Rarity in boreal stream insects: patterns, causes and consequences Per-Ola Hoffsten Umeå 2003 Animal Ecology Department of Ecology and Environmental Science Umeå University Sweden ISBN 91-7305-529-8 © Per-Ola Hoffsten 2003 Printed by VMC, KBC, Umeå University Cover photos, and drawings in the summary, by the author. Organization Document name Umeå University DOCTORAL Animal Ecology DISSERTATION Department of Ecology and Environmental Science Date of issue SE-901 87 Umeå November 2003 Author: Per-Ola Hoffsten Title: Rarity in boreal stream insects: patterns, causes and consequences Abstract: Patterns of site occupancy among boreal stream insects were studied in central Sweden with focus on sparsely distributed species and the role of dispersal and niche limitations. In the study of dispersal limitation, I found that effects of an extraordinarily harsh winter in small to medium-sized streams were strongest in sites located in small streams and far from lake outlets. Species richness and the total abundance of macroinvertebrates and trout returned to pre-disturbance levels after three years. However, some species showed slow recolonization and the proportion of holoaquatic taxa was still reduced after three years. In a second study, I found a positive correlation between site occupancy in stream caddisflies and morphological traits associated with fast and energy-efficient flight, whereas specialized spring caddisflies showed a negative correlation to these traits compared to stream species. This suggested that streams, but not springs, select for strong dispersal ability in caddisflies. In a survey of springs in central Sweden, hydrogeology was found to be a useful predictor of the occurrence of spring specialists. Two of these, Crunoecia irrorata Curtis and Parachiona picicornis (Pictet), were found exclusively in glaciofluvial springs, characterized by a stable discharge and temperature. Less specialized members of the spring fauna (i.e. species also occurring in streams, ponds or lakes) also occurred in moraine and limestone springs characterized by more unstable conditions. Niche limitations were studied by contrasting large-scale distributions of closely related rare and common stoneflies. Differences in temperature requirements in the juvenile stages and life cycles suggested that the rare species, Isogenus nubecula Newman, was restricted by a limited tolerance to low stream temperatures, whereas the two common species, Isoperla grammatica (Poda) and Diura nanseni (Kempny), appeared to have a broader tolerance to climatic conditions in the study area. In a second study of niche limitations, macroinvertebrate assemblages in 88 streams in Central Sweden showed a nested distribution pattern. Most species deviating from expected distributions occurred in small streams, indicating competitive exclusion from species-rich sites, predator avoidance, or specialization to unique habitat features of small streams. In the last paper, the longitudinal distribution of filter-feeding caddisflies in a lake-outlet stream demonstrated patterns concordant to feeding specialization. Key words Central Sweden, site occupancy, dispersal limitation, natural disturbances, recolonization, flight-morphology, metapopulation, habitat stability, spring specialists, Crunoecia irrorata, niche limitation, Isogenus nubecula, nestedness, longitudinal distribution. Language: English ISBN: 91-7305-529-8 Number of pages: 136 Signature: Date: 7th November 2003 To Veronica and Dennis CONTENTS List of papers.....................................................................................................5 Introduction.......................................................................................................7 Study area..........................................................................................................9 Results and discussion ....................................................................................10 Dispersal limitations – The role of disturbances in stream communities.....................................10 – Dispersal abilities in adult caddisflies.....................................................11 – Stable habitats .........................................................................................12 Niche limitations – Rarity toward the geographical range border..........................................14 – Nested patterns in central Swedish streams ............................................15 – Longitudinal distribution of caddisflies in a lake-outlet stream..............16 Concluding remarks ........................................................................................17 Literature cited ................................................................................................18 Acknowledgements.........................................................................................23 Appendices: Papers I–VI LIST OF PAPERS This thesis is a summary and discussion of the following papers that will be referred to in the text by their Roman numerals. I. Hoffsten, P.-O. 2003. Effects of an extraordinarily harsh winter on macroinvertebrates and fish in boreal streams. – Archiv für Hydrobiologie 157: 505–523. II. Hoffsten, P.-O. Morphometric variation, dispersal ability and distribution in adult caddisflies. – Submitted manuscript. III. Hoffsten, P.-O. & Malmqvist, B. 2000. The macroinvertebrate fauna and hydrogeology of springs in central Sweden. – Hydrobiologia 436: 91–104. VI. Hoffsten, P.-O. & Malmqvist, B. Rarity and commonness in three perlodid stoneflies. –Manuscript. V. Malmqvist, B. & Hoffsten, P.-O. 2000. Macroinvertebrate taxonomic richness, community structure and nestedness in Swedish streams. – Archiv für Hydrobiologie 150: 29–54. VI. Hoffsten, P.-O. 1999. Distribution of filter-feeding caddisflies (Trichoptera) and plankton drift in a Swedish lake-outlet stream. – Aquatic Ecology 33: 377–386. Papers I, III, V and VI are reproduced with kind permission from the publishers. Summary 7 Introduction In population and community ecology rarity is often defined by species abundances and range sizes (Gaston, 1994). However, for organisms whose habitats are patchily distributed, e.g. stream insects, another important variable is the incidence of occupancy (frequency of occurrence). This variable is related to a species’ habitat specificity, which Rabinowitz (1981) used in her often quoted classification of rare species, since narrow habitat specificity might lead to a sparse distribution (Brown, 1984). Since I consider habitat specificity a cause of rarity, rather than a definition of it, I prefer to use occupancy (see also Kunin & Gaston, 1997). However, no generally accepted definition of rarity exists (Gaston & Lawton, 1990). Using the variables abundance, occupancy and range size, it is clear that species can be rare in several different ways. Out of eight possible combinations of the variables, seven characterize rare species (Tab. 1). It is also clear that species can be more or less rare, since all three variables are continuous, and that rarity can be measured at different scales: local, regional and global. My work has focused on causes behind low occupancy. This is partly because aquatic insects in my study area in central Sweden have wide geographical distributions, and partly because I soon got interested in how species’ different dispersal abilities could affect patterns of occupancy. Given that rarity is a relative state, rare species have been defined as those showing low site occupancy relative to other species among the studied animals. This has generally meant site occupancy less than 25 percent (cf. Gaston, 1994). The spatial scale of my studies ranges from regional to within a stream. Table 1. Eight different combinations of the factors local abundance, incidence of occupancy and geographic range size, of which only the first (locally abundant in a high proportion of sites over a large geographic range) characterize a common species. Modified after Rabinowitz (1981). Local abundance Occupancy Geographic range High High Large Low High Large High Low Large Low Low Large High High Small Low High Small High Low Small Low Low Small 8 Per-Ola Hoffsten The dominating organism group in boreal streams both in terms of number of species and number of individuals is often insects. Because most stream insects are amphibiotic, i.e. juvenile development takes place in the water whereas adults are terrestrial and often fully winged, both life-stages are important for distributions. While adult dispersal between habitat patches is a necessary first step in the colonization of a new site, the physiological tolerances and competition and predation sensitivity of larvae in relation to local conditions may limit the establishment of a population. Most stream insects show behavioural and physiological adaptations to seasonal variation in flow, water temperature and other factors typical of the boreal zone (Giller & Malmqvist, 1998). However, adaptations may not be enough in the face of unpredictable environmental conditions, i.e. disturbances (Resh et al., 1988), and local populations risk to go extinct from time to time. Amphibiotic stream insects have therefore, and because they are generally considered efficient dispersers, been suggested as candidates for metapopulations and metacommunites (Bohonak & Jenkins, 2003). According to metapopulation theory, local abundance and incidence of occupancy are interdependent (Hanski, 1982; Hanski et al., 1993). As dispersal rate is a key