bay Open Lacustrine 20.53 29.3
Au Gres Saginaw Bay Open Lacustrine 37.03 43
Pinconning Saginaw Bay Open Lacustrine 12.22 22
Quanicassee Saginaw Bay Open Lacustrine 38.09 255.8
Fish Point Saginaw Bay Open Lacustrine 21.45 40.65
Alpena Thunder Bay Open Lacustrine 167.32 176
Open in a new tab

Table 2.

Mean total nitrogen (TN) and total phosphorus (TP) concentrations sampled summer 2014 (Schoen unpublished) from wetland (WL) and nearshore (NS) habitats of a subset of the study systems. ND = below detection.

Site TN mg L−1 TP mg L−1
Grand WL 1.474 0.077
NS 0.554 ND
Kalamazoo WL 1.132 0.056
NS 0.514 0.002
Duck WL 0.517 0.011
NS 0.370 0.004
Mismer WL 0.367 0.009
NS 0.357 ND
Wildfowl WL 0.858 0.034
NS 0.946 0.020
Au Gres WL 0.636 0.019
NS 0.421 0.014
Pinconning WL 0.734 0.016
NS 0.477 0.020
Quanicassee WL 1.104 0.058
NS 2.168 0.020
Open in a new tab

All samples were collected between 15 June and 15 August 2012; fish and invertebrate samples were collected concurrently at each site. We collected wetland and nearshore zooplankton using triplicate 100 m subsurface horizontal tows with a 40 cm diameter, 80 μm mesh plankton net. In the wetlands, care was taken not to include benthic biota and detritus by avoiding sediments and macrophytes. Plankton tows were taken before fish and benthos sampling that might have suspended sediments and epiphytes. In the few cases when benthic or epiphytic materials were encountered, the sample was discarded, the net rinsed, and the tow was retaken. Nearshore tows were taken in at least 3 m depth to avoid benthic materials, and outside river plumes (where present) and far enough offshore to avoid wetland influence. Plankton samples were sieved onto 50 μm mesh Nitex filters, rinsed with deionized water, and frozen pending analysis. We collected macroinvertebrates using 500 μm mesh D-net sweeps within all vegetation zones (cf. Uzarski et al. 2004, Uzarski et al. 2017) in each wetland. Nearshore benthic macroinvertebrates were collected using PONAR grabs taken along three transects parallel to shore. Macroinvertebrates were sorted by taxon and then frozen pending analysis. To ensure that primary consumers were used to establish the food web baselines, we omitted predatory taxa (e.g., odonates) as well as those whose δ15N indicated they occupied higher trophic positions.

The Michigan Department of Natural Resources sampled nearshore fish using experimental gill-nets set in nearshore habitats adjacent to wetland habitats. However, no fish were collected from Pentwater nearshore. We sampled wetland fish with a combination of modified fyke-nets and boat electrofishing. After collection, fish were placed on ice and frozen within 24 hours. In the lab, muscle tissue samples were taken from anterior dorsal portions of individual fish. For smaller specimens, we analyzed whole fillets. Fish and invertebrate tissue samples were rinsed in deionized water, dried at 55°C, ground, and 0.5 mg packed into a tin capsule for stable isotope analysis. Samples were analyzed for δ13C and δ15N values using a Costech 4010 EA (Costech analytical Technologies Inc., Valencia, CA, U.S.A.) and Thermo Delta Plus XP isotope ratio mass spectrometer (Thermo Fisher Scientific Inc., Waltham, MA, U.S.A.). We analyzed 68 zooplankton samples, 277 benthic invertebrate samples, and 448 fish samples. The analytical error, expressed as the standard deviation of the stable isotope values of replicate analyses of reference materials, was 0.18 for δ13C and 0.21 for δ15N. Reference materials included walleye Sander vitreus muscle tissue used throughout the project to determine instrument precision and drift, as well as a commercially-obtained sulfanilamide standard (Sigma-Aldrich; δ13C = −28.58; δ15N=−3.1), which yielded the same estimates of precision as the walleye standard.

We quantified the nutritional contributions of wetland and nearshore food-web pathways to the common species of fishes collected using mean isotope values of zooplankton, benthos, and fishes in a multisource dual-isotope mixing model (IsoSource 1.3.1; Phillips and Gregg 2003; Phillips et al. 2005). IsoSource is an iterative probabilistic model that calculates source contributions to a stable isotope mixture when the number of sources is too large to permit a unique solution (i.e., number of sources > number of isotope systems + 1). All possible combinations of each source are examined in small increments, and combinations that sum to the observed mixture within a small tolerance (ca. 0.1 ‰) are considered feasible solutions. Each multivariate solution (i.e., each combination of source contributions that satisfies the model tolerance) is equally valid. However, each source has a univariate distribution of possible contributions, and the frequency of occurrence of a particular contribution in that distribution reflects the likelihood of that proportional contribution (Semmens et al., 2013). Correct reporting of model results requires that the distribution of feasible solutions be described (Phillips and Gregg 2003). When a distribution of source contributions is well constrained (i.e., it has a relatively narrow range of possible contributions) and symmetrical, the mean (i.e., the most frequent contribution of that source) describes the most likely contribution, and the standard deviation may be used to express dispersion about the mean. Distribution characteristics determine whether the mean, median or mode reflects the most likely value, and whether standard deviations, confidence intervals and quartiles are best used to express dispersion. Because source distributions are often skewed, we used medians and quartiles to describe source contributions, expressed as nutrition obtained through nearshore food-web pathways by summing contributions of nearshore zooplankton and benthos sources.

Calculating isotope mixing across trophic levels requires knowledge of trophic fractionation of isotopes (i.e., the 15N and 13C enrichment of a consumer relative to its food). We used a nitrogen isotope fractionation factor of 3.4 ‰ (Cabana and Rasmussen 1994) and carbon isotope fractionation of 1.0 ‰ (Vander Zanden and Rasmussen 2001). Those factors were used to adjust fish isotope values according to the trophic position of each fish sample, as calculated from the difference in δ15N between the food web base (mean of nearshore and wetland benthos and plankton sources) and the fish. To verify that those calculations were realistic, we compared our calculated trophic positions with those from FishBase (www.fishbase.org) and those calculated using both diet and stable isotope techniques (Vander Zanden et al. 1997).

We modelled the diets of fish species that were common in the wetland and nearshore habitats, which resulted in a total of 114 species · habitat combinations. We used several approaches to evaluate the degree to which fish were nutritionally supported by wetland and nearshore habitats during recent (seasonal or annual) life history. First, if the proportion of nearshore nutrition obtained from the mixing model was in the vicinity of 0.5, it would mean that the consumer obtained equivalent support from the two habitats. Therefore, mid-range (e.g., 0.4 – 0.6) values suggest high degrees of multihabitat support, whereas high (primarily nearshore nutrition) and low (primarily wetland nutrition) values indicate predominantly single-habitat support. Second, if fish from the two adjacent habitats had similar distributions of nearshore nutrition, it would suggest common foraging histories, either by traveling between the habitats or by an interaction of the habitats through hydrologic mixing or movement of prey. We tested for differences between wetland-caught and nearshore-caught fish using two-sample t-tests of arcsine-transformed data. Finally, if fish with high proportions of nutrition from one of the habitats were captured in the other habitat, it was an obvious (though not quantitative) indication of multihabitat use. These measures were compared among wetland sites that differed in hydrogeomorphic type and fetch exposure to the nearshore. We examined the influence of wetland type on sources of nutrition to wetland- and nearshore-caught fish by testing for differences in nearshore nutrition among fish from systems differing in hydrogeomorphic type, modified effective fetch (MEF), and maximum fetch (MaxF). We pooled fish species within each habitat and arcsine-transformed proportions of nearshore nutrition for their use in ANOVA.

We examined the influence of fish species on multiple-habitat use by comparing the intraspecific range of nearshore nutrition across systems. If fish species determined the degree to which they obtained nutrition from multiple habitats - for example, if species A fed mostly from wetland habitat and species B consistently fed from both habitats - we would expect each species to have a relatively narrow range in their proportions of nearshore nutrition when compared among all sites. Also, those ranges would differ in their magnitudes, with species A having low proportions of nearshore nutrition and species B having values near 0.5. Conversely, if fish species did not determine amount of multihabitat feeding, and foraging behavior varied with location, we would expect a relatively broad range in proportion of nearshore nutrition when compared among sites. We compared the range of median values for each species across sites and expressed the overall variability at each habitat as the difference between the 1st quartile associated with the minimum median value and the 3rd quartile associated with the maximum median value. We limited this analysis to species that were collected in four or more systems.

In an earlier study (Schoen et al. 2016), we used otolith microchemical analyses to trace movement patterns between wetland and nearshore habitats by yellow perch collected from the same sites as in this study. Otolith analyses revealed several life-history patterns in habitat use, including exclusive wetland residence, wetland use as juveniles followed by nearshore residence as adults, annual migrations between habitats, and combinations thereof. We have now enhanced that analysis by discriminating annual migrants that primarily used wetland habitats from those that primarily used lake habitats, according to amount of intra-annual growth accumulated in each habitat based on otolith ring analysis. We compared the estimates of wetland and nearshore habitat use according to time spent in those habitats (based on otolith analyses) with habitat use according to nutritional support (based on stable isotope mixing model results) by comparing the two measures for the same individual yellow perch. We also included fish for which we had stable isotope data but no microchemistry data, to a total of 124 individual yellow perch. We used the output of the isotope mixing model from individual yellow perch to address four objectives: 1) to determine whether life-history category predicted nutrition; 2) to test whether wetland-caught and nearshore-caught fish had similar nutritional sources (as above); 3) to test for discrepancies between habitat of capture and habitat of dominant nutritional support; and 4) to test for the influence of wetland hydrogeomorphology on the contributions of multiple habitats to fish nutrition.

Results

Wetland primary consumers analyzed for stable isotope ratios to characterize wetland benthic sources included amphipods, isopods, chironomids, mayflies, and snails. Richness of primary consumers in nearshore habitats was lower, but amphipods, oligochaetes, and chironomids were the most abundant taxa and plentiful enough to characterize nearshore benthic sources. Zooplankton were collected from each habitat but were not characterized taxonomically. Bulk zooplankton samples could therefore have contained taxa that varied in trophic position, which could add uncertainty to our calculations of the trophic positions of higher consumers; however, our trophic position estimates were in agreement with other published estimates (see below).

Fish species commonly collected in wetland habitats for stable isotope analysis included yellow perch Perea flavescens, rock bass Ambloplites rupestris, bluegill Lepomis macrochirus, pumpkinseed Lepomls glbbosus, largemouth bass Micropterus salmoides, smallmouth bass Micropterus dolomieu, northern pike Esox lucius, walleye Sander vitreus, bowfin Amia calva, black bullhead Ameiurus melas, brown bullhead Ameiurus nebulosus, bluntnose minnow Pimephales notatus, and round goby Neogobius melanostomus. Fishes captured in nearshore habitats included those species as well as lean lake trout Salvelinus namaycush, lake whitefish Coregonus clupeaformis, and freshwater drum Aplodinotus grunniens. Although ours was not a comprehensive collection, it includes major common species typical of Great Lakes coastal habitats (Jude and Pappas 1992, Keough et al. 1996, Brazner 1997).

We did not adjust δ13C values for lipid content prior to use in the mixing model. The ratio of carbon:nitrogen in our fish samples were generally < 3.5 and in rare instances exceeded 4.0. At low C:N values such as these, normalization of δ13C for lipid content is not recommended (e.g., Hoffman et al. 2015) and would not influence isotope mixing results. Further, use of lipid-corrected fish values in the mixing model would have required arithmetic correction of invertebrate values, which has been shown to be unreliable (Logan et al. 2008, Kiljunen et al. 2006).

Stable Isotope analyses

Across sites, we found a consistent difference between wetland and nearshore habitats in the δ13C values of their respective biota. At Lake Michigan drowned river mouth wetlands and Les Cheneaux protected lacustrine wetland sites, nearshore primary consumers were generally 13C-enriched relative to wetland primary consumers (Fig. 2a, b, Electronic Supplementary Material (ESM) Table S1). However, at four open lacustrine wetlands (Alpena, Pinconning, Wildfowl Bay, and Fish Point) nearshore primary consumers were 13C-depleted relative to those of wetland habitats (Fig. 2c, ESM Table S1). Fish caught in the nearshore and wetland habitats at the Lake Michigan drowned river mouth systems often had δ13C values that corresponded with those of primary consumers from the habitat of capture. In contrast, at many open and protected lacustrine wetland sites, the δ13C values of nearshore- and wetland-caught fish were interspersed (e.g., Fig. 2b). At every site except the Grand River, and in 51 of 128 species-habitat combinations (40%), the range in fish δ13C was outside that of the primary consumers in the habitat of capture (ESM Table S1), indicating the fish might have obtained substantial nutritional support from a habitat other than that of capture.

Figure 2.

Figure 2.

Open in a new tab

δ13C and δ15N values of biota from three representative wetland-nearshore sites: a) at the Grand River drowned river mouth site on Lake Michigan; b) at the Duck Bay protected lacustrine wetland site on northern Lake Huron’s Les Cheneaux region; c) at the Wildfowl Bay lacustrine wetland site on Lake Huron’s Saginaw Bay. Black dots: wetland primary consumers (zooplankton and benthos); black triangles: wetland fishes; open circles: nearshore primary consumers; open triangles: nearshore fishes.

At one open and two protected lacustrine wetland sites (Au Gres, Mismer Bay, and Government Bay), nearshore and wetland zooplankton samples had nearly identical stable isotope ratios. In contrast, wetland benthic invertebrates were 13C-depleted relative to nearshore benthos. This suggests there may have been strong hydrologic mixing between wetland and nearshore waters at these sites. In those cases, three sources were used in the mixing model (nearshore benthos, wetland benthos, zooplankton) and half of the zooplankton contribution to fish nutrition from the mixing model was attributed to nearshore habitats.

Comparisons of our calculated fish trophic positions to previously published estimates showed them to be in good agreement (Table 3). This is important, as trophic position was used along with trophic fractionation factors to standardize fish δ13C and δ15N values prior to use in the mixing model. Inclusion of juveniles resulted in lower trophic position estimates for some species.

Table 3.

Comparison of trophic position (TP) calculated in this study with those from the literature: diet-based estimates in FishBase (www.fishbase.org), and diet- and stable isotope (SI)-based estimates from Vander Zanden et al. (1997; VZ).

Fish species TP this study TP FishBase (diet) TP VZ (diet) TP VZ (SI)
walleye 3.6 – 4.7 4.5 ± 0.0 3.9 – 4.5 4.1 – 4.9
largemouth bass 3.2 – 5.3 3.8 ± 0.4 3.6 – 4.5 3.9 – 4.4
yellow perch 2.6 – 4.6 3.7 ± 0.2 3.5 – 4.5 3.0 – 4.3
bluegill 3.1 – 4.6 3.2 ± 0.2
northern pike 3.6 – 4.0 4.1 ± 0.4 4.2 – 4.5 3.4 – 4.5
smallmouth bass 2.9 – 4.5 3.6 ± 0.2 3.5 – 4.5 3.6 – 4.7
rock bass 3.4 – 4.8 3.4 ± 0.2 3.5 – 4.1 3.5 – 4.4
pumpkinseed 3.0 – 3.5 3.3 ± 0.1 3.0 – 3.5 2.8 – 4.2
black bullhead 2.9 – 4.1 3.8 ± 0.4
bowfin 3.6 – 4.8 3.8 ± 0.7
round goby 3.0 – 3.9 3.3 ± 0.1
Open in a new tab

Isotope mixing

Frequently, the habitat in which fish were captured supplied the majority of their nutrition as estimated through isotope mixing; however, substantial support (proportions > 0.30) often came from the adjacent habitat. Of the 113 species-habitat pairs, 37 (33%) had less than half of their nutrition originating from their habitat of capture, based on median values (Table S2).

Wetland-caught fish species at the drowned river mouth wetland sites on Lake Michigan had significantly lower amounts of nearshore-based nutrition than did nearshore-caught fish. There were significant differences in the amount of nearshore nutrition between wetland-caught and nearshore-caught fish (all fish species pooled) at the Grand River site (p=0.039) and the Kalamazoo River site (p< 0.001), indicating differences in foraging patterns between nearshore and wetland fish and relatively little food-web exchange between habitats. Wetland-caught fish tended to have proportions of nearshore nutrition ≤ 0.4, and nearshore fish tended to have proportions of nearshore nutrition ≥ 0.6 (e.g, Fig 3a; Table S2). At Pentwater, wetland-caught fish had low proportions (≤ 0.4) of nearshore nutrition. No nearshore fish were collected at Pentwater.

Figure 3.

Figure 3.

Open in a new tab

Mixing model results expressed as proportion nearshore-derived nutrition for fish species sampled from nearshore habitats (upper panels) and wetland habitats (lower panels) at the same three representative sites as in Fig. 1. Species abbreviations as in Table 1. Boxes span 1st to 3rd quartiles, with vertical line at median; whiskers indicate 10th and 90th percentiles; dots indicate 5th and 95th percentiles.

At the open and protected lacustrine wetland sites of Saginaw Bay, Alpena, and Les Cheneaux, proportions of nutrition from the nearshore for most fish species were near the middle of the range, with no statistical difference between fish caught in nearshore and conspecifics caught in wetland habitats (Fig 3b, c; ESM Table S2). Only two of the nine lacustrine wetlands (the open lacustrine site at Wildfowl Bay, p = 0.003, and the protected lacustrine site at Mismer Bay, p=0.018; two-sample t-tests; all fish species pooled within habitats) showed significant differences between the nutritional sources of wetland- and nearshore-caught fish, indicating differences in foraging habits between fish from the two habitats.

Estimates of wetland exposure to lake waves differed among hydrogeomorphic types. The drowned river mouth systems had the lowest MEF and MaxF values among all study sites, protected lacustrine sites were intermediate, and open lacustrine sites had the highest MEF and MaxF values (Table 1). Proportions of nearshore nutrition in wetland-caught fish (all species pooled within hydrogeomorphic types) differed significantly among types (ANOVA, p<0.001; F=10.897; r2=0.230), with fish from drowned river mouth systems having significantly lower nearshore nutrition than those from protected or open lacustrine sites. Nearshore nutrition did not differ significantly between wetland fish from protected and open lacustrine sites. Nearshore nutrition in wetland-caught fish increased significantly with MEF (p<0.001; F= 6.034; r2=0.509) and MaxF (p<0.001; F= 6.034; r2=0.509). Nearshore nutrition in nearshore-caught fish differed significantly among all hydrogeomorphic types, with highest nearshore nutrition in drowned river mouth systems and lowest nearshore nutrition in open lacustrine sites (p<0.001; F=15.512; r2=0.478). Nearshore nutrition in nearshore-caught fish decreased significantly with MEF (p<0.001; F=6.642; r2=0.719) and MaxF (p<0.001; F=6.642; r2=0.719).

Comparing the influence of fish species on sources of nutrition across sites, yellow perch and rock bass caught in wetland habitats tended to have lower median values of nearshore nutrition than did nearshore-caught conspecifics, whereas northern pike median values were similar in wetland and nearshore habitats (Table 4). The medians for wetland-caught yellow perch (and largemouth bass and bluegill, which were caught only in wetlands) indicated predominantly wetland nutrition (overall proportions 0.05-0.61); whereas, nearshore-caught yellow perch and rock bass tended toward nearshore nutrition (proportions 0.43-0.98); northern pike from both habitats and nearshore walleye had median values that bracketed 0.5, indicating a strong mix of nutrition from both habitats. The range of values for each species varied from 0.49 for wetland northern pike to 0.69 for wetland rock bass. These broad ranges suggest variable habitat support across systems, i.e., that site had a relatively strong effect on the amount of multi-habitat support of species. For a more comprehensive view of the variability in these data, see ESM Table S2.

Table 4.

Excerpt from Table S2. Median proportions of nearshore nutrition of fish species (YP=yellow perch; NP=northern pike; RKB=rock bass; LMB=largemouth bass; BLG=bluegill; WAE=walleye) from wetland (WL) and nearshore (NS) habitats. Quartile range indicates the span between the first quartile associated with the minimum median value and the third quartile associated with the maximum median value, and reflect relative variability in nutritional support among sites.

Site WL YP NS YP WL NP NS NP WL RKB NS RKB WL LMB WL BLG NS WAE
Grand R. 0.05 0.98 0.18 0.13 0.18
Kalamazoo R. 0.39 0.88 0.13 0.39 0.14
Pentwater R. 0.27 0.29 0.52 0.27
Duck Bay 0.53 0.81 0.80 0.87 0.72 0.90 0.28 0.78
Mismer Bay 0.41 0.90 0.69 0.63 0.75 0.67
Government Bay 0.61 0.44 0.41 0.39 0.39 0.45 0.37
Pinconning 0.41 0.83 0.63 0.34 0.82
Wildfowl 0.54 0.36 0.21 0.63
Quanicassee 0.60 0.68 0.63 0.61 0.51 0.66
Au Gres 0.05 0.47 0.41 0.31 0.41
Fish Point 0.77 0.56 0.55 0.32 0.57
Alpena 0.43 0.53 0.65 0.45
Min – max 0.05-0.61 0.43-0.98 0.41-0.80 0.39-0.87 0.13-0.72 0.45-0.90 0.13-0.61 0.14-0.51 0.37-0.82
Quartile Range 0.63 0.65 0.49 0.59 0.69 0.55 0.64 0.53 0.54
Open in a new tab

Otolith-isotope comparison

Yellow perch individuals classified as wetland residents via otolith microchemistry (n=6) had median proportions of nearshore nutrition ranging from 0.005 to 0.44 based on isotope mixing; those classified as nearshore residents (n=12) had nearshore nutrition from 0.42 to 0.96 (ESM Table S3). Overall, 5/6 (83%) of the individuals classified as wetland residents had most (proportions > 0.50) of their nutrition from wetland habitats and 10/12 (83%) of individual yellow perch classified as nearshore residents had most of their nutrition from the nearshore. Annual migrants classified as primarily using nearshore habitats (n=16) obtained from 0.05 to 0.995 nearshore nutrition, and 81% of them had >0.50 of their nutrition from nearshore habitats; annual migrants classified as primarily using wetland habitats (n=20) had from 0 to 0.96 nearshore nutrition, and 70% of those individuals had >0.50 of their nutrition from wetland habitats. Young-of-year (n=38), which we were unable to classify reliably from otolith microchemistry, had median estimates of nearshore nutrition that ranged from zero to 0.995. Of those fish, 18% were captured in wetland habitats but had high proportions of nearshore nutrition based on the stable isotope mixing model, indicating that young-of-year routinely move between habitats.

Individual yellow perch at drowned river mouth sites obtained high proportions of their nutrition from their habitat of capture (Fig. 4a; ESM Table S3). As observed in the species-level analyses, individual yellow perch at protected and open lacustrine wetland sites had variable amounts of nearshore nutrition, often mid-range, indicating substantial inputs from both habitats (Fig. 4b, c). There were significant differences in the proportion of nearshore nutrition between nearshore-caught and wetland-caught individual yellow perch at the Grand River (p<0.001) and Kalamazoo River (p=0.006) drowned river mouth sites, but no significant differences at any of the lacustrine wetland sites. At drowned river mouth sites, 30 of 32 (94%) had more than half of their nutrition matching the habitat in which they were captured. At lacustrine wetland sites, 58 of 92 (63%) of individuals captured had over half of their nutrition matching the habitat in which they were captured.

Figure 4.

Figure 4.

Open in a new tab

Mixing model results for individual yellow perch sampled from nearshore and wetland habitats at the three representative sites; box dimensions as in Fig. 3. Life-history habitat categories determined from otolith microchemical analyses indicated next to the boxes: WL (wetland resident); NR (nearshore resident); M/N (annual migrant between habitats using mostly nearshore); M/W (annual migrant using mostly wetland); YOY (young-of-year). When no life-history type is indicated, otoliths were not analyzed.

Occupancy based on habitat of capture and life-history patterns identified from otolith analyses often predicted the major habitat supporting fish nutrition, but there were notable exceptions. For numerous individual yellow perch, the habitat providing most of their nutrition was not the habitat in which they were captured or in which they spent most of their time based upon otolith microchemical analyses (Table 5). For example, we captured yellow perch in the nearshore that were classified as migrants primarily using the nearshore, but which had very little nearshore diet contributions (e.g., at Duck Bay). We also captured fish in the nearshore that were classified as migrants that primarily used wetlands, but which had high nutritional support from the nearshore (e.g., at Pinconning). We captured one fish in the Quanicassee wetland that was classified as a nearshore resident but which had almost no nearshore diet contribution. This suggests the fish was targeting prey that were exported to the nearshore, or there was an error in either the otolith classification or the isotope mixing model.

Table 5.

Excerpt from Table S3. Individual yellow perch for which the habitat of occupancy, based on place of capture (Capt. Hab.) or life history classification from otolith microchemical analyses (LH), did not provide most of the nutrition based on stable isotope mixing model results. Capture habitat was wetland (W) or nearshore (N); life history classifications are wetland resident (WR), nearshore resident (NR), annual migrants (M/W for primarily wetland use; M/N for primarily nearshore use) and young of year (Y). Proportion nearshore nutrition given as median (1st quartile, 3rd quartile).

Site Capt. Hab. LH Prop. Nearshore
Duck N M/N 0.28 (0.25, 0.31)
N M/N 0.19 (0.17, 0.21)
Mismer W Y 0.65 (0.62, 0.66)
Government W 0.71 (0.65, 0.77)
N NR 0.42 (0.32, 0.53)
N 0.31 (0.30, 0.32)
Pinconning W Y 0.65 (0.53, 0.76)
W Y 0.83 (0.77, 0.90)
W Y 0.59 (0.48, 0.70)
W 0.87 (0.82, 0.92)
W 0.71 (0.59, 0.82)
W 0.77 (0.69, 0.87)
N M/W 0.85 (0.78, 0.91)
N M/W 0.93 (0.90, 0.96)
Quanicassee W Y 0.995 (0.99, 1.00)
W Y 0.72 (0.65, 0.81)
W Y 0.995 (0.99, 1.00)
W 0.63 (0.50, 0.76)
W M/N 0.05 (0.03, 0.07)
Au Gres N 0.43 (0.40, 0.46)
Fish Point W 0.68 (0.56, 0.79)
W Y 0.55 (0.39, 0.70)
W 0.61 (0.47, 0.74)
W Y 0.66 (0.53, 0.78)
Open in a new tab

Discussion

Our analyses shed light on the contributions coastal wetland and lake habitats make to the nutrition of coastal fishes and advance our understanding of factors that influence multi-habitat use by coastal fishes in large lakes. Among the places and fish species we examined, use of resources from multiple habitats was widespread. It was also reciprocal in that fish caught in both wetland and nearshore often obtained substantial amounts of nutrition from the alternative habitat. Our quantification of food web subsidies from coastal wetlands to the nearshore is consistent with the characterization of coastal wetlands as energy-rich systems that export resources to adjacent zones (Odom 1968, Wetzel 1992, Brazner et al. 2000). Further, our results go beyond the “outwelling” perspective of wetlands as sources of energy and materials (Odom 1968; see also Polis et al. 1997) because there is reciprocity in the mutual energetic support of nearshore and wetland habitats. Our measures of nearshore contributions to wetland biota concur with earlier estimates from coastal habitats in the other Laurentian Great Lakes (Sierszen et al. 2012a, Hoffman et al. 2015) and enhance our recognition of interacting organisms and resources among intact Great Lakes habitats.

Discrepancies between results of different measures of habitat resource use are particularly informative to our understanding of how fishes use habitat. Capture of fish in habitats that did not provide the majority of their nutrition may reflect relatively short-term (e.g., diel) occupancy, emigration from nursery habitats by young-of-year (e.g., Brazner et al. 2001), or feeding on prey resources that were exported from adjacent areas. Similarly, when otolith analyses indicate patterns in movement or occupancy that are not reflected in nutrition, it can relate to movement for life functions other than feeding, such as spawning migration (when little feeding is done) or to access thermal or other physiological refugia. Movements of Atlantic salmon into tributary waters for thermal refuge were tracked through telemetry, but were not reflected in stable isotope values (Cunjak et al. 2005). Also, seasonal movements among different habitats may sometimes not be reflected in otolith profiles due to seasonal depression of otolith accretion (Clement et al. 2014).

The majority of variability we measured in coastal fish stable isotope values was in δ13C, which was often outside the measured range of primary consumers in the habitat of capture. Differences between the δ13C of coastal wetland and nearshore biota can be traced to differences in the δ13C of dissolved inorganic carbon in those habitats (Keough et al. 1996, 1998). Higher trophic-level consumers with δ13C values outside those of their capture habitat have been described in other coastal systems of lakes Superior, Michigan, Erie, and Ontario (Keough et al. 1996, Hoffman et al. 2010, Sierszen et al. 2012a), and may be expected given accounts of movements among coastal habitats by Great Lakes fishes (Jude and Pappas 1992, Brazner et al. 2001, Landsman et al. 2011, Midwood and Chow-Fraser 2015). These conditions necessitated our use of all wetland and nearshore sources in the isotope mixing model.

Wetland hydrogeomorphology influenced the use of multiple habitats by coastal fishes. Analyses of variance among wetland classes and across measures of wetland exposure to nearshore waters revealed consistent patterns in the use of resources from nearshore and wetland. As the interface between wetland and nearshore became more open to exchange (i.e., from drowned river mouth to protected lacustrine to open lacustrine wetland, and at higher measures of modified effective fetch and maximum fetch) fish from both habitats exhibited less nutritional reliance on the habitat in which they were captured, and higher use of resources from the adjacent habitat. Boundary conditions can influence their accessibility and the amount of habitat coupling by mobile consumers (Dolson et al. 2009). In earlier analyses, a Great Lakes coastal wetland separated from the lake by a large barrier spit had a food web isotopically dissimilar to that of the adjacent nearshore (Keough et al. 1996), whereas there was complete isotopic overlap between the food webs of a morphologically open coastal wetland and nearshore waters (Brazner et al. 2000).

Fish species appeared to differ somewhat in their support from multiple habitats. Yellow perch and rock bass (and largemouth bass and bluegill, which were only captured in wetlands) tended toward support from their site of capture, whereas northern pike and walleye had stronger tendencies toward multihabitat support. Nonetheless, all species we examined had wide ranges in their nutrition across sites, indicating that site had a relatively strong effect on foraging patterns.

At the Lake Michigan and Les Cheneaux sites, nearshore primary consumers and most nearshore fish were 13C-enriched relative to wetland biota, a pattern commonly observed in other Great Lakes coastal wetlands attributable to respiration and remineralization of carbon in wetland habitats (Keough et al. 1996, 1998, Sierszen et al. 2012a). In wetlands situated on Saginaw Bay and Thunder Bay (Alpena), nearshore biota were 13C-depleted relative to those of wetland habitats. This suggests that the nearshore food webs in those bays were influenced by 13C-depleted inputs from the Saginaw River and Thunder Bay River, respectively, which probably carried strong heterotrophic signals. Alternatively, high productivity in those wetlands may have resulted in decreased C isotope fractionation during C uptake and photosynthesis, and led to 13C enrichment of wetland biota.

Understanding the use of aquatic habitats in the Great Lakes is valuable to fishery management and to identify critical habitats for protection and restoration (Landsman et al. 2011). Our findings enhance our understanding of the ways multiple habitats support food webs in large lakes. They also indicate that complementary types of information are provided by these different measures of habitat use. The reciprocity we found in the support of coastal fishes by adjacent habitats, revealed by nutritional and occupancy information, and the role of hydrogeomorphology in regulating those interactions, are valuable advancements. Coastal fishes in the Great Lakes use a mosaic of interacting habitats, and hydrologic linkages among those habitats maintain diversity of fishes, which act as metacommunities (Bouvier et al. 2009, Sierszen et al. 2012a, Midwood and Chow-Fraser 2015). The use of restored access to spawning and nursery habitats by coastal fish (e.g., Oele et al. 2015) makes a compelling argument for restoration of lost inter-habitat connections. Our results substantiate the value of protecting coastal habitats and restoring coastal wetlands that are hydrologically connected to the Great Lakes.

Supplementary Material

Supplementary S1-S3

Acknowledgements

We thank David Clapp, Dave Fielder, Bill Wellenkamp, and the Michigan Department of Natural Resources Fisheries Division for assisting in fish collections, and Ryan Wheeler, Tom Langer, Bridget Wheelock, Lindsey Adams, and Kaley Genther for assisting in the collection and preparation of samples for analysis. Thanks to John Brazner and the journal reviewers for critical comments that substantially improved the manuscript. The research described in this article has been funded in part by the U.S. Environmental Protection Agency as part of the Great Lakes Restoration Initiative under proposal EPA GLNPO-2010-H-3-984-758 and the Great Lakes Fish and Wildlife Restoration Act under FWS Agreement Number: F14AP00520. The research was also supported by the Garden Club of America Coastal Wetlands Award, Center for Coastal Resources Management at the Virginia Institute of Marine Science, College of William and Mary. The views expressed in this article are those of the authors and do not necessarily reflect the views or policies of the U.S. EPA, Michigan DNR, or Michigan DEQ. This paper is Contribution Number XXX of the Central Michigan University Institute for Great Lakes Research.

References

  1. Albert DA, Wilcox DA, Ingram J, Thompson TA, 2005. Hydrogeomorphic classification for Great Lakes coastal wetlands. J. Great Lakes Res 31 (Suppl. 1), 129–146. doi: 10.1016/S0380-1330(05)70294-X [DOI] [Google Scholar]
  2. Bouchard V, 2007. Export of organic matter from a coastal freshwater wetland to Lake Erie: an extension of the outwelling hypothesis. Aquat. Ecol 41, 1–7. [Google Scholar]
  3. Bouvier LD, Cottenie K, Doka SE, 2009. Aquatic connectivity and fish metacommunities in wetlands of the lower Great Lakes. Can. J. Fish. Aquat. Sci 66, 933–948. [Google Scholar]
  4. Brazner JC, 1997. Regional, habitat, and human development influences on coastal wetland and beach fish assemblages in Green Bay, Lake Michigan. J. Great Lakes Res 23, 36–51. [Google Scholar]
  5. Brazner JC, Campana SE, Tanner DK, Schram ST, 2004. Reconstructing habitat use and wetland nursery origin of yellow perch from Lake Superior using otolith elemental analysis. J. Great Lakes Res 30, 492–507. [Google Scholar]
  6. Brazner JC, Sierszen ME, Keough JR, Tanner DK, 2000. Assessing the ecological importance of coastal wetlands in a large lake context. Verh. Internat. Verein. Limnol 27, 1950–1961. [Google Scholar]
  7. Brazner JC, Tanner DK, Morrice JA, 2001. Fish-mediated nutrient and energy exchange between a Lake Superior coastal wetland and its adjacent bay. J. Great Lakes Res 27, 98–111. [Google Scholar]
  8. Cabana G, Rasmussen JB, 1994. Modeling food chain structure and contaminant bioaccumulation using stable nitrogen isotopes. Nature 372, 255–257. [Google Scholar]
  9. Clement M, Chiasson AG, Veinott GL, Cairns DK, 2014. What otolith microchemistry and stable isotope analysis reveal and conceal about anguillid eel movements across salinity boundaries. Oecologia 175, 1143–1153. [DOI] [PubMed] [Google Scholar]
  10. Cunjak RA, Roussel J-M, Gray MA, Dietrich JP, Cartwright DF, Munkittrick KR, Jardine TD, 2005. Using stable isotope analysis with telemetry or mark-recapture data to identify fish movement and foraging. Oecologia 144, 636–646. [DOI] [PubMed] [Google Scholar]
  11. Dolson R, McCann K, Rooney N, Ridgway M, 2009. Lake morphometry predicts the degree of habitat coupling by a mobile predator. Oikos 118, 1230–1238. [Google Scholar]
  12. Eshenroder R, Sideleva VG, Todd TN, 1999. Functional convergence among pelagic sculpins of Lake Baikal and deepwater ciscoes of the Great Lakes. J. Great Lakes Res 25, 847–855. [Google Scholar]
  13. Hecky RE, Hesslein RH, 1995. Contributions of benthic algae to lake food webs as revealed by stable isotope analysis. J. N. Am. Benth. Soc 14, 631–653. [Google Scholar]
  14. Hoffman JC, Kelly JR, Peterson GS, Cotter AM, 2015. Landscape-scale food webs offish nursery habitat along a river-coast mixing zone. Estuaries and Coasts 38, 1335–1349. [Google Scholar]
  15. Hoffman JC, Peterson GS, Cotter AM, Kelly JR, 2010. Using stable isotope mixing in a Great Lakes coastal tributary to determine food web linkages in young fishes. Estuaries and Coasts 33, 1391–1405. [Google Scholar]
  16. Hoffman JC, Sierszen ME, Cotter AM, 2015. Fish tissue Iipid-C:N relationships for correcting δ13C values and estimating lipid content in aquatic food-web studies. Rapid Commun. Mass Spectrom 29, 2069–2077. [DOI] [PubMed] [Google Scholar]
  17. Hogan Z, Baird IG, Radtke R, Vander Sanden MJ, 2007. Long distance migration and marine habitation in the tropical Asian catfish, Pangasius krempfi. J. Fish Biol 71, 818–832. [Google Scholar]
  18. Howes DE, Morris M, Zacharias M, 1999. British Columbia Estuary Mapping System. Resources Inventory Committee, Province of British Columbia, Canada: ISBN 0-7726-3911-6. [Google Scholar]
  19. Jude DJ, Pappas J, 1992. Fish utilization of Great Lakes coastal wetlands. J. Great Lakes Res 18, 651–672. [Google Scholar]
  20. Keough JR, Sierszen ΜE, Hagley CA, 1996. Analysis of a Lake Superior coastal food web with stable isotope techniques. Limnol. Oceanogr 41, 136–146. [Google Scholar]
  21. Keough JR, Hagley CA, Ruzycki E, Sierszen ME, 1998. δ13C composition of primary producers and role of detritus in a freshwater coastal ecosystem. Limnol. Oceanogr 43, 734–740. [Google Scholar]
  22. Keough JR, Thompson TA, Guntenspergen GR, Wilcox DA, 1999. Hydrogeomorphic factors and ecosystem responses in coastal wetlands of the Great Lakes. Wetlands 19, 821–834. [Google Scholar]
  23. Kiljunen M, Grey J, Sinisalo T, Harrod C, Immonen H, Jones RI, 2006. A revised model for lipid-normalizing values from aquatic organisms, with implications for isotope mixing models. J. Appl. Ecol 43, 1213–1222. [Google Scholar]
  24. Landsman SJ, Nguyen VM, Gutowsky LFG, Gobin J, Cook KV, Binder TR, Lower N, McLaughlin RL, Cooke SJ, 2011. Fish movement and migration studies in the Laurentian Great Lakes: Research trends and knowledge gaps. J. Great Lakes Res 37, 365–379. [Google Scholar]
  25. Logan JM, Jardine TD, Miller TJ, Bunn SE, Cunjak RA, Lutcavage ME, 2008. Lipid corrections in carbon and nitrogen stable isotope analyses: comparison of chemical extraction and modelling methods. J. Anim. Ecol 77, 838–846. [DOI] [PubMed] [Google Scholar]
  26. Meixler MS, Arend KK, Bain MB 2005. Fish community support in wetlands within protected embayments of Lake Ontario. J. Great Lakes Res 31 (Suppl. 1): 188–196. [Google Scholar]
  27. Midwood JD, Chow-Fraser P, 2015. Connecting coastal marshes using movements of resident and migratory fishes. Wetlands 35, 69–79. [Google Scholar]
  28. Morrice JA, Trebitz AS, Kelly JR, Sierszen ME, Cottter AM, Hollenhorst T, 2011. Determining sources of water to Great Lakes coastal wetlands: a classification approach. Wetlands 31, 1199–1213. [Google Scholar]
  29. Odom EP, 1968. A research challenge: Evaluating the productivity of coastal and estuarine water. Proceedings of the Second Sea Grant Conference Univ. of Rhode Island, pp. 63–64. [Google Scholar]
  30. Oele DL, Hogan JD, McIntire PB, 2015. Chemical tracking of northern pike migrations: If we restore access to breeding habitat, will they come? J. Great Lakes Res 41, 853–861. [Google Scholar]
  31. Pangle KL, Ludsin SA, Fryer JJ, 2010. Otolith microchemistry as a stock identification tool for freshwater fishes: Testing its limits in Lake Erie. Can. J. Fish. Aquat. Sci 67, 1475–1489. [Google Scholar]
  32. Parker AD, Cooper MJ, Ruetz CR III Coulter DP, Uzarski DG, 2012. Chemical and physical factors associated with yellow perch abundance in Great Lakes coastal wetlands: patterns within and among wetland types. Wetl. Ecol. Manag 20, 137–150. [Google Scholar]
  33. Phillips DL, Gregg JW, 2003. Source partitioning using stable isotopes: coping with too many sources. Oecologia 136, 261–269. [DOI] [PubMed] [Google Scholar]
  34. Phillips DL, Newsome SD, Gregg JW, 2005. Combining sources in stable isotope mixing models: alternative methods. Oecologia 144, 520–527. doi: 10.1007/s00442-004-1816-8. [DOI] [PubMed] [Google Scholar]
  35. Polis GA, Anderson WB, Holt RD, 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annu. Rev. Ecol. Syst 28, 289–316. [Google Scholar]
  36. Schindler DE, Scheuerell MD, 2002. Habitat coupling in lake ecosystems. Oikos 98, 177–189. [Google Scholar]
  37. Schoen LS, Student JJ, Hoffman JC, Sierszen ME, Uzarski DG, 2016. Reconstructing fish movements between coastal wetland and nearshore habitats of the Great Lakes. Limnol. Oceanogr 61, 1800–1813. [Google Scholar]
  38. Semmens BX, Ward EJ, Parnell AC, Phillips DL, Bearhop S, Inger R, Jackson A, Moore JW, 2013. Statistical basis and outputs of stable isotope mixing models: Comment on Fry (2013). Mar. Ecol. Prog. Ser 490, 285–289. [Google Scholar]
  39. Sierszen ME, Brazner JC, Cotter AM, Morrice JA, Peterson GS, Trebitz AS, 2012a. Watershed and lake influences on the energetic base of coastal wetland food webs across the Great Lakes Basin. J. Great Lakes Res 38, 418–428. [Google Scholar]
  40. Sierszen ME, Hrabik TR, Stockwell JD, Cotter AM, Hoffman JC, Yule DL, 2014. Depth gradients in food-web processes linking habitats in large lakes: Lake Superior as an exemplar ecosystem. Freshwat. Biol 59, 2122–2136. [Google Scholar]
  41. Sierszen ME, Morrice JA, Trebitz AS, Hoffman JC, 2012b. A review of selected ecosystem services provided by coastal wetlands of the Laurentian Great Lakes. Aquat. Ecosys. Health Mgmt 15, 92–106. [Google Scholar]
  42. Stockwell JD, Yule DL, Hrabik TR, Sierszen ME, Isaac EJ, 2014. Habitat coupling in a large lake system: delivery of an energy subsidy by an offshore planktivore to the nearshore zone of Lake Superior. Freshwat. Biol 59, 1197–1212. [Google Scholar]
  43. Trebitz AM, Hoffman JC, 2015. Coastal wetland support of Great Lakes fisheries: progress from concept to quantification. Trans. Am. Fish. Soc 144, 352–372. [Google Scholar]
  44. Uzarski DG, Brady VJ, Cooper MJ, Wilcox DA Albert DA, Axler R, Bostwick P, Brown TA, Ciborowski JJH, Danz NP, Gathman J, Gehring T, Grabas G, Garwood A, Howe R, Johnson LB, Lamberti GA, Moerke A, Murry B, Niemi G, Norment CJ, Ruetz CR III, Steinman AD, Tozer D, Wheeler R, O’Donnell TK, Schneider JP, 2017. Standardized measures of coastal wetland condition: implementation at the Laurentian Great Lakes basin-wide scale. Wetlands 37, 15–32. DOI: 10.1007/s13157-016-0835-7. [DOI] [Google Scholar]
  45. Uzarski DG, Burton TM, Genet JA, 2004. Validation and performance of an invertebrate index of biotic integrity for Lakes Huron and Michigan fringing wetlands during a period of lake level decline. Aquat. Ecosys. Health Mgmt 7(2), 269–288. [Google Scholar]
  46. Uzarski DG, Burton TM, Cooper MJ, Ingram JW, Timmermans STA, 2005. Fish habitat use within and across wetland classes in coastal wetlands of the five Great Lakes: Development of a fish-based index of biotic integrity. J. Great Lakes Res 31 (Suppl. 1), 171–187. [Google Scholar]
  47. Vadeboncoeur Y, Vander Zanden MJ, Lodge DM, 2002. Putting the lake back together: Reintegrating benthic pathways into lake food web models. BioScience 52, 44–54. [Google Scholar]
  48. Vander Zanden MJ, Rasmussen JB, 2001. Variation in δ15N and δ13C trophic fractionation: implications for aquatic food web studies. Limnol. Oceanogr 46, 2061–2066. [Google Scholar]
  49. Vander Zanden MJ, Cabana G, Rasmussen JB, 1997. Comparing trophic position of freshwater fish calculated using stable nitrogen isotope ratios (δ15N) and literature dietary data. Can. J. Fish. Aquat. Sci 54, 1142–1158. [Google Scholar]
  50. Vander Zanden MJ, Clayton MK, Moody EK, Solomon CT, Weidel BC, 2015. Stable isotope turnover and half-life in animal tissues: a literature synthesis. PLoS ONE 10, e0116182. doi: 10.1371/journal.pone.0116182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Vander Zanden MJ, Vadeboncoeur Y, 2002. Fishes as integrators of benthic and pelagic food webs in lakes. Ecology 83, 2152–2161. [Google Scholar]
  52. Wetzel RG, 1983. Limnology, Second edition Holt, Rinehart and Winston, Inc. Orlando, FL. [Google Scholar]
  53. Wetzel RG, 1992. Wetlands as metabolic gates. J. Great Lakes Res 18, 529–532. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary S1-S3
NIHMS1521669-supplement-Supplementary_S1-S3.pdf (57.6KB, pdf)

RESOURCES