Patterns of trophic niche divergence between invasive and native fishes in wild communities are predictable from mesocosm studies

Abstract

Ecological theory attempts to predict how impacts for native species arise from biological invasions. A fundamental question centres on the feeding interactions of invasive and native species: whether invasion will result in increased interspecific competition, which would result in negative consequences for the competing species, or trophic niche divergence, which would facilitate the invader's integration into the community and their coexistence with native species. Here, the feeding interactions of a highly invasive fish, topmouth gudgeon Pseudorasbora parva, with three native and functionally similar fishes were studied to determine whether patterns of either niche overlap or divergence detected in mesocosm experiments were apparent between the species at larger spatial scales. Using stable isotope analysis, their feeding relationships were assessed initially in the mesocosms (1000 L) and then in small ponds (<400 m(2) ) and large ponds (>600 m(2) ). In the mesocosms, a consistent pattern of trophic niche divergence was evident between the sympatric fishes, with niches shifting further apart in isotopic space than suggested in allopatry, revealing that sharing of food resources was limited. Sympatric P. parva also had a smaller niche than their allopatric populations. In eight small ponds where P. parva had coexisted for several years with at least one of the fish species used in the mesocosms, strong patterns of niche differentiation were also apparent, with P. parva always at a lower trophic position than the other fishes, as also occurred in the mesocosms. Where these fishes were sympatric within more complex fish communities in the large ponds, similar patterns were also apparent, with strong evidence of trophic niche differentiation. Aspects of the ecological impacts of P. parva invasion for native communities in larger ponds were consistent with those in the mesocosm experiments. Their invasion resulted in divergence in trophic niches, partly due to their reduced niche widths when in sympatry with other species, facilitating their coexistence in invaded ecosystems. Our study highlights the utility of controlled mesocosm studies for predicting the trophic relationships that can develop from introductions of non-native species into more complex ecosystems and at larger spatial scales.

Keywords: freshwater ecosystems; invasive species; stable isotope analysis; trophic niche width; trophic relationships.

Figure 1

Grey ellipses enclose the trophic niche of Cyprinus carpio (a), Gasterosteus aculeatus (b) and Tinca tinca (c) in the experimental mesocosms in the presence (solid grey) and absence (dashed grey) of Pseudorasbora parva. Black ellipses enclose the trophic niche of sympatric (solid black) and allopatric (dashed black) populations of P. parva. Note there is no overlap of the sympatric invasive P. parva (solid black) and native fish (solid grey) trophic niches in any of the treatments.

Figure 2

Food web structure in the eight small ponds, where the only fish present were Pseudorasbora parva (all ponds), Gasterosteus aculeatus (ponds a, b, c, d, g, h) and Tinca tinca (ponds e, f, g, h). Ellipses enclose the trophic niche of P. parva (solid black), G. aculeatus (solid grey) and T. tinca (dashed grey). Circular data points represent individual P. parva (closed black), G. aculeatus (closed grey) and T. tinca (open grey). Mean values (± standard error, n = 1–8) of resources are represented by open diamonds. Invertebrates included mayflies, Gammarus pulex, Asellus aquaticus and Chironomids. P.leni. = signal crayfish, Pacifastacus leniusculus. Note there is no overlap between invasive P. parva (black) and native fish (grey) trophic niches in any pond.

Figure 3

Food web structure in the three natural large pond communities in Belgium (a and b) and Wales (c). Ellipses enclose the trophic niche of Pseudorasbora parva (solid black), T. tinca (dashed grey), Cyprinus carpio (dashed black) and Gasterosteus aculeatus (solid grey). Data points represent individual P. parva (closed black), T. tinca (open grey), C. carpio (open black) and G. aculeatus (closed grey). Other community members are represented by means ± standard error (n = 3–14). C.aura. = goldfish, Carassius auratus; S.eryt. = common rudd, Scardinius erythropthalmus; C.gibe. = gibel carp, Carassius auratus; L.deli. = sunbleack, Leucaspius delineates; R.ruti. = roach, Rutilus rutilus; B.bjoe. = silver bream, Blicca bjoerkna; R.amar. = European bitterling, Rhodeus amarus. Note there is minimal niche overlap of invasive P. parva (solid black) and the model native fish, except C. carpio (dashed black), across all three ponds.