Introduction
Natural selection can evoke adaptive phenotypic divergence of populations (1-3), which can lead to the formation of distinct populations or ecotypes (4, 5), and might even initiate speciation (6). In freshwater fish species, underlying ecological factors responsible for diversifying patterns in populations may include different predation regimes (7, 8), parasite occurrence (9), or different hydrodynamic conditions (10, 11). An environment that is characterized either by running or standing water may trigger divergence in fish populations with regard to morphological, physiological, developmental, or behavioral traits (12). To reduce the drag in the current, stream fishes often have a more streamlined body shape (13), which has been found, e.g. in pumpkinseeds (Lepomis gibbosus ) or rock bass (Ambloplites rupestris ) (14). Furthermore, divergence can be based on resource use, often referred to as trophic polymorphism (15, 16). Often, adaptations to different habitat types are intertwined with trophic niche partitioning (17). For example, in three-spined stickleback (Gasterosteus aculeatus ) (e.g. 18) and juvenile sockeye salmon (Oncorhynchus nerka ) (19), the lake ecotypes have a more streamlined body, as this morphology can be seen as beneficial to forage at relatively high speed and to cover larger areas in the lake habitat. In contrast, a deeper body generally supports a higher maneuverability in a structurally complex stream habitat when searching for the more cryptic benthic prey (20, 21).
In this study, I investigate the patterns of divergence in the European minnow (Phoxinus phoxinus ), a common freshwater fish that has paid little attention despite its large distribution (22, 23). It is an understudied fish species, despite its profound ecological impact when introduced to new areas, where it can become invasive and has the potential to modify native ecosystems (24-26). Minnows were recently found to show phenotypic adaptions in lake versus stream habitats: Collin and Fumagalli (27) studied minnow populations in Switzerland, and Ramler, Palandacic (28) investigated minnows in Northern Italy and the Danube basin. The studies found opposing results: Collin and Fumagalli (27) found stream minnows being more streamlined, a body form that is beneficial to reduce the drag in the current. In contrast, Ramler, Palandacic (28) reported that a streamlined body form was more pronounced in lake minnows compared to stream minnows and lake minnows also had larger heads compared to stream minnows. This might be due to habitat-induced changes in head structures linked to different modes of foraging, as it is known, for example from European perch (Perca fluviatilis ) (e.g. 29, 30). However, evidence on trophic niche divergence, incorporating morphological adaptations in minnows inhabiting lake versus stream habitats is missing.
In this study, I have analyzed stomach contents to understand the trophic niches of minnows in the lake Ånnsjön, Central Sweden and its tributaries. This method has the advantage to provide a direct insight into the foraging ecology, proving information on ingested prey with a high taxonomic resolution (31, 32). By combining the resource use assessment with morphological analyses by geometric morphometrics, I aimed to link the changes in body morphology to the individual resource use in the respective habitats. Streams are generally characterized by a lower abundance of zooplankton (e.g. 33). Therefore, I predict that in stream minnows, the dietary contribution of zooplankton would be lower compared to lake minnows. Furthermore, I predict that there is a relationship between morphological distance and dietary preference, indicating a specific morphology when consuming specific prey.