Discussion:
Minnows caught in Ånnsjön generally ingested more zooplankton, compared to minnows caught in the adjacent streams, which ingested benthic invertebrates and terrestrial insects to a higher degree. Thus, my results are in line with previous findings on minnow diet (22, 42). Furthermore, morphology differed between minnows caught in the lake and streams. This habitat-specific body shape was associated with the use of a specific set of resources, indicating that minnow morphology is an adaption to a specific foraging mode. While the streamlined morphology of lake minnows can be seen as an advantage to swim in a uniform environment, stream minnows were characterized by a downward pointed snout and a deep body that could help feeding on benthic prey while maneuvering in the structurally complex stream habitat (20, 21), where zooplankton is typically scarce (33).
Resource polymorphism will lead to intraspecific divergence within a single population (15, 16), and can be seen as an early stage of speciation (6). As predicted from niche evolution theory, morphologically divergence will reduce competition as less prey items are shared (43, 44), and this pattern could also be demonstrated empirically in the perch-roach system (45). Intraspecific differentiation may initially emerge from phenotypic plasticity (46), and depending on the stability of the selective regime, divergent phenotypes may become genetically fixed (47, 48). Unfortunately, genetic data for the minnows of this study is not available. Future studies to investigate the level of genetic differentiation are needed to characterize the position of the morphotypes of the European minnow in Sweden along the specification continuum that could vary from adaptive variation to complete and reproductive isolation (6, 49). However, results of morphological divergence, but also resource use showed a strong variation within the habitats, between the different locations. Individuals caught at location S3 were feeding to a greater extent on terrestrial insects, compared to individuals caught in the other two stream locations. Furthermore, individuals caught at location L3 were feeding to a greater extent on benthic invertebrates and ingested fewer zooplankton than individuals caught at L1 and L2 and body shape were more similar to stream minnows. Such variation indicates a strong degree of plasticity in resource-morph formation, which would suggest that differences between minnows of the two adjacent habitats might not be genetically fixed. In contrast to the other lake habitats, water plants were abundant at L3, which could provide suitable microhabitats for benthic invertebrates. To understand the driving forces behind the variation in the degree of divergence, estimates of prey abundances at the different locations are needed.
In contrast to my results that are in accordance with the ones of Ramler, Palandacic (28), Collin and Fumagalli (27) found minnows inhabiting stream habitats in Switzerland to be more streamlined compared to conspecifics living in lakes. However, besides attributing these morphological adaptions to the hydrodynamic conditions occurring in stream habitats, they further reported a high predation pressure present in the lake habitats. A deeper body morphology can be seen as advantageous under such kind of ecological conditions, as muscle mass may enhance a rapid acceleration to escape predators (13, 50, 51). Only very few studies have tried to identify the specific factors that may shape the degree of variation in individual specialization (but see 52, 53, 54). Potentially, predation pressure could influence the strength and direction on the correlation of minnow morphology and diet, but further lab experiments are needed to resolve this relationship.
Scandinavian mountain lakes are characterized by a low species richness of fish. In many of these lakes, that are often remote, minnows were introduced as anglers used them as life bait (26). They are considered as being invasive, due to the fact that they can reach high densities, as for example, in Ånssjön (34). Næstad and Brittain (25) further showed that they have the ability to modify lake food webs, thus being responsible for a zoobenthos assemblage with a dominance of Chironomidae and Oligochaeta, and a low abundance of Gammarus lacustris . As lake minnows show a strong diet overlap with juvenile brown trout (24), they are also regarded as one of the factors contributing to the reduced recruitment and growth of the native brown trout in lake habitats (26). My results presented herein indicate that minnows inhabiting stream habitats may rely on different resources than the individuals inhabiting lakes. Therefore, interspecific competition target at different species in these contrasting habitats and patterns observed in lake habitats cannot be directly transferred to the interactions occurring in stream habitats. Nonetheless, the introduction of minnows into stream habitats may also pose a similar threat for the native fish populations of the stream, if a diet overlap would occur. Certainly, future studies need to determine the consequences of minnow invasions on the stream food webs.