Human-Induced Long-Term Shifts in Gull Diet from Marine to

Aug 24, 2015 - Procellaria Research & Consulting, 944 Dunsmuir Road, Victoria, British Columbia V9A 5C3, Canada. Environ. Sci. Technol. , 2015, 49 (18...
2 downloads 5 Views 1MB Size
Article pubs.acs.org/est

Human-Induced Long-Term Shifts in Gull Diet from Marine to Terrestrial Sources in North America’s Coastal Pacific: More Evidence from More Isotopes (δ2H, δ34S) Keith A. Hobson,*,† Louise K. Blight,‡,§ and Peter Arcese‡ †

Environment Canada, 11 Innovation Boulevard, Saskatoon, Saskatchewan S7N 3H5, Canada Department of Forest and Conservation Sciences, University of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada § Procellaria Research & Consulting, 944 Dunsmuir Road, Victoria, British Columbia V9A 5C3, Canada ‡

ABSTRACT: Measurements of naturally occurring stable isotopes in tissues of seabirds and their prey are a powerful tool for investigating long-term changes in marine foodwebs. Recent isotopic (δ15N, δ13C) evidence from feathers of Glaucous-winged Gulls (Larus glaucescens) has shown that over the last 150 years, this species shifted from a midtrophic marine diet to one including lower trophic marine prey and/or more terrestrial or freshwater foods. However, long-term isotopic patterns of δ15N and δ13C cannot distinguish between the relative importance of lower trophic-level marine foods and terrestrial sources. We examined 48 feather stable-hydrogen (δ2H) and -sulfur (δ34S) isotope values from this same 150-year feather set and found additional isotopic evidence supporting the hypothesis that gulls shifted to terrestrial and/or freshwater prey. Mean feather δ2H and δ34S values (±SD) declined from the earliest period (1860−1915; n = 12) from −2.5 ± 21.4‰ and 18.9 ± 2.7 ‰, respectively, to −35.5 ± 15.5 ‰ and 14.8 ± 2.4 ‰, respectively, for the period 1980−2009 (n = 12). We estimated a shift of ∼30% increase in dependence on terrestrial/ freshwater sources. These results are consistent with the hypothesis that gulls increased terrestrial food inputs in response to declining forage fish availability.



INTRODUCTION A key question in conservation science is not whether most of the world’s species will be influenced by rapid anthropogenic change but how they will respond to change and whether or not management actions can conserve viable populations into the future.1−4 Coastal regions have undergone dramatic human population increases, and coastal marine ecosystems have been considerably affected by a host of factors such as overfishing, habitat loss, pollution, and development of adjacent terrestrial regions.5 Modeling offers one approach to exploring the consequences of such anthropogenic change, but most biotic responses remain poorly understood. In addition, retrospective or comparative analyses, commonly applied in climate or meteorological science, remain rare because of a paucity of long-term biological data and a lack of knowledge on historical ecosystems states (see ref 6). One of the tools available for retrospective environmental modeling is the measurement of stable isotope ratios of the common light elements (C, N, H, O, S) in biotic materials that have been archived in museums or other collections. These isotopic values can remain unchanged in archived materials, and so, such materials can be extremely valuable recorders of long-term trends in environmental processes (nutrient cycling), as well as of animal diets and precise sources of nutrients, and have formed the basis of numerous paleoclimate and paleodietary reconstructions.7,8 Recently, Blight et al. (2014)9 examined head and flight feathers of adult and immature Glaucous-winged Gulls (Larus glaucescens) from a nonmigratory population breeding in the © 2015 American Chemical Society

inshore waters of British Columbia (BC, Canada) and Washington (USA), with feathers sampled over a 150-year period (1860−2009). Those feathers were largely available from museum specimens and, given timing of moult and regrowth, represented an isotopic record of individual diets from the physiologically demanding pre- and postbreeding periods over time. Although Larus gulls are generalist feeders, successful breeders consume a high proportion of fish prey, at least during the breeding season.10 Such a diet allows for population persistence as a result of prey-switching during times of limited food availability, but also means that Glaucouswinged Gull breeding parameters may be a sensitive indicator of environmental change. In our study area, populations of small “forage” fishes were harvested in commercial or recreational fisheries from the 1890s onward, in many cases to the point of disappearing (or nearly so) from the system.9,11,12 Pollution and climate change have also affected local marine foodwebs.13 Blight et al. (2014),9 therefore, tested for evidence of long-term changes in marine foodweb of the study area and how these birds may have responded, especially to a reduction in high-quality mesopelagic prey fishes such as herring (Clupea pallasii) and eulachon (Thaleichthys pacif icus). That study showed a gradual 150-year decline in stableReceived: Revised: Accepted: Published: 10834

April 29, 2015 August 20, 2015 August 24, 2015 August 24, 2015 DOI: 10.1021/acs.est.5b02053 Environ. Sci. Technol. 2015, 49, 10834−10840

Article

Environmental Science & Technology

Figure 1. Study area showing the extent of the Salish Sea, with dots representing those colonies where birds were collected (1860−2009) and feathers subsequently sampled for this study.

evapotranspiration.21 Little research has been conducted on the use of δ2H measurements to trace the relative use of marine and terrestrial/freshwater inputs to diets of animals, but we expected that a decline in feather δ2H values through time would be evidence that birds had consumed an increasing proportion of terrestrial- or freshwater-derived nutrients.22 Stable sulfur isotope measurements have long been recognized as useful indicators of marine versus terrestrial foods in diets of animals because marine sulfates are typically considerably enriched over terrestrial sources.23,24 Some exceptions to this are cases where terrestrial foodwebs may be based on volcanic soils, or in freshwater marshes where sulfate reduction occurs under anaerobic conditions.25 The advantages of using δ34S over δ13C measurements to trace animal use of marine resources is that δ34S values span a greater isotopic range (>20 ‰ vs ∼8 ‰) between these reservoirs; that plant δ34S values are not influenced by photosynthetic pathways, such as C3 versus C4; and that they are not significantly altered by trophic level23−27 (but, see ref 28). Feathers are a good source of the sulfur-bearing amino acid cysteine, and so δ34S analysis of feathers is expected to provide a faithful means of tracking protein pathways in birds. As with δ2H measurements, we expected a decline in gull feather δ34S values through time to indicate increases in terrestrial/ freshwater-derived protein during feather synthesis. Thus, to test the two hypotheses of gull diet change, we examined a subsample of the flight feathers used by Blight et al. (2014)9 for stable-hydrogen (δ2H) and stable-sulfur (δ34S) isotope values. We anticipated that long-term trends in these two isotopes would allow us to test whether gull diets had shifted to more terrestrial foods or to a diet including more marine invertebrates.21,22

nitrogen (δ15N) and -carbon (δ13C) isotope values consistent with a long-term shift toward more terrestrial foods (e.g., garbage) or the inclusion of increasing amounts of invertebrate marine prey. Those results were also consistent with the hypothesis that the marine foodweb on which the gulls depended historically had declined in quality (i.e., lower abundance, size-at-age, and lipid content of forage fish9,14,15), while human refuse at garbage dumps and agricultural development on the mainland increased the availability of terrestrial foods over time. However, δ15N and δ13C data could not rule out the alternative hypothesis that the gulls had primarily shifted their diets within the marine foodweb, from marine fish to lower trophic-level marine foods, such as mussels (Mytilus spp.) and sea stars (Asteroidea), given that terrestrial foods and marine invertebrates are both consumed by gulls in the region.13,16 Although less likely based on δ13C values of forage fish measured by Blight et al. (2014),9 δ13C values of marine prey may have also declined through this period due to the Seuss effect,17 that is, the isotopic depletion of the carbon in the atmosphere and oceans since the industrial revolution because of the burning of (isotopically light) fossil fuels. Fortunately, other stable isotope measurements can provide additional information on whether an organism has utilized marine versus terrestrial sources of foods. Stable hydrogen isotope analyses have proven to be useful for the investigation of movements and origins of wildlife at continental scales.18 This derives from the fact that δ2H values in precipitation vary primarily according to ambient temperature, distance from source, and elevation, according to well-understood kinetic and equilibration processes.19 In general, we expect terrestrial and freshwater foodwebs to be depleted in 2H compared to marine ones,20 although there are notable exceptions related to isotopic enrichment of terrestrial sources of water driven by 10835

DOI: 10.1021/acs.est.5b02053 Environ. Sci. Technol. 2015, 49, 10834−10840

Article

Environmental Science & Technology



by ref 35 for scaup (Aythya af finis; δ2Hf = 0.952, *δ2Hp = −27.8) provided an δ2Hf estimate of −96.3‰. For this terrestrial δ2Hf end point, we applied a ±12‰ error based on ref 36. The Fraser River (Figure 1), although showing seasonal fluctuations, shows values of approximately −135‰ near Vancouver33 so the above transfer function predicts that birds feeding exclusively on a lower Fraser River foodweb would have a mean δ2Hf value of −156.5 ± 12‰. We considered, then, a 3source (marine, terrestrial, Fraser), two-isotope (δ2H, δ34S) mixing model to estimate approximate inputs of these sources to “modern” gull diets (1980−2009 δ2H = −35.5 ± 15.5‰; δ34S = 14.8 ± 2.4‰; n = 12). MixSir was run using 1 000 000 iterations.

METHODS For δ2H and δ34S analysis, a subsample of 48 primary feather samples from adult birds previously analyzed for δ15N and δ13C by ref 9 was chosen systematically to span the period 1860− 2009. Feather samples came from birds collected in the Salish Sea (the inshore marine waters adjacent to southwest BC, Canada, and northwest Washington, USA; Figure 1) and archived in museum collections (all sources in Acknowledgments). Samples from 2008 to 2009 were the exception; these were collected as moulted feathers from the Glaucous-winged Gull colony at Mandarte Island, BC (48.63°N, 123.28°W). Early moult primary feathers (generally P129) were selected to represent diet at the beginning of the nesting season (late winter-early spring), when breeding birds are present at or near their nesting colonies. Feather segments were washed using a 2:1 chloroform:methanol solution to remove surface oils and air-dried according to methods described previously.9 Stablehydrogen isotope analyses of feathers were conducted using the comparative equilibration method30 through use of calibrated keratin δ2H reference materials. Stable-hydrogen isotope measurements were performed on H2 derived from hightemperature (1350 °C) flash pyrolysis of 350 ± 10 μg feather subsamples using continuous-flow isotope-ratio mass spectrometry (CFIRMS). Measurement of three keratin laboratory reference materials, CBS (Caribou hoof = −197 ± 0.79‰), SPK (Spectrum commercial = −121.6 ± 0.56‰), and KHS (Kudu horn = −54.1 ± 0.33‰) corrected for linear instrumental drift had typical mean within-run (n = 5) measurement variance of