Influence of reed (Phragmites australis) belts in the Baltic Sea archipelago on pike (Esox lucius) and other coastal fish species

Abstract: Eutrophication, near shore building and human disturbances by dredging and shore alteration have led to increased expansion of the common reed (Phragmites australis) in the archipelago of the Baltic Sea. Reed has an important ecological function such as nursery habitat for many fish species. Pike (Esox lucius) is a predatory fish whose larvae and young-of-the year fish find both food and shelter in coastal reed beds. But due to the increased amount of reed, more homogenous reed belts are formed, the overall biodiversity is reduced, and dense reed belts can reduce pike foraging. During the last decades, pike populations in the Baltic Sea have declined and are now mainly found in the inner bays of the archipelago but seem to have declined also in these core areas. No study has yet studied how pike abundance in inner archipelagos is related to reed characteristics like reed area, perimeter and heterogeneity. Here I study the impact of reed on abundance primarily of pike, but also of other coastal fish species: perch (Perca fluviatilis), roach (Rutilus rutilus) and three-spine stickleback (Gasterosteus aculeatus). More specifically I tested if more extensive and heterogeneous reed belts have more pike than homogenous reed belts. I conducted a spatial analysis for pike catch per unit effort (CPUE) from angling in relation to reed perimeter and area among 24 bays in the Stockholm archipelago. Pike CPUE was positively associated with both reed area and perimeter. The data showed that below 0.5 ha reed or a reed perimeter of 2500 m pike populations started to decline drastically, and there was no indication of lowered pike density in bays with the highest amounts of reed. Of the other coastal species, roach also showed a positive correlation with reed cover while perch abundance showed a positive correlation with pike abundance. Wave exposure was negatively correlated with pike and positively correlated with three-spined stickleback, indicating a transition zone between pike and sticklebacks along an exposure gradient. To study if reed management by cutting reed impacts pike populations, I did a angling survey in two coastal bays to test if pike utilized the more heterogeneous reed cut areas over homogenous reed belts. Unfortunately, too few pike were caught to allow statistical analysis, longer time series of pike abundance data are necessary. This study concluded that there is a positive association between pike abundance and reed, and there is a lower reed limit threshold for stable occurrence of pike. I could not find that very extensive reed belts would be negative for pike, nor that reed management by cutting reed would be beneficial but more data is required for a more certain conclusion.