Influences of Gene Flow, Demographics, and Habitat Fragmentation

Research Collection

Doctoral Thesis

Habitat fragmentation of alpine streams: implications for genetic structure and species diversity of aquatic insects

Author(s): Monaghan, Michael Thomas

Publication Date: 2002

Permanent Link: https://doi.org/10.3929/ethz-a-004385646

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use.

ETH Library Diss. ETH No. 14561

Habitat Fragmentation of Alpine Streams: Implications for Genetic Structure and Species Diversity of Aquatic Insects

A Dissertation submitted to the ETH ZÜRICH (SWISS FEDERAL INSTITUTE OF TECHNOLOGY)

for the degree of Doctor of Natural Sciences

presented by

Michael Thomas Monaghan B.S. Nat. Res. Manage. The Ohio State University M.S. Biol. Idaho State University born 13 October, 1971 Columbus, Ohio, USA

accepted on the recommendation of

Prof. Dr. J.V. Ward, examiner Dr. M. Sartori, co-examiner PD Dr. P. Spaak, co-examiner PD Dr. C.T. Robinson, co-examiner

Zürich, 2002

Chapters 2 and 3 have been published:

Monaghan, M.T., P. Spaak, C.T. Robinson, and J.V. Ward. (2001). Genetic differentiation of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in fragmented alpine streams. Heredity 86:395-403.

Monaghan, M.T., P. Spaak, C.T. Robinson, and J.V. Ward. (2002). Population genetic structure of 3 Alpine stream insects: influences of gene flow, demographics, and habitat fragmentation. Journal of the North American Benthological Society 21: 114-131.

Chapter 6 is in press:

Monaghan, M.T., M. Hieber, C.T. Robinson, P. Spaak, and J.V. Ward. (In Press). Spatial patterns of Ephemeroptera, Plectopera, and Trichoptera species diversity in fragmented alpine streams. Verhandlungen Internationale Vereinigung für Theoretisch und Angewandte Limnologie 28. pour marie, qui m'a toujours fait sentir que j'appartenais ici

Table of Contents

Summary 1

Zusammenfassung 4

1. Introduction and outline of the thesis 7

2. Genetic differentiation of Baetis alpinus Pictet 15 (Ephemeroptera: Baetidae) in fragmented alpine streams

3. Population genetic structure of 3 alpine stream insects: 31 influences of gene flow, demographics, and habitat fragmentation

4. Local differentiation and regional homogeneity: lack of 59 equilibrium in the population genetic structure of an aquatic insect

5. Taxon richness and nestedness of benthic macroinvertebrates in 73 fragmented alpine streams: implications for freshwater conservation

6. Spatial patterns of Ephemeroptera, Plecoptera, and Trichoptera 93 diversity in fragmented alpine streams

7. Thesis conclusions and recommendations for future research 105

Curriculum vitae 111

Acknowledgements 115

SUMMARY

Habitat fragmentation of alpine streams, as defined herein, results from the presence of standing water bodies along stream courses that separate streams into discrete flowing reaches. The standing water bodies may be of natural (lakes) or anthropogenic (reservoirs) origin. This thesis consists of five studies that investigated the effects of alpine stream habitat fragmentation on the genetic structure and species richness of benthic macroinvertebrates. Three studies were conducted to examine the genetic structure of stream insect populations and two studies examined the composition of species assemblages. The first study used allozyme electrophoresis to examine the genetic structure of 23 populations of Baetis alpinus Pictet (Ephemeroptera: Baetidae) to determine whether lakes or reservoirs act as barriers to gene flow. Estimates of genetic differentiation (q) indicated little or no genetic difference between populations along 4 nonfragmented reference streams or across 2 lakes and 2 reservoirs, but marked differentiation across 5 lakes. Differentiation was unrelated to distance between fragments, but occurred only if lakes were situated in valleys that have been ice-free throughout the Holocene. If standing water bodies act as barriers to dispersal, the low q-values across geologically younger lakes and across reservoirs suggest that genetic differentiation was not detectable within the first 100 to 1000 years of habitat fragmentation. The second study analyzed 2 additional species of alpine stream insects and investigated the hierarchical genetic structure within and among the headwaters of major drainages of the Swiss Alps (Rhine, Inn, and Ticino rivers). Rhithrogena loyolaea (Heptageniidae) exhibited little genetic differentiation within and among streams but significant differentiation among drainages, suggesting that dispersal occurs among habitat fragments and among different streams but not across watersheds. Allogamus auricollis (Limnephilidae) did not exhibit genetic differentiation at any scale, suggesting that dispersal occurs throughout the geographical range of the study. In contrast, B. alpinus showed moderate to substantial differentiation among streams, but a distinct lack of genetic differentiation among major drainages. Although a definitive explanation for this surprising result is not immediately apparent, the data suggest a lack of equilibrium between gene flow and genetic drift in B. alpinus. This lack of equilibrium results from historical gene flow that continues to mask reduced dispersal. A synthesis of the results from these three species and from additional species in other studies indicated that four consistent 2 patterns of genetic differentiation emerge when multiple spatial scales are considered. The patterns are indicative of taxon-specific dispersal ability and reflect whether taxa are in gene flow - genetic drift equilibrium. The third study further examined the relative importance of historical and contemporary processes in determining genetic structure of B. alpinus. The study used amplified fragment length polymorphism (AFLP) with the hypothesis that a molecular DNA marker such as AFLP would be more sensitive to genetic change than allozyme markers, and would therefore detect more recent population differentiation. Interestingly, results were very similar for the two markers. Population differentiation occurred only at the geologically older lakes, and small-scale differentiation and large- scale homogeneity suggested a lack of genetic drift - gene flow equilibrium. An examination of the variance in q for pairwise comparisons of all populations revealed that variance was larger for AFLP than for allozymes. Variance was unrelated to geographic distance between populations, suggesting gene flow between populations can be quite limited regardless of distance. The study concluded that the difference in variance of q reflects two stages of range expansion following recolonization of the Alps after the retreat of glaciers. The less sensitive allozymes reflect rapid upstream gene flow into alpine streams from a few glacial refugia, while AFLP markers reflect more recent, small-scale local differentiation within and among streams. The fourth and fifth studies investigated the effects of natural and anthropogenic habitat fragmentation on macroinvertebrate communities among the same study sites used for the population genetics studies. Total taxon richness, all sites combined, was 69 ranging from 6 to 27 for individual sites. While total richness was not significantly different between fragmented and unfragmented sites, both Ephemeroptera and Diptera (excluding Simuliidae) richness was significantly reduced in stream fragments. Assemblages in fragments were not nested subsets of unfragmented sites, indicating that site- and stream-specific habitat conditions, rather than extinction and lack of subsequent recolonization, were primary determinants of assemblage structure in all sites. Collectively, the results suggest taxon turnover among sites and among streams (b-diversity) is an important component of biodiversity in alpine streams. The fifth study comprised a finer-scale study of how the presence of lake-outlet habitats may affect Ephemeroptera, Plecoptera, and Trichoptera (EPT) richness in fragmented streams. Richness and turnover generally increased along longitudinal gradients of 200-300 m that 3 encompassed sites above lakes, lake outlet sites, and sites farther downstream. Eight of ten fragmented sites had reduced EPT richness, but total richness was greater in a given fragmented stream than in an unfragmented reference stream. This finding supports the contention that

turnover at multiple spatial scales ( b -diversity) is an important measure of biodiversity in aquatic systems, and that lake outlets provide habitat conditions not present in unfragmented streams. One general conclusion from the combined studies is that dispersal ability is a species-specific trait; however, vagility generally is high for the benthic macroinvertebrates inhabiting alpine streams. The large number of species found only in a few widely separated fragments suggests they are capable of dispersal and that the suitability of local habitat conditions (e.g., flow and temperature regimes, disturbance, sediment structure, food resources, water velocity, water chemistry, competition, and predation) probably is the primary reason a species is present or absent in a fragment. A second conclusion concerned differences between natural and anthropogenic fragmentation in streams. Lakes and reservoirs had similar effects on taxon richness, but the effects on genetic differentiation were more pronounced in lakes. This suggests lakes and reservoirs can affect different aspects of total biodiversity and that the reduction of taxon diversity is a more immediate threat of reservoir construction. The thesis concludes with a discussion of potential avenues for future research on habitat fragmentation using population genetic techniques.

Zusammenfassung

Die Habitate in alpinen Flüssen werden durch stehende Gewässer fragmentiert, die sich im Flussverlauf befinden, diesen somit unterbrechen und in einzelne Abschnitte unterteilen. Bei diesen stehenden Gewässern kann es sich entweder um natürliche Seen oder um künstliche Stauseen handeln. Die vorliegende Arbeit beinhaltet fünf Studien, in denen die Effekte der Habitatfragmentierung in alpinen Flüssen auf die genetische Struktur und den Artenreichtum benthischer Makroinvertebraten untersucht wurden. Drei der Untersuchungen befassen sich mit der genetischen Struktur der Insektenpopulationen in Fliessgewässern, während sich zwei der Studien mit den Artenzusammensetzungen befassen. In der ersten Studie wurde die genetische Struktur von 23 Baetis alpinus- Populationen mit Hilfe von Allozymelektrophorese untersucht, um zu klären, ob natürliche Seen oder Stauseen den Genfluss begrenzen. Abschätzungen des Genflusses ergaben geringe oder keine genetische Differenzierung entlang vier unfragmentierter Referenzflüsse sowie über zwei natürliche Seen und zwei Stauseen, jedoch eine deutliche Differenzierung über fünf natürliche Seen. Die Differenzierung war dabei unabhängig von der geographischen Distanz zwischen den Fragmenten, zeigte sich jedoch ausschließlich bei Seen in Tälern, die während des Holozäns eisfrei waren. Wenn stehende Gewässer die Verbreitung der untersuchten Art begrenzen, so lassen die niedrigen q-Werte vermuten, daß eine genetische Differenzierung innerhalb der ersten 100-1000 Jahre nach der Habitatfragmentierung mittels Allozymen nicht erfassbar ist. In der zweiten Untersuchung wurden die Allozyme zweier weiterer Arten alpiner Fliessgewässerinsekten analysiert sowie die hierarchische genetische Struktur innerhalb und zwischen größeren Wassereinzugsgebieten der Schweizerischen Alpen (Rhein, Inn, Ticino) untersucht. Rithrogena loyolaea (Heptageniidae) wies innerhalb von und zwischen verschiedenen Flüssen geringe genetische Differenzierungen auf, zeigte jedoch zwischen den verschiedenen Wassereinzugsgebieten eine signifikante Differenzierung. Diese Ergebnisse lassen Rückschlüsse auf die Fähigkeit dieser Art zu, sich zwischen Habitatfragmenten und zwischen verschiedenen Flüssen auszubreiten. Allogamus auricollis (Limnephilidae) hingegen wies keine genetische Differenzierung auf, was auf eine Ausbreitung dieser Art über die gesamte geographische Reichweite dieser Studie hindeutet. Im Gegensatz dazu, zeigte B. alpinus eine leichte bis erhebliche genetische Differenzierung zwischen Flüssen, jedoch keine 5

Differenzierung zwischen den Wassereinzugsgebieten. Diese Ergebnisse deuten darauf hin, dass Genfluss und genetische Drift miteinander nicht im Gleichgewicht sind, und dass dieses Ungleichgewicht aus dem historischen Genfluss resultiert, der die reduzierte Verbreitung dieser Art verdeckt. Eine Synthese dieser Ergebnisse und anderer Publikationen deutet darauf hin, dass die genetische Struktur von Fliessgewässerinsekten vier Müstern folgt. Diese Muster kennzeichnen die Taxon-spezifische Verbreitungsfähigkeit und ob sich Genfluss und genetische Drift bei den jeweiligen Taxa im Gleichgewicht befinden. Die dritte Studie befasst sich mit der relativen Bedeutung historischer und gegenwärtiger Prozesse für die genetische Struktur von B. alpinus. Zu diesem Zweck wurden die Allozymdaten mit Ergebnissen eines zusätzlichen molekularen Markers (AFLP) verglichen. Die Muster der genetischen Differenzierung waren für beide Marker nahezu identisch. Dieses Ergebnis unterstreicht die Annahme eines begrenzten Genflusses über kürzere Distanzen. Die Varianzen von q waren unabhängig von der geographischen Distanz, waren jedoch für AFLP grösser als für die Allozyme. Die unterschiedlichen Varianzen von q scheinen zwei Stadien der Ausbreitung widerzuspiegeln, die auf die Wiederbesiedlung der Alpen nach dem Rückzug der Gletscher folgen. Die weniger sensitiven Allozyme geben einen schnellen, stomaufwärts gerichteten Genfluss aus wenigen

glazialen Rückzugsgebieten in alpine Flüsse wieder, während die AFLP - Marker eine jüngere, kleinräumige genetische Differenzierung innerhalb von und zwischen verschiedenen Flüssen zeigen. Die vierte und fünfte Untersuchung befassen sich mit den Effekten natürlicher und anthropogener Habitatfragmentierung auf Makroinvertebraten-Gemeinschaften. Der Artenreichtum betrug gemittelt über alle Probenahmestellen 69 und variierte zwischen 6 und 27 an den einzelnen Stellen. Während sich die unfragmentierten und fragmentierten Stellen hinsichtlich der Gesamtartenanzahlen nicht signifikant voneinander unterschieden, war der Artenreichtum der Ephemeroptera und Diptera (ohne Simuliidae) in fragmentierten Flüssen signifikant reduziert. Die Gemeinschaften in fragmentierten Habitaten waren jedoch keine Untergruppen der Gemeinschaften in unfragmentierten Habitaten. Dieses Ergebnis deutet darauf hin, daß in erster Linie Stellen- und Fluss- spezifische Habitatbedingungen für die Strukturen der Gemeinschaften verantwortlich waren und nicht Extinktion und eine mangelnde Rekolonisierung. Zusammenfassend legen die Ergebnisse nahe, dass der Taxon-Turnover zwischen den verschiedenen Stellen und Flüssen (b- 6

Diversität) von wesentlicher Bedeutung für die Biodiversität ist. In der fünften Studie wurden die Auswirkungen von Seeausflüssen auf den Artenreichtum von Ephemeroptera, Plecoptera und Trichoptera (EPT) in fragmentierten Flüssen untersucht. Artenreichtum und Turnover nahmen generell oberhalb von Seen flussabwärts, an Seeausflüssen und flussabwärts der Seen zu. Acht von zehn fragmentierten Flüssen wiesen einen reduzierten EPT-Artenreichtum auf. Die gesamte Artenanzahl war dabei jedoch in fragmentierten Flüssen grösser als in einem unfragmentierten Referenzfluss. Dieses Resultat unterstützt die Annahme, dass der Turnover in verschiedenen räumlichen Grössenordnungen ein wichtiges Mass für die Biodiversität in Fliessgewässern ist. Zusammenfassend lässt sich sagen, daß die Ausbreitung von Arten -- obwohl diese in Abhängigkeit von der Art sehr begrenzt sein kann -- ein weitverbreitetes Mekmal der meisten benthischen Makroinvertebraten- Gemeinschaften in Flüssen ist. Die hohen Anzahlen von Arten, die ausschliesslich in einigen weit voneinander entfernten Habitatfragmenten gefunden wurden, läßt eine ausgeprägte Verbreitungsfähigkeit dieser Arten vermuten. Lokale Habitatbedingungen (z.B. Abfluss- und Temperaturregime, Störungen, Sedimentstruktur, Nahrungsressourcen, Fliessgeschwindigkeit, Wasserchemie, Konkurrenz, Räuberdrück and Parasitismus) bestimmen vermutlich in erster Linie die Präsenz oder Abwesenheit einer Art in einem Habitatfragment. Die vorliegende Arbeit schliesst mit einer Diskussion möglicher zukünftiger Richtungen in der Erforschung von Habitatfragmention mit Hilfe populationsgenetischer Methoden ab. 7

CHAPTER 1

INTRODUCTION AND OUTLINE OF THE THESIS

'There have been many kings of Babylon who helped to fortify the city…There were also two queens…the later, Nitocris,… constructed em- bankments on both sides of the river of remarkable strength and height, and a long way above the city, close beside the river, dug a basin for a lake some forty-seven miles in circumference…The purpose of the excavation and of the diversion of the river was to…prevent a direct voyage downstream to the city. A boat would be faced with a devious course, and at the end of her trip she would have to make the tedious circuit of the lake.' Herodotus, The Histories (from the translation by Aubrey de Sélincourt, revised by John Marincola, Penguin Classics, 1996, London, p. 72-73).

In his discussion of the influence of geography on the history of the Mediterranean region, Braudel (1972) discussed how the earliest human inhabitants long remained in the mountains "...because the plains were originally a land of stagnant waters and malaria, or zones through which the unstable river beds passed," and that "...the thickly populated plains which today are the image of prosperity were the culmination of centuries of painful collective effort." Indeed, humans have manipulated and altered streams and rivers in order to control and use the aquatic resources for thousands of years. It is comparatively recent that a large-scale, collective effort has been made to understand how the alteration of these aquatic ecosystems affects their natural functioning and biological characteristics (see Hynes 1970, Ward and Stanford 1979). One prominent means by which humans have altered streams and rivers is through the construction of dams. The construction of large dams (> 15 m high) to create reservoirs for water storage in Europe dates back to 8 at least the 2nd century A.D. in Spain (Leonard and Crouzet 1999). By 1900 the number had grown to approximately 400, and today there are more than 3300 large dams and major reservoirs located throughout Western Europe, predominantly used for hydropower, water supply, irrigation, and transport (Leonard and Crouzet 1999). In the simplest sense, the construction of dams and reservoirs alters the natural physical structure of the stream or river on which it was built by creating an area of standing water and by changing, reducing, or halting the flow of water. One notable feature of streams and rivers is that a very similar alteration of physical connectivity occurs naturally by the formation of lakes. Lakes are formed by a variety of geological processes, including landslides, glacial activity (ice-scour, cirque, and paternoster lakes) and deposition (moraine-dammed lakes, kettle lakes), volcanism, and tectonic action (Wetzel 2001). This thesis attempts to investigate how the fragmentation of streams into discreet flowing reaches by natural and anthropogenic standing water bodies affects stream organisms living in "fragments" of flowing water.

A Definition of Habitat Fragmentation Habitat fragmentation is the process by which large, continuous areas of habitat are divided into smaller, often discreet patches (Saunders et al. 1991). Habitat fragmentation occurs at both evolutionary and ecological time scales, resulting from both natural and anthropogenic processes. Long-term, natural processes include Pleistocene climate warming, causing the retreat of boreal plants and animals upward to isolated mountain peaks (Brown 1971, Templeton et al. 1990) and the isolation of islands following Pleistocene sea-level rise (Cox and Moore 2000). Short- term processes of fragmentation often are human-caused, such as cutting of tropical rainforests (Lovejoy et al. 1986, Laurance and Bierregaard 1997) and urbanization of the landscape (Bolger et al. 1997).

Biological Consequences of Habitat Fragmentation Natural habitat fragmentation can result in allopatric speciation and endemism (reviewed by Maynard Smith 1975). Anthropogenic habitat fragmentation is considered to be a major threat to the preservation of natural biodiversity (sensu Noss 1990, Johnson et al. 1996) because of its 9 rapid rate of occurrence and multiple harmful effects on natural ecological systems. Two important effects of habitat fragmentation on biodiversity are alterations to natural levels of species diversity of communities and to levels of genetic diversity of populations.

Changes in species richness and assemblage structure Two well-documented effects of habitat fragmentation are the reduction of species richness (Wilcox and Murphy 1985, Klein 1989, Newmark 1991, Kattan et al. 1994) and the change in assemblage structure (Margules et al. 1994, Davies and Margules 1998) within fragments. These changes are caused by a variety of inter-related processes, including an increase in the proportion of edge habitat, a reduction in habitat area, and an isolation of habitat fragments (Saunders et al. 1991, Andren 1994). These three processes are potentially confounding (Robinson et al. 1992); therefore, their inter-relationships require careful investigation. There is a great deal of theoretical and empirical work applicable to the study of habitat fragmentation and its effects on organisms. For example, reduced species richness in fragments is, in part, a function of species-area relationships and the fact that larger areas contain more kinds of habitat (Rosenzweig 1995, Bender et al. 1998). An increase in the amount of edge relative to the amount of interior habitat can create an "edge effect." The alterations to habitat near the edge of fragments, such as changes to light and temperature regime, can reduce species richness and alter assemblage structure (e.g., Lovejoy et al. 1986) because of the close relationship of most species with their habitat (Southwood 1977). Investigating the influence of habitat alteration on species assemblages is the subject of a large body of work in many ecosystems, including streams and rivers (Ward 1992). Isolation of fragments also can result in the formation of subpopulations which may have a greater probability of extinction because of environmental or demographic stochasticity (Pimm et al. 1995). The resulting metapopulation may become more dependent on dispersal among subpopulations for persistence (Stacey and Taper 1992, Hanski 1998). Naturalists, ecologists, and geneticists have long understood that most species are not homogeneously distributed across their biogeographic range, but rather are spatially clumped and often divided into subpopulations (Wright 1931, Nicholson 1933, Pianka 1983, Cox and Moore 2000). The theory of island biogeography (MacArthur and Wilson 1967) predicts that species number on oceanic islands is a function of 10 island size and distance from the mainland source when extinction and immigration of populations are in equilibrium. The metapopulation concept (Levins 1970, Hanski and Gilpin 1997) considers the interactions of local population dynamics and regional dispersal and how they affect population persistence.

Changes in the genetic structure of populations The reduction of total habitat area and the isolation of habitat fragments can create smaller, more isolated subpopulations. This process can decrease effective population size and lead to a loss of genetic variation through the process of genetic drift (Lacy 1987, Frankham 1997, van Dongen et al. 1998, Morden and Loeffler 1999). Reduced genetic diversity, in turn, can increase the probability of local extinction (Saccheri et al. 1998) through mechanisms of inbreeding depression (Charlesworth and Charlesworth 1987) or by reducing the natural level of adaptive variation (Allendorf and Leary 1986). Gene flow, resulting from the dispersal of organisms among habitat fragments, can counteract the reduction of genetic diversity in subpopulations that occurs by genetic drift (Lacy 1987, Grant and Grant 1992, Stangel et al. 1992). By examining the spatial distribution of neutral genetic variation (i.e., variation that is not influenced by natural selection), we can estimate the relative importance of gene flow and drift in determining population genetic structure (Slatkin 1985). On the one hand, it provides measures of genetic diversity and the consequences of reduced effective population size. On the other hand, it provides estimates of gene flow among subpopulations, and can provide insights as to how important the dispersal of individuals is for maintaining species populations in fragmented habitats (Stacey and Taper 1992, Hastings and Harrison 1994, Hanski 1998).

Research Objectives and Contents of the Thesis Fragmentation by standing water bodies is one of many forms of habitat fragmentation that threaten to alter the natural biodiversity patterns of streams and rivers (Zwick 1992, Allan and Flecker 1993, Dynesius and Nilsson 1994, Power et al. 1996, Ward and Tockner 2001). The abundance of streams, lakes, and reservoirs in the Swiss Alps provides ample opportunity to study the effects of habitat fragmentation on the genetic structure and species richness of stream benthic 11 macroinvertebrates. The goal of this thesis was to examine the effect of natural and anthropogenic stream habitat fragmentation on these two aspects of biodiversity. The first set of objectives of the thesis concerned the effects of fragmentation on genetic diversity and gene flow. The following specific questions were addressed: (1) is the genetic diversity of populations reduced in fragments and, if so, is the reduction related to dispersal ability, the distance between fragments, and the type and age of the fragmenting feature (i.e. natural lakes formed following glacial retreat and anthropogenic reservoirs constructed in the 1900s), and (2) is gene flow reduced among habitat fragments and, if so, is reduction related to these same features? I attempted to answer these questions beginning with a study of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in chapter 2. Two lines of evidence suggested that historical processes were contributing to the genetic structure of B. alpinus: a lack of genetic differentiation across reservoirs and the combination of small-scale differentiation and large-scale homogeneity. Having found high levels of genetic differentiation between populations separated by lakes, I extended the investigation in chapter 3 with a study of two additional species, Rhithrogena loyolaea Navàs (Ephemeroptera: Heptageniidae) and Allogamus auricollis Pictet (Trichoptera: Limnephilidae), both of which have presumably better dispersal abilities. Chapter 4 reports an investigation using amplified fragment length polymorphism (AFLP) to examine whether genetic drift and gene flow are in equilibrium for B. alpinus. The second set of objectives of the thesis concerned the effect of fragmentation on species richness. The questions asked were (1) is taxon richness reduced in habitat fragments relative to unfragmented streams, (2) are the effects of lakes and reservoirs on taxon richness different, (3) how important are the processes of isolation and habitat alteration for determining the assemblage composition in fragments, and (4) how does small-scale taxon turnover in fragmented streams affect taxon richness along the stream course. The study reported in chapter 5 addressed the first three questions by examining the reduction in taxon richness in lake- and reservoir-fragmented streams and used an analysis of assemblage nestedness to examine the relative importance of isolation and habitat alteration. To further examine small-scale taxon turnover, chapter 6 investigated the specific effect of lake-outlet habitat on total taxon richness in fragmented streams. The thesis concludes with chapter 7, providing a 12 synthesis of the results and an attempt to answer some general ecological questions regarding the effect of habitat fragmentation on the genetic structure and species richness of stream macroinvertebrates.

Literature Cited Allan, J. D., and A. S. Flecker. 1993. Biodiversity conservation in running waters. BioScience 43:32-43. Allendorf, F. W., and R. F. Leary. 1986. Heterozygosity and fitness in natural populations of animals. Pages 57-76 in M. E. Soulé (editor). Conservation biology: the science of scarcity and diversity. Sinauer Associates, Sunderland, Massachusetts. Andren, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: a review. Oikos 71:355-366. Bender, D. J., T. A. Contreras, and L. Fahrig. 1998. Habitat loss and population decline: a meta-analysis of the patch size effect. Ecology 79:517-533. Bolger, D. T., A. C. Alberts, R. M. Sauvajot, P. Potenza, C. McCalvin, D. Tran, S. Mazzoni, and M. E. Soulé. 1997. Response of rodents to habitat fragmentation in coastal southern California. Ecological Applications 7:552-563. Braudel, F. 1972. The Mediterranean and the Mediterranean world in the age of Philip II (translated from the French by Siân Reynolds). Harper and Row (Torchbooks), New York. Brown, J. H. 1971. Mammals on mountaintops: nonequilibrium insular biogeography. American Naturalist 105:467-478. Charlesworth, D., and B. Charlesworth. 1987. Inbreeding depression and its evolutionary consequences. Annual Review of Ecology and Systematics 18:237-268. Cox, C. B., and P. D. Moore. 2000. Biogeography: an ecological and evolutionary approach, 6th edition. Blackwell Science, Oxford. Davies, K. F., and C. R. Margules. 1998. Effects of habitat fragmentation on carabid beetles: experimental evidence. Journal of Animal Ecology 67:460-471. Dynesius, M., and C. Nilsson. 1994. Fragmentation and flow regulation of river systems in the northern third of the world. Science 266:753-762. Frankham, R. 1997. Do island populations have less genetic variation than mainland populations? Heredity 78:311-327. Grant, P. R., and B. R. Grant. 1992. Darwin's finches: genetically effective population sizes. Ecology 73:766-784. Hanski, I. 1998. Metapopulation dynamics. Nature 396:41-49. Hanski, I., and M. E. Gilpin. 1997. Metapopulation biology: ecology, genetics, and evolution. Academic Press, San Diego. Hastings, A., and S. Harrison. 1994. Metapopulation dynamics and genetics. Annual Review of Ecology and Systematics 25:167-188. Herodotus. 1996. The histories (from the translation by A. de Sélencourt, revised by J. Marincola). Penguin Classics, London. Hynes, H. B. N. 1970. The ecology of running waters. Liverpool University Press, Liverpool. Johnson, K. H., K. A. Vogt, H. J. Clark, O. J. Schmitz, and D. J. Vogt. 1996. Biodiversity and the productivity and stability of ecosystems. Trends in Ecology and Evolution 11:372-377. Kattan, G. H., H. Alvarez-Lopez, and M. Giraldo. 1994. Forest fragmentation and bird extinctions: San Antonio eighty years later. Conservation Biology 8:138-146. Klein, B. C. 1989. Effects of forest fragmentation on dung and carrion beetle communities in central Amazonia. Ecology 70:1715-1725. Lacy, R. C. 1987. Loss of genetic diversity from managed populations: interacting effects of drift, mutation, immigration, selection, and population subdivision. Conservation Biology 1:143-158. 13

Laurance, W. F., and R. Bierregaard. 1997. Tropical forest remnants: ecology, management, and conservation. University of Chicago Press, Chicago. Leonard, J., and P. Crouzet. 1999. Lakes and reservoirs in the EEA area: topic report 1/1999. European Environment Agency, Office for Official Publications of the European Communities, Luxembourg. Levins, R. 1970. Extinction. Pages 75-107 in M. Gerstenhaber (editor). Some mathematical questions in biology; lecture notes on mathematics in the life sciences. The American Mathematical Society, Providence, Rhode Island. Lovejoy, T. E., A. B. Bierregaard, A. B. Rylands, J. R. Malcolm, C. E. Quintela, L. H. Harper, K. S. Brown, A. H. Powell, G. V. N. Powell, H. O. R. Schubart, and M. B. Hays. 1986. Edge and other effects of isolation on Amazon forest fragments. Pages 257-285 in M. E. Soulé (editor). Conservation biology: the science of scarcity and diversity. Sinauer Associates, Sunderland, Massachusetts. MacArthur, R. M., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, New Jersey. Margules, C. R., G. A. Milkovits, and G. T. Smith. 1994. Contrasting effects of habitat fragmentation on the scorpion Cercophonius squama and an amphipod. Ecology 75:2033- 2042. Maynard Smith, J. 1975. The theory of evolution. Cambridge University Press, Cambridge. Morden, C. W., and W. Loeffler. 1999. Fragmentation and genetic differentiation among subpopulations of the endangered Hawaiian mint Haplostachys haplostachya (Lamiaceae). Molecular Ecology 8:617-625. Newmark, W. D. 1991. Tropical forest fragmentation and the local extinction of understory birds in the eastern Usambara Mountains, Tanzania. Conservation Biology 5:67-78. Nicholson, A. J. 1933. The balance of animal populations. Journal of Animal Ecology 2:132- 178. Noss, R. F. 1990. Indicators for monitoring biodiversity: a hierarchical approach. Conservation Biology 4:355-364. Pianka, E. R. 1983. Evolutionary ecology, 3rd edition. Harper and Row, New York. Pimm, S. L., G. J. Russell, J. L. Gittleman, and T. M. Brooks. 1995. The future of biodiversity. Science 269:347-350. Power, M. E., W. E. Dietrich, and J. C. Finlay. 1996. Dams and downstream aquatic biodiversity: potential food web consequences of hydrologic and geomorphic change. Environmental Management 20:887-895. Robinson, G. R., R. D. Holt, M. S. Gaines, S. P. Hamburg, M. L. Johnson, H. S. Fitch, and E. A. Martinko. 1992. Diverse and contrasting effects of habitat fragmentation. Science 257:524-526. Rosenzweig, M. L. 1995. Species diversity in space and time. Cambridge University, Cambridge. Saccheri, I., M. Kuussaari, M. Kankare, P. Vikman, W. Fortelius, and I. Hanski. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494. Saunders, D. A., R. J. Hobbs, and C. R. Margules. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology 5:18-32. Slatkin, M. 1985. Gene flow in natural populations. Annual Review of Ecology and Systematics 16:393-430. Southwood, T. R. E. 1977. Habitat, the templet for ecological strategies? Journal of Animal Ecology 46:337-365. Stacey, P. B., and M. Taper. 1992. Environmental variation and the persistence of small populations. Ecological Applications 2:18-29. Stangel, P. W., M. R. Lennartz, and M. H. Smith. 1992. Genetic variation and population structure of red-cockaded woodpeckers. Conservation Biology 6:283-292. Templeton, A. R., K. Shaw, E. Routman, and S. K. Davis. 1990. The genetic consequences of habitat fragmentation. Annals of the Missouri Botanical Garden 77:13-27. 14 van Dongen, S., T. Backeljau, E. Matthysen, and A. A. Dhondt. 1998. Genetic population structure of the winter moth (Operophtera brumata L.) (Lepidoptera, Geometridae) in a fragmented landscape. Heredity 80:92-100. Ward, J. V. 1992. Aquatic insect ecology 1: biology and habitat. John Wiley and Sons, New York. Ward, J. V., and J. A. Stanford, editors. 1979. The ecology of regulated streams. Plenum, New York. Ward, J. V., and K. Tockner. 2001. Biodiversity: towards a unifying theme for river ecology. Freshwater Biology 46:807-819. Wetzel, R. G. 2001. Limnology, 3rd edition. Academic Press, San Diego. Wilcox, B. A., and D. D. Murphy. 1985. Conservation strategy: the effects of fragmentation on extinction. American Naturalist 125:879-887. Wright. 1931. Evolution in Mendelian populations. Genetics 16:97-159. Zwick, P. 1992. Stream habitat fragmentation - a threat to biodiversity. Biodiversity and Conservation 1:80-97. 15

CHAPTER 2

GENETIC DIFFERENTIATION OF BAETIS ALPINUS

PICTET (EPHEMEROPTERA: BAETIDAE) IN FRAG-

MENTED ALPINE STREAMS

with P. Spaak, C. T. Robinson, and J.V. Ward (2001) Heredity 86:365-403.

Abstract The interpretation of low q-values as evidence of high levels of gene flow among habitat fragments may be confounded by population genetic structures that are indicative of historical rather than present-day levels of gene flow. We examined the genetic structure of 23 populations of Baetis alpinus (Insecta: Ephemeroptera) living in alpine streams fragmented by lakes (ca 10,000 years old), reservoirs (ca 100 years old), and in unfragmented streams, to examine if lakes act as barriers to gene flow and to investigate the temporal resolution of allozyme markers. Estimates of gene flow indicated little or no genetic divergence along 4 unfragmented reference streams and across 2 lakes and 2 reservoirs (q = 0.004 to 0.041), but marked differentiation across 4 lakes (q = 0.092 and 0.362) and across one reservoir that was a lake enlarged by a dam (q = 0.075). Differentiation was unrelated to distance between fragments, but occurred only if lakes occurred in valleys that have been ice-free throughout the Holocene. We suggest that standing water bodies act as barriers to gene flow in B. alpinus and that low q values observed between fragments separated by reservoirs do not indicate high levels of gene flow but rather show that genetic differentiation was not detectable within the first 100 to 1000 years of habitat fragmentation. 16

Introduction Habitat fragmentation occurs at both evolutionary and ecological time scales, resulting from both natural and anthropogenic processes. Long- term, natural processes include Pleistocene climate warming and the retreat of boreal plants and animals upward to isolated mountain peaks (Brown 1971, Templeton et al. 1990). Short-term processes of fragmentation often are human-caused, such as clearcutting of forests and urbanization. Consequences include the loss of total habitat area and isolation of fragments. Isolation can reduce dispersal among fragments, increase local extinction (Saunders et al. 1991), and lead to creation of a metapopulation (Hastings and Harrison 1994, Hanski 1998). Both simulation (Lacy 1987) and empirical (van Dongen et al. 1998) studies document a decrease in genetic diversity among fragmented populations which potentially can reduce fitness and cause local extinction (Saccheri et al. 1998). Reduced genetic diversity may be the result of decreased population size and decreased gene flow (Frankham 1997), both of which reduce the effective population size. Accordingly, high vagility often is invoked to explain a lack of genetic differentiation among fragmented populations (e.g. Stangel et al. 1992, Hickerson and Wolf 1998, Ramirez and Haakonsen 1999). The multiple time scales over which fragmentation occurs potentially can confound the interpretation of population genetic structure because current population genetic structure may not reflect current levels of gene flow (Larson et al. 1984, Bossart and Prowell 1998). Much habitat fragmentation is human-caused and therefore relatively recent, occurring at time scales of 10s to 100s of years. Studies of recent fragmentation that detect no differentiation and conclude high levels of gene flow (e.g. studies cited above) may be observing the genetic signature resulting from historical rather than current levels of gene flow. This is because genetic drift may be too slow to result in detectable differentiation among populations in landscapes recently altered by humans. Such temporal resolution of genetic markers is of general importance to population genetics and requires further investigation (Bossart and Prowell 1998). Alpine streams provide a good opportunity to examine different time scales at which differentiation may be observed in the genetic signature of fragmented populations. Alpine streams often are fragmented by the presence of lakes or reservoirs that may act as barriers to gene flow, as these standing water habitats are unsuitable for most organisms adapted to running water. Many alpine lakes were formed following glacier retreat at 17 the end of the Pleistocene (i.e. 1000s to 10,000 years ago), while most reservoirs were constructed for hydroelectric power generation in the 20th Century (e.g. 10s to 100 years ago). As such, fragments of different age are produced by similar processes. For stream insects, gene flow is achieved via larval drift (downstream transport with the current) and adult flight after emergence from the water. The fragmentation of streams by standing water may restrict downstream larval drift, upstream adult flight, or both. The result can be the genetic differentiation of populations upstream and downstream of standing water bodies. Alternatively, gene flow may be inversely related to the distance animals must travel, resulting in a negative relationship between genetic differentiation and distance between fragments. We used allozyme electrophoresis to investigate population genetic structure of the mayfly, Baetis alpinus Pictet, 1843 (Ephemeroptera: Baetidae), in alpine streams in Switzerland. We first examined whether genetic diversity was reduced in fragmented populations relative to populations living in unfragmented streams. We then analyzed whether lakes and reservoirs fragment populations and act as barriers to gene flow in a stream insect, with a focus on how habitat fragmentation at various temporal scales may be reflected by genetic population structure. Our primary hypothesis was that genetic differentiation would be evident at only evolutionary time scales, i.e. would occur between populations separated by lakes but not between populations separated by reservoirs. Our alternative hypothesis was that gene flow was reduced between fragments separated by greater distances, regardless of time since fragmentation.

Methods

Study sites Fragmentation was studied at 13 headwater sites of the Rhine, Inn, and Ticino Rivers in the Swiss Alps (Fig. 1). Each site consisted of paired sampling locations upstream and downstream of a potential dispersal barrier (i.e. lake or reservoir) or at comparable distances along an unfragmented reference stream (Fig. 1). Six streams were fragmented by a natural lake and 3 streams were fragmented by reservoirs (Table 1). One of the reservoirs (Ritom) is a natural lake that has been enlarged by a dam (Knoll-Heinz 1991). Four study sites were unfragmented alpine streams. 18

Switzerland

1 Arosa Rhine R. 10 5,6 Upper Julierpass 1 Silvaplana 7 San Bernadino 1 Inn R. 4,9 Schwellisee 8 3 Bianco 2 2 Ticino R. Minor 11,12 Marmorera 1 13 Livigno 2 2 N Upper/Lower Jöri Cadagno/Ritom 50 km 2

Figure 1. Map of the study sites in Switzerland and schematics of the paired sampling design for lakes and reservoirs (ovals), and for unfragmented streams (line). Sites are designated as: 1-Schwellisee, 2-Bianco, 3-Minor, 4-Cadagno, 5-Upper Jörisee, 6-Lower Jörisee, 7-Livigno, 8-Marmorera, 9-Ritom, 10-Arosa, 11-Julierpass, 12-Silvaplana, 13-San Bernadino.

There were no tributaries entering any of the study streams between sample sites. For clearer presentation of results, upstream sampling locations at each site are designated with a 1 and downstream locations a 2 (e.g. Cadagno-1 and Cadagno-2 are upstream and downstream of Lake Cadagno, respectively). Three sampling locations were shared by multiple study sites and thus used in 2 separate analyses: Cadagno-2 occurred between Lake Cadagno and the reservoir Ritom, and Upper Jörisee-2 occurred between the upper and lower Jörisee lakes. The third location (Julierpass-1) was used for comparison along both a short reference reach (Julierpass) and a long reference reach (Silvaplana) by pairing it with 2 different downstream sampling locations. Upstream sample locations ranged in elevation from 1700 to 2525 m a.s.l., with all but 4 locations occurring above 1900 m a.s.l. (Table 1). Water-surface distance between fragments ranged from 280 to 10,000 m and elevation difference between fragment locations ranged from 4 to 250 m (Table 1).

Study animal The mayfly Baetis alpinus is a widespread and abundant alpine species (Humpesch 1979, Breitenmoser-Würsten and Sartori 1995) occurring in 1st to 4th order streams between 200 and 2600 m in elevation 19

Table 1. The 13 study sites examined in the present study. Each site consists of paired sampling locations upstream and downstream of the fragmenting feature or along the unfragmented reference stream. Lake and reservoir names are from maps. Reference streams were unnamed and therefore designated by location. Study Site Distance between Elevation of upper Elevation change between fragments location fragments (m) (m) (m) Lakes Schwellisee 350 1935 5 Lago Bianco 525 2080 4 Puox Minor 375 2340 15 Lago Cadagno 1075 1940 40 Upper Jörisee 550 2525 30 Lower Jörisee 975 2495 175

Reservoirs Livigno 10000 1910 155 Marmorera 7750 1700 250 Ritom 4500 1900 120

Unfragmented Arosa 375 1940 10 Julierpass 625 2310 105 Silvaplana 3225 2310 220 San Bernadino 280 2225 25 that do not exceed 20 °C (Sartori and Landolt 1999). The life cycle of B. alpinus is quite plastic and is dependent largely upon environmental conditions, namely temperature and elevation. Humpesch (1979) observed a bivoltine life cycle in streams at 615 m and a univoltine life cycle at 1355 m in Austria, while Lavandier (1988) observed a univoltine life cycle in streams at 1920 m and semi-voltine life cycle at 2190 m in France. Periods of emergence and flight typically are asynchronous and extend several months (Humpesch 1979, Kukula 1997). Upstream bias in adult flight has been observed in B. alpinus (Thomas 1975, Lavandier 1982) and Hershey et al. (1993) reported flight distances of 1.6 to 1.9 km for another species of Baetis.

Field collection and allozyme electrophoresis Between 30 and 50 late instar larvae were collected on a single day from each sampling location using a kick-net (250 µm mesh), kept alive for 1-2 hours in stream water, and then flash-frozen in liquid nitrogen. In the laboratory, larvae were thawed and tentatively identified using a dissecting microscope (60x). Head capsules then were removed and preserved in 20

Table 2. Enzyme systems, presumptive loci scored, and observed subunit structure for Baetis alpinus. Peptidase substrates were Leu-Gly-Gly, Lue-Ala, and Phe-Pro for Pep-B, C, and D, respectively. Buffer systems are those designated by Richardson et al. (1986). Locus No. Alleles Subunit Structure E.C. Number Buffer System Mpi 9 Monomer 5.3.1.8 A Pgi 7 Dimer 5.3.1.9 I Pgm 5 Monomer 2.7.5.1 I Pep-B 7 Dimer 3.4.11 or 13 I Pep-C 8 Dimer 3.4.11 or 13 I Pep-D 8 Dimer 3.4.11 or 13 I

70% ethanol solution for later final taxonomic distinction by examining the mandibles with a dissecting microscope and the maxillary palpae with a light microscope (400x). The remainder of each animal was ground in 80 mL crushing buffer (diH2O, NADP, b-mercaptoethanol). Using cellulose acetate electrophoresis and the methods and stain recipes of Hebert and Beaton (1989), we screened 25 enzyme systems using individuals from a subset of sampling sites to identify polymorphic loci that could be scored consistently. We found 6 satisfactory enzyme systems and identified 44 alleles among 6 presumptive loci (Table 2). The final data analysis was based on at least 25 animals from each of 23 sampling locations (all data, including sample size for each locus and allele and genotype frequencies are available from the authors).

Data analysis We examined genetic diversity using the mean number of alleles per locus, % polymorphic loci (95% criterion), and observed and expected

Hardy-Weinberg heterozygosity values (HWobs and HWexp , respectively). All values were calculated using GENEPOP (version 3.1d; Raymond and

Rousset, 1995). Mean number of alleles per locus, HWexp , and HWobs were compared between locations in fragmented (n = 16) and unfragmented (n = 7) streams using ANCOVA on all data from each locus (i.e., not mean values) with locus as the covariant. Percent polymorphic loci was compared using a student’s t-test on transformed (arcsin square root) data. Genotype diversity was calculated using Simpson’s (1 – S (p2)) index and also compared using a students’ t -test. Deviations from H -W equilibrium were examined by calculating an inbreeding coefficient (f ). Significant difference from zero was determined using the probability test (complete enumeration or the Markov chain method depending on the number of alleles) available in GENEPOP. A Bonferroni correction was applied to 21 adjust the p-value of significance to account for the 138 tests (23 populations, 6 loci). Multi-locus values of q were computed for the paired sampling locations at each site using GENEPOP. Levels of population differentiation were determined using the ranges suggested by Hartl and Clark (1997). We observed a large number of significant deviations from Hardy-Weinberg equilibrium prior to comparing population pairs (presented below), indicating that assumptions of allele neutrality (Slatkin 1985) may not be valid. We therefore calculated 2 sets of q values; the first was averaged over all loci and the second was calculated using only loci in H-W equilibrium (including monomorphic loci) in both locations at a site (see Table 3). Results of both were similar and so only the results from all 6 polymorphic loci are presented. Unbiased genetic distance (Nei 1978) was estimated using BIOSYS-1 (Swofford and Selander 1981). We report genetic distances for each paired study site from the matrix of all pairwise comparisons.

Results

Genetic diversity in fragmented and unfragmented locations The mean number of alleles per locus ranged from 2.8 to 4.7 among all sampling locations (Fig. 2a) and was not significantly different between fragmented (mean ± 1 SD: 3.97 ± 1.25) a nd unfragmented (4.22 ± 1.33;

ANCOVA; F1,135 = 1.03, p = 0.31) sample locations. Neither HWexp nor HWobs was significantly different between fragmented and unfragmented locations (Fig. 2b; ANCOVA; F1,135 = 0.19; p = 0.66 and F1,135 = 0.01, p = 0.76, respectively). Mean HWexp was significantly greater than HWobs at all locations (probability test; all p-values < 0.05; Fig. 2b). Examined individually, 39 of 138 loci exhibited heterozygote deficiencies as indicated by significant positive f values after Bonferroni correction (Table 3). No significant differences in percent of loci polymorphic and the Simpson’s index of genotype diversity (Student’s t-test, p = 0.65 and 0.90, respectively) occurred between fragmented and unfragmented locations. All sites had genotype diversity indices greater than 0.86 (10 of the 22 sites has values above 0.95) indicating that nearly every individual within each location had a unique 6-locus genotype.

A 6 Lake-fragmented

s Reservoir-fragmented u

c 5 Unfragmented o l

r 4 e p

s 3 e l e

l 2 l a

. 1 o N 0 Figure 2. Genetic B diversity (based on 6 0.8 polymorphic loci) of

y HW HW

t obs exp i Baetis alpinus in all s

o 0.6 sampling locations g

y measured as (A) mean z

o number of alleles per r 0.4 e

t locus (error bars e

H indicate 1 SD) and (B) 0.2 HWexp and HWobs (error W

H bars indicate 1 SE). 0.0

Genetic differentiation between fragments θ values averaged across all loci indicated moderate to substantial differentiation between the lake-fragmented locations Schwellisee, Bianco, Minor, and Cadagno (θ ranged from 0.092 to 0.362; Fig. 3a). θ values were well below 0.05 at the other 2 sites separated by lakes (the Upper and Lower Jörisee) indicating little or no genetic differentiation. θ values also indicated little or no differentiation between locations separated by the reservoirs Livigno and Marmorera, although the value at lake/reservoir Ritom (θ = 0.075; Fig. 3a) indicated significant differentiation similar to 4 of the lakes. θ ranged from 0.004 to 0.041 in unfragmented reference streams (Fig. 3a), implying little or no genetic differentiation. Values of Nei’s (1978) genetic distance ranged from 0.009 to 0.228 and patterns were consistent with those observed for θ (Fig. 3b). Specifically, the 5 sites exhibiting the largest genetic distance were the lakes Schwellisee, Minor, Cadagno, Bianco, and the lake/reservoir Ritom. The smallest genetic distance values also were observed for the lake- fragmented sites Upper and Lower Jöriseen (Fig 3b). 23

Table 3. f-values for each locus at each sampling location. An asterix (*) indicates significant difference from zero following Bonferoni correction. A dash (--) indicates loci that were monomorphic. Location Mpi PepB PepC PepD Pgi Pgm Schwellisee-1 .176 .431 -- .364* -.029 -.078 Schwellisee-2 .749* .287 -- .891* -.040 .085 Bianco-1 .699* .531* -- .799* -.051 .375* Bianco-2 .587* .474 -- .769* .092 .696* Minor-1 .004 .130 -- .793* -.068 .355 Minor-2 .673* .165 -- .715* -.015 .229 Cadagno-1 .999* .405 .999* .581 .607* .078* Cadagno-2 .454 .123 -- .652* -.155 .054 Upper Jörisee-1 .573 .381 -- .760* -.212 .091 Upper Jörisee-2 -.067 .224 -.011 .514 .020 .065 Lower Jörisee-2 .276 .454* -.021 .639* .284 .340 Livigno-1 .484 .096 -- .652 .372 .060 Livigno-2 .454* .057 -- .374 -.005 .087 Marmorera-1 .648 .331 -- .309 .082 .533 Marmorera-2 .625* .088 .999* .373 -.203 .380 Ritom-2 .641* .418* -.062 .253 .042 -.188 Arosa-1 .242 .718* -.010 .144 .010 .285 Arosa-2 .069 .484* -- .566* -.140 .392 Julierpass-1 .690* .318 -- .861* .053 -.098 Julierpass-2 .669* .458 -- .426 -.115 .210 Silvaplana-2 .513* .212 -- .796* -.042 .414 San Bernadino-1 .758* .599* -- .709* .375 .248 San Bernadino-2 .190 .420* -- .583* .075 .024

No clear pattern existed between the values and distance between fragments (Fig. 4; Pearson r = -0.25). The same was true for relationships between and site elevation (r = -0.25) and between and change in elevation between fragments (r = -0.43; neither relationship is shown).

Discussion

Genetic diversity in fragmented and unfragmented streams We observed no reduction in genetic diversity in populations of Baetis alpinus in fragmented relative to unfragmented habitats. Genetic diversity in fragmented populations can be reduced first by an initial decrease in population size upon fragmentation, and further by loss of alleles due to inbreeding in small populations. Gene flow between fragments may counteract both of these processes and maintain genetic diversity (Slatkin 1985). Based upon calculated values of , gene flow was reduced among 24

A 0.40 Lake Reservoir 0.35 Unfragmented Stream 0.15 q 0.10

0.05

0.00 B 0.25

0.20 Figure 3. Genetic differentiation between 0.15 paired locations at each study site estimated using 0.10

Distance (a) q averaged across all 0.05 loci, and (b) Nei’s (1978) unbiased genetic distance. Nei's (1978) Genetic 0.00 The dotted line (a) represents q = 0.05, above

Minor which indicates genetic Ritom Arosa Bianco Livigno

Cadagno differentiation (Hartl and Julierpass Silvaplana Marmorera Schwellisee Clark 1997). Upper Jörisee Lower Jörisee San Bernadino

some fragments (discussed below), suggesting that large effective population sizes may maintain genetic diversity. The observed levels of inbreeding provide contradictory evidence for large populations as discussed below. Overall genetic diversity measured with these 6 allozyme loci was high, with a total of 44 alleles and all HWexp values between 0.285 and 0.504. These values of HWexp are elevated because only polymorphic loci were used in the study, but overall diversity shows that allozymes provided sufficient resolution for detecting any genetic differentiation that occurred among populations.

Genetic differentiation at evolutionary and ecological time scales Based upon multi-locus estimates of q and genetic distance, we conclude that some alpine lakes act as significant barriers to gene flow for B. alpinus. Direct measures of dispersal of another species of Baetis indicated that a large number of adults flew 1.6 to 1.9 km upstream of the emergence site (Hershey et al. 1993). It is likely that a behavioral response of halted flight dispersal and immediate female oviposition by B. alpinus 25

0.4 Lake-fragmented 0.3 Reservoir-fragmented Unfragmented 0.2 q 0.1

0.0

0 2000 4000 6000 8000 10000 Distance between fragments (m) Figure 4. Scatterplot examining the relationship between q (averaged across all loci) and geographic distance between fragments.

occurs when the water surface changes from flowing (the stream) to standing (the lake) (Michel Sartori, Musée cantonal de Zoologie, Lausanne, Switzerland, personal communication). Such behavior also has been documented for other taxa of aquatic insects (Richardson and Mackay 1991). Our initial hypothesis addressed evolutionary time scales and predicted that differentiation would be observed between lake-separated populations but not between reservoir-separated populations and not between locations along unfragmented streams. We did observe marked differentiation across 4 lakes (Schwellisee, Bianco, Minor, and Cadagno) and little or no differentiation above and below reservoirs and in all unfragmented reference streams in accordance with our hypothesis. However, the lowest values of q observed in the study also were across lakes (Upper and Lower Jörisee). We observed differentiation (q = 0.075) across the reservoir Ritom, but the fact that Ritom was a natural lake enlarged by a dam further supports the conclusion that differentiation occurred only between sites fragmented by lakes. The lack of a relationship between geographic distance and q suggested that reduced gene flow over greater distances was not responsible for the observed patterns of q. Substantial differences occurred at Schwellisee where the distance between sample pairs was 350 m, but not 26 at Livigno where the distance was 10 km. We also observed no relationship between genetic separation and elevation or change in elevation. We suggest that low q values at the reservoirs Livigno and Ritom resulted from the fact that fragmentation is recent and thus population differentiation is not yet detectable. Sweeney et al. (1986) observed no genetic differentiation between populations of mayflies (Ephemerella subvaria and Eurylophella verisimilis) above and below reservoirs of the Delaware River, USA. We suggest, based on our results, that this may not be indicative of current gene flow patterns, but that population differentiation has not been manifested in allele frequencies of the genetic markers. However, the lack of genetic differentiation at the 2 Jörisee lakes requires consideration of alternative hypotheses that address shorter time scales (see below).

Genetic differentiation and glacial activity in the valley We suggest that mechanisms operating at intermediate time scales are responsible for the lack of differentiation observed at the 2 Jörisee lakes. The geologic history of all 13 study sites is relatively similar. All of our study sites occur in valleys that were ice-covered during the last glacial maximum when snowline depression was at approximately 1300 m. Glacial retreat occurred between 13,000 and 10,000 years ago and drainage networks, including lakes, began to form. In valleys where deglaciation was not complete, Holocene readvancement has occurred up to 3 times in the last 150 to 3000 years as recorded in the Aletsch and Gorner Glaciers in Switzerland (Holzhauser 1995). This is the case in the valley containing the Jöri lakes, where the existing Jöri Glacier extended into the lower Jörisee itself during the Little Ice Age (1600 - 1850 A.D.) (Maisch 1992). In contrast, the valleys containing Schwellisee and Minor have been ice- free since the last maximum (Maisch 1992, 1995) indicating that the drainage systems are on the order of 10,000 years old. The same is likely true for Lake Cadagno based on the lack of a glacier at present; one study documents its age at 8000 years (Del Don et al. 1998). Because Ritom is located in the same valley as Cadagno and at a lower elevation, it likely has been ice-free for a comparable amount of time. It is unclear whether Lake Bianco has been reglaciated, but the absence of a glacier at present and its south-facing aspect suggests it has been ice-free throughout the Holocene.

Heterozygosity and inbreeding in Baetis alpinus We observed pronounced heterozygote deficiencies in the present study. Only 3 of 23 populations were in Hardy-Weinberg equilibrium for all loci, even after Bonferroni correction. Similar findings have been reported for another species of Baetis (Schmidt et al. 1995) as well as other stream invertebrates (Bunn and Hughes 1997, Hughes et al. 1998). As an explanatory mechanism, Schmidt et al. (1995) and Bunn and Hughes (1997) proposed extinction and recolonization of local populations, specifically that populations were the result of only a few ovipositing females each generation. Even relatively conservative estimates of fecundity accounted for estimated population sizes of the streams investigated (Bunn and Hughes 1997). In the present study, we observed even more pronounced heterozygote deficiencies than did Schmidt et al. (1995) and Bunn and Hughes (1997), thus presenting further evidence that sampled individuals may be offspring of only a few adult matings. The genetic examination of 25 ovipositing females within a given reach of stream would perhaps offer some explanation for this consistent pattern.

Conclusions Our results suggest that historical rather than present-day levels of gene flow are reflected in the existing patterns of population genetic structure of Baetis alpinus (Larson et al. 1984). Without more information regarding the temporal resolution of genetic markers, we suggest caution should be taken when studies of habitat fragmentation conclude that levels of gene flow remain high. For example, if the present study had been limited to only the reservoirs, high levels of gene flow may have been invoked as the causal mechanism. We suggest, however, that low q values are probably not indicative of current levels of gene flow, but rather reflect a slow rate of divergence of fragmented populations, and that the allozyme markers used were able to detect genetic isolation of populations only after 1000+ years of reduced gene flow. Because reservoirs tend to be larger in surface area and elongate in shape (Ryding and Rast 1989), differentiation across reservoirs may occur at a faster rate and be even more pronounced than across lakes.

Acknowledgements The authors especially appreciate the assistance of Mäggi Hieber during field sampling. Field and laboratory assistance also was provided by Peter Burgherr, Christine Calvino, Christina Jolidon, Sandra Lass, 28

Florian Malard, Friederike Mösslacher, Marcos de la Puenta Nilsson, Karsten Rinke, Sven Schalla, and Bettina Wagner. Gigi Ostrow helped with field and laboratory work and provided critical comments on an early draft of the manuscript. We appreciate the help of Urs Uehlinger during study site selection and we thank Thomas Scheurer and Flurin Filli for their assistance and encouragment in the Swiss National Park. Much gratitude is owed Christian Ohlendorf and Gerhard Mohler for their help obtaining and interpreting literature on the Quaternary geology of the study sites. Finally, we thank John Brookfield and two anonymous r eferees for comments that improved the manuscript. Research was funded by Swiss National Science Foundation grant No. 31-50444.97/1.

Literature Cited Bossart, J. L., and D. P. Prowell. 1998. Genetic estimates of population structure and gene flow: limitations, lessons, and new directions. Trends in Ecology and Evolution 13:202-206. Breitenmoser-Würsten, C., and M. Sartori. 1995. Distribution, diversity, life cycle and growth of a mayfly community in a prealpine stream system (Insecta, Ephemeroptera). Hydrobiologia 308:85-101. Brown, J. H. 1971. Mammals on mountaintops: nonequilibrium insular biogeography. American Naturalist 105:467-478. Bunn, S. E., and J. M. Hughes. 1997. Dispersal and recruitment in streams: evidence from genetic studie s. Journal of the North American Benthological Society 16:338-346. Del Don, C., K. Hanselmann, R. Peduzzi, and R. Bachofen. 1998. Orographical and geochemical description of the meromictic Alpine Lake Cadagno. Documenta dell 'Istituto Italiano di Idrobiologia 63:5-9. Frankham, R. 1997. Do island populations have less genetic variation than mainland populations? Heredity 78:311-327. Hanski, I. 1998. Metapopulation dynamics. Nature 396:41-49. Hartl, D. L., and A. G. Clark. 1997. Principles of population genetics, 3rd edition. Sinauer, Sunderland, Massachusetts. Hastings, A., and S. Harrison. 1994. Metapopulation dynamics and genetics. Annual Review of Ecology and Systematics 25:167-188. Hebert, P. D. N., and M. J. Beaton. 1989. Methodologies for allozyme analysis using cellulose acetate electrophoresis. Helena Laboratories, Beaumont, Texas. Hershey, A. E., J. Pastor, B. J. Peterson, and G. W. Kling. 1993. Stable isotopes resolve the drift paradox for Baetis mayflies in an Arctic River. Ecology 74:2315-2325. Hic kerson, L. L., and P. G. Wolf. 1998. Population genetic structure of Arctomecon californica Torrey and Fremont (Papaveraceae) in fragmented and unfragmented habitat. Plant Species Biology 13:21-33. Holzhauser, H. 1995. Gletscherschwankungen innerhalb der letzten 3200 Jahre am beispiel des grossen Aletsch- and des Gornergletschers. Neue Ergebnisse. Pages 101-122 in Gletscher in ständigen Wandel: Jubiläums-Symposium der Schweizerischen Gletscherkommission 1993 Verbier (Valais, Schweiz). Hochschulverlag AG, ETH Zürich. Hughes, J. M., S. E. Bunn, D. A. Hurwood, and C. Cleary. 1998. Dispersal and recruitment of Tasiagma ciliata (Trichoptera: Tasimiidae) in rainforest streams, south-eastern Australia. Freshwater Biology 39:117-127. Humpesch, U. H. 1979. Life cycles and growth of Baetis spp. (Ephemeroptera: Baetidae) in the laboratory and in two stony streams in Austria. Freshwater Biology 9:467-479. 29

Knoll-Heinz, F. 1991. Piora: Konzept für die Erhaltung einer Landschaft. WWF Sezione Svizzera Italiano, Bellinzona. Kukula, K. 1997. The life cycles of three species of Ephemeroptera in two streams in Poland. Hydrobiologia 353:193-198. Lacy, R. C. 1987. Loss of genetic diversity from managed populations: interacting effects of drift, mutation, immigration, selection, and population subdivision. Conservation Biology 1:143-158. Larson, A., D. B. Wake, and K. P. Yanev. 1984. Measuring gene flow among populations having high levels of genetic fragmentation. Genetics 106:293-308. Lavandier, P. 1982. Evidence of upstream migration by female adults of Baetis alpinus Pict. (Ephemeroptera) at high altitude in the Pyrenees. Annales de Limnologie 18:55-59. Lavandier, P. 1988. Semivoltinisme dans des populations de haute montagne de Baetis alpinus Pictet (Ephemeroptera). Bulletin de la Société d'Histoire Naturelle de Toulouse 124:61-64. Maisch, M. 1992. Die Gletscher Graubündens; Habil Schrift Geograph. Inst. Univ. Zürich, Teil A und B. Physische Geographie 33:1-428. Maisch, M. 1995. Gletscherschwundphasen im Zeitraum des ausgehended Spätglazials (Egesen-Stadium) und seit dem Hochstand von 1850 sowie Prognosen zum künftigen Eisrückgang in den Alpen. Pages 81-100 in Gletscher in ständigen Wandel: Jubiläums- Symposium der Schweizerischen Gletscherkommission 1993 Verbier (Valais, Schweiz). Hochschulverlag AG, ETH, Zürich. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590. Ramirez, M. G., and K. E. Haakonsen. 1999. Gene flow among habitat patches on a fragmented la ndscape in the spider Argiope trifasciata (Araneae: Araneidae). Heredity 83:580-585. Raymond, M., and F. Rousset. 1995. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86:248-249. Richardson, B. J., P. R. Baverstock, and M. Adams. 1986. Allozyme electrophoresis: a handbook for animal systematics and population studies. Academic Press, San Diego. Richardson, J. S., and R. J. Mackay. 1991. Lake outlets and the distribution of filter-feeders: an assessment of hypotheses. Oikos 62:370-380. Ryding, S. O., and W. Rast. 1989. The control of eutrophication of lakes and reservoirs. UNESCO, Paris. Saccheri, I., M. Kuussaari, M. Kankare, P. Vikman, W. Fortelius, and I. Hanski. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494. Sartori, M., and P. Landolt. 1999. Atlas de distribution des éphémères de Suisse (Insecta, Ephemeroptera). Centre Suisse de Cartographie de la Faune, Neuchatel. Saunders, D. A., R. J. Hobbs, and C. R. Margules. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology 5:18-32. Schmidt, S. K., J. M. Hughes, and S. E. Bunn. 1995. Gene flow among conspecific populations of Baetis sp. (Ephemeroptera): Adult flight and larval drift. Journal of the North American Benthological Society 14:147-157. Slatkin, M. 1985. Gene flow in natural populations. Annual Review of Ecology and Systematics 16:393-430. Stangel, P. W., M. R. Lennartz, and M. H. Smith. 1992. Genetic variation and population structure of red-cockaded woodpeckers. Conservation Biology 6:283-292. Sweeney, B. W., D. H. Funk, and R. L. Vannote. 1986. Population genetic structure of two mayflies (Ephemerella subvaria, Eurylophella versimilis) in the Delaware River drainage basin (USA). Journal of the North American Benthological Society 5:253-262. Swofford, D. L., and R. B. Selander. 1981. Biosys-1: a FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72:281-283. Templeton, A. R., K. Shaw, E. Routman, and S. K. Davis. 1990. The genetic consequences of habitat fragmentation. Annals of the Missouri Botanical Garden 77:13-27. Thomas, A. G. B. 1975. Ephéméroptères du sud-ouest de la France. I. migrations d'imagos a haute altitude. Annales de Limnologie 11:47-66. 30 van Dongen, S., T. Backeljau, E. Matthysen, and A. A. Dhondt. 1998. Genetic population structure of the winter moth (Operophtera brumata L.) (Lepidoptera, Geometridae) in a fragmented landscape. Heredity 80:92-100. 31

CHAPTER 3

POPULATION GENETIC STRUCTURE OF 3 ALPINE

STREAM INSECTS: INFLUENCES OF GENE FLOW,

DEMOGRAPHICS, AND HABITAT FRAGMENTATION

with P. Spaak, C.T. Robinson, and J. V. Ward, (2002) Journal of the North American Benthological Society 21: 114-131.

Abstract Estimating scales of dispersal for benthic macroinvertebrates using neutral genetic markers requires consideration of genetic, demographic, and historical influences on population genetic structure. We used allozyme electrophoresis to investigate the population genetic structure of 3 species of alpine stream insects among major drainages of the Swiss Alps (Rhine, Inn, and Ticino rivers), among streams within each drainage, and within single streams. Within streams we examined reaches that were fragmented by lakes or resevoirs and unfragmented reaches. Rhithrogena loyolaea (Heptageniidae) exhibited little genetic differentiation (q) within (q = 0.01 - 0.03) and among (q = 0.02- 0.03) streams but significant differentiation among drainages (q = 0.08), suggesting that dispersal occurs among stream fragments and among stream valleys. Allogamus auricollis (Limnephilidae) did not exhibit genetic differentiation at any scale, suggesting that dispersal occurs throughout the geographical range of the study. In contrast, Baetis alpinus (Baetidae) showed moderate to substantial differentiation both within (q = 0.08 - 0.39) and among (q = 0.06 - 0.09) streams, but showed a distinct lack of genetic differentiation among major drainages of the Alps (q = 0.01). An explanation for this surprising result is not immediately apparent, but we suggest that low q values at large spatial scales reflect historical rather than present-day levels of gene flow and that genetic population structure of B. alpinus 32

reflects a lack of equilibrium between gene flow and genetic drift. Historical gene flow may continue to mask reduced dispersal and recurring processes of recruitment may lead to random changes in genetic signatures. We conclude that demographic processes affect small-scale patterns and historical processes affect large-scale patterns. The simultaneous study of multiple spatial scales helps determine the relative importance of each. A synthesis of our results and data for other species from published studies indicated that 4 consistent patterns of genetic differentiation emerged when multiple spatial scales were investigated. These patterns are indicative of taxon-specific dispersal ability within and among streams and whether taxa are in gene flow - genetic drift equilibrium.

Introduction Dispersal can be an important factor maintaining populations of species in fragmented habitats. Dispersal can counteract local extinction directly via immigration (Stacey and Taper 1992) and can ensure population persistence at larger spatial scales by maintaining a metapopulation structure (Hanski 1998). Dispersal among habitat fragments also may provide sufficient gene flow to maintain the genetic diversity within fragments, thereby indirectly reducing the probability of local extinction (Saccheri et al. 1998). Dispersal of organisms among habitat fragments often is studied using population genetics, where levels of gene flow are inferred from the spatial distribution of neutral alleles (Slatkin 1985). Such an approach can circumvent some of the practical difficulties involved with directly measuring the dispersal of organisms. Population genetic techniques also directly estimate genetic diversity, which can be reduced as a consequence of habitat fragmentation (e.g., Morden and Loeffler 1999). On the other hand, genetic structure can be influenced by historical and demographic processes. These processes can confound patterns interpreted as present-day levels of gene flow because they may result in nonequilibrium between genetic drift (random loss of alleles within a population) and gene flow (movement of alleles among populations). Historical and demographic processes likely affect population genetic signatures at large and small scales, respectively. For some species, Pleistocene climate changes have resulted in large-scale habitat changes 33 such that populations have become increasingly fragmented. Large-scale population genetic signatures may continue to reflect the older, more continuous habitat distribution, whereas small-scale population genetic structure may better reflect present-day levels of gene flow (Barber 1999). Different patterns at different spatial scales can result from the fact that equilibrium between genetic drift and gene flow is more rapidly reached at smaller than larger spatial scales (Hellberg 1994). On the other hand, rapid population turnover within and between generations can lead to temporal variation in genetic signatures at small spatial scales (e.g., Piertney and Carvalho 1995). These rapid, small-scale changes in genetic structure may confound large-scale estimates of gene flow because of the resulting nonequilibrium at small scales (Wade and McCauley 1988). Stream benthic macroinvertebrates face habitat fragmentation at a variety of spatial scales, and so understanding the spatial scale of dispersal and the processes that may affect population genetic signatures at different spatial scales are important for predicting potential consequences of habitat fragmentation. River systems often traverse several biomes, effectively isolating their headwaters by biogeographic barriers downstream (Minshall 1988). At smaller spatial scales, rapid changes in longitudinal habitat characteristics may isolate species locally (Ward 1994) and drainage divides may limit dispersal among streams. Within individual streams, lakes and reservoirs create discrete flowing reaches separated by unsuitable habitat for many stream macroinvertebrates. Evidence from genetic studies of stream benthic macroinvertebrates suggests that both historical and demographic processes may influence their population genetic signature. In an earlier study of the mayfly, Baetis alpinus Pictet (Baetidae), we found that populations were genetically differentiated among habitat fragments in alpine streams and concluded that dispersal over lakes was limited (Monaghan et al. 2001; see chapter 2). Genetic differentiation was unrelated to lake size, but occurred only if lakes were situated in valleys that were ice-free throughout the Holocene. We concluded that the low levels of genetic differentiation observed between fragments separated by reservoirs (100 y old) and more recently formed lakes (100s - 1000s y old) did not indicate high levels of gene flow but rather indicated that nonequilibrium between genetic drift and gene flow has prevented genetic differentiation since fragmentation. With regard to demographic processes that may affect population genetic signatures, many studies of stream macroinvertebrates have observed pronounced levels of reduced heterozygosity (inbreeding) and attribute this 34 finding to oviposition by a few females (Schmidt et al. 1995, Bunn and Hughes 1997). Such recurring demographic processes may confound large-scale genetic structure because of rapid fluctuations in the spatial distribution of alleles at small scales. The aim of our study was to investigate how habitat fragmentation at multiple spatial scales affected 3 species of stream insects. We examined larval population genetic structure of B. alpinus, Rhithrogena loyolaea Navàs (Heptageniidae), and Allogamus auricollis Pictet (Limnephilidae) using allozyme electrophoresis. We estimated levels of gene flow at multiple spatial scales: among major drainages of the Swiss Alps, among streams, and within streams. Our 1st objective was to examine whether habitat fragmentation within streams had similar effects on genetic diversity and gene flow of R. loyolaea and A. auricollis as was previously observed for B. alpinus (Monaghan et al. 2001; see chapter 2). Our 2nd objective was to determine explicit scales of dispersal by examining whether or not genetic patterns were consistent across multiple spatial scales. We hypothesized that a lack of consistency among scales indicates that different processes affect patterns at different scales.

Methods

Study sites The study was conducted in 3 major drainages in the Swiss Alps, constituting the headwaters of the Rhine, Inn, and Ticino rivers (Fig. 1). Within each major drainage we sampled either 3 or 4 streams and within each stream we sampled either 2 or 3 sites (Table 1), resulting in a total of 25 sampling sites. At the within-stream scale, we sampled 2 types of streams in each major drainage: those that contained potential dispersal barriers (a lake or reservoir) and those that did not (unfragmented). Upstream sampling sites were numbered 1 and downstream sites 2 (Fig. 1) for purposes of data presentation. Sites located between lakes were called site 2 of the upper lake. For example, Upper Jöri-2 was the lower site of Upper Jöri lake and was the upper site of Lower Jöri lake (Fig. 1). Upstream sites ranged in elevation from 1100 m to 2525 m, with 19 of 25 locations occurring above 1900 m (Table 1). Sites within streams were 280 m to 10 km apart. The change in elevation within streams ranged 35

Switzerland

Arosa 1 Upper Julierpass Rhine R. 4 2 Lower Julierpass 1 Muesa 6 Ticino River Inn R. 9 3 2 T 11 8 icino R. 7 5 1 10 1 Schwellisee 2 Minor Bianco Upper/Lower Jöri N Marmorera Cadagno/Ritom Livigno 50 km 2

Figure 1. The 11 study streams in the Swiss Alps (top) and schematic representations of sampling sites within each stream (bottom). Streams on the map are designated as: 1- Schwellisee, 2-Upper/Lower Jöri, 3-Marmorera, 4-Arosa, 5-Minor, 6-Livigno, 7- Upper/Lower Julierpass, 8-Bianco, 9-Cadagno/Ritom, 10-Muesa, 11-Ticino River. Ovals in the schematics represent lake or reservoir habitat and numbers indicate sampling sites along the stream. Distance between sampling sites in a stream ranged from 280 m to 10 km (see Table 1). from 4 to 250 m (Table 1). There were no tributaries entering any of the study streams between sample sites.

Study Animals Baetis alpinus is a widespread and abundant alpine species (Humpesch 1979, Breitenmoser-Würsten and Sartori 1995). Larvae are eurythermal and occur between 200 and 2600 m in elevation (Sartori and Landolt 1999). The B. alpinus life cycle is plastic, ranging from bivoltine to semivoltine depending upon elevation (Humpesch 1979, Lavandier 1988). Adults exhibit pronounced upstream flight bias (Thomas 1975, Lavandier 1982) and our previous finding indicated dispersal was limited within fragmented streams. Rhithrogena loyolaea also is widespread but occurs within a more limited elevation range of 1300 to 2600 m (Sartori and Landolt 1999). Larvae are cold stenotherms (Vincon and Thomas 1987) with semivoltine (2 - 3 y) development (Lavandier 1981, Olechowska 1981). Its flight behavior is not as well studied as B. alpinus, although Thomas (1975) observed upstream bias and considerable altitude gains over B. alpinus, suggesting it may be a stronger flier. Lavandier (1981) also documented upstream flight bias but gave no estimates of distance. Allogamus auricollis is locally very abundant and is univoltine (Waringer 1986, Graf et al. 1993). Limnephilids are considered strong 36

Table 1. Geographical characteristics of the study streams and sampling sites in the 3 major drainages (see Fig. 1). Lake and reservoir names are from maps. Unfragmented reference streams, with the exceptions of Muesa and the Ticino River, were unnamed and therefore designated by location. Elevation data are for the upper location at each site. asl = above sea level. Major drainage Dispersal barrier Distance between Elevation Elevation and stream fragments (m) (m asl) change (m) Rhine Schwellisee Lake 350 1935 5 Upper/Lower Jöri Lakes (2) 550/975 2525/2495 30/175 Marmorera Reservoir 7750 1700 250 Arosa None 375 1940 10 Inn Minor Lake 375 2340 15 Livigno Reservoir 10,000 1910 155 Upper/ Lower None 625/3225 2310 105/220 Julierpassa Ticino Bianco Lake 525 2080 4 Cadagno/Ritomb Lakes (2) 1075/4500 1940/1090 40/120 Muesa None 280 2225 25 Ticino Riverc None 8000 1100 110 a Julierpass was a single stream with 3 sampling sites. For Upper Julierpass, the uppermost site was compared with a site ~0.6 km downstream. For Lower Julierpass, the uppermost site (= upstream site of Upper Julierpass) was compared with a site ~3.2 km downstream. b A single stream fragmented by a lake (Cadagno) and a reservoir that is an enlarged lake (Ritom). c The Ticino River was sampled only for Allogamus auricollis (see text). fliers (Svensson 1974) but we know of no studies that investigated the flying ability of A. auricollis.

Sample collection and allozyme electrophoresis Late-instar larvae were collected using a 250-mm mesh kicknet, kept alive for 1 to 2 h in stream water, flash-frozen in liquid N, and stored for between 4 and 8 mo prior to allozyme electrophoresis in the laboratory. Collection of animals occurred on a single day at each site in summer 1999. Initially, all sites except the Ticino River were sampled for B. alpinus and R. loyolaea. Rhithrogena loyolaea was not observed at all sites, including both streams that were fragmented by a single reservoir (Livigno, Marmorera). Thus, A. auricollis was collected at these sites for the purpose of comparing it with B. alpinus. The Ticino River was then sampled only to provide an unfragmented reference stream for A. auricollis. In the laboratory, larvae were thawed, identified, and ground in ~80 mL of crushing buffer (diH2O, NADP, b-mercaptoethanol). Cellulose acetate electrophoresis (Hebert and Beaton 1989) was used to screen 25 enzyme systems for each species using individuals from a subset of 37

Table 2. Locus name, enzyme system (including peptidase substrate), International Enzyme Commission (E.C.) number, running buffer, and number of alleles scored for each locus. Buffer systems are those indicated by Richardson et al. (1986). Blanks indicate that the respective locus was not resolved successfully for the species. No. of alleles Locus Enzyme E.C. Buffer Rhithrogena Allogamus Baetis Number loyolaea auricollis alpinus Gda Guanine deaminase 3.5.4.3 I 7 Mpi Mannose-phosphate 5.3.1.8 A 4 4 9 isomerase Pep-a Peptidase (val-leu) 3.4.11 or 13 I 4 Pep-B Peptidase (leu-gly-gly) 3.4.11 or 13 I 7 Pep-C-1 Peptidase (leu-ala) 3.4.11 or 13 I 4 5 Pep-C-2 Peptidase (leu-ala) 3.4.11 or 13 I 4 Pep-D Peptidase (phe-pro) 3.4.13.9 I 5 7 Pgi Phosphoglucose isomerase 5.3.1.9 I 8 5 8 Pgm Phosphoglucomutase 2.7.5.1 I 4 4 8

sampling sites to identify polymorphic loci. Five and 6 polymorphic loci were identified for R. loyolaea and A. auricollis, respectively, and Monaghan et al. (2001; see chapter 2) reported on 6 polymorphic loci for B. alpinus (Table 2). Data analysis was based on at least 25 animals from each sampling location when possible.

Data analysis Mean number of alleles per locus (A) and expected Hardy-Weinberg heterozygosity (HWexp) were calculated for each locus at each sampling location using BIOSYS-1 (Swofford and Selander 1981). A and

HWexpwere compared between fragmented and unfragmented sites using ANOVA blocked by locus. A and HWexp for genetically differentiated and undifferentiated streams were compared in the same way when within- stream genetic differentiation was moderate or greater (q > 0.05, see below). Deviations from HW equilibrium were examined by calculating the inbreeding coefficient (f) and testing for significance using GENEPOP version 3.1d (M. Raymond and F. Rousset, Université de Montpellier II, Montpellier, France). Significant (Bonferroni-corrected) positive values of f indicate heterozygote deficiency and significant negative values indicate heterozygote excess. Linkage disequilibrium also was assessed using GENEPOP. Genetic differentiation of populations was determined by estimating q, a measure of the relative fixation of alternate alleles in different subpopulations (Weir and Cockerham 1984). Values of q, 95 % confidence intervals, and significant difference from 0 were examined 38 using FSTAT version 2.9.1 (J. Goudet, Université de Lausanne, Switzerland). We used the option that compares genotype frequencies rather than allele frequencies because of significant deviation from HW equilibrium (see Results). When q was significant, the degree of genetic differentiation was assessed using the ranges specified by Hartl and Clark (1997), where q < 0.05 indicates little differentiation, 0.05 - 0.15 indicates moderate differentiation, 0.15 - 0.25 indicates great differentiation, and > 0.25 indicates very great differentiation. Levels of genetic differentiation were assessed at 3 levels of a spatial hierarchy: 1) among the 3 major drainages (Rhine, Inn, and Ticino headwater populations), 2) among streams within each major drainage, and 3) within streams. At the within- stream scale, 7 streams contained potential dispersal barriers and 4 streams did not. When populations were genetically differentiated within streams (e.g., B. alpinus data from Monaghan et al. 2001; see chapter 2), we calculated q among streams and among drainages once for all populations and once using only those populations from streams that were not genetically differentiated. Thus, we could examine whether small-scale differentiation affected large-scale patterns.

Results Genetic diversity in fragmented and unfragmented streams Allele frequencies and locus n-sizes are reported in Appendices 1-3. A was not significantly different between fragmented and unfragmented sample locations for either R. loyolaea (F1,52 = 0.18, p = 0.68) or A. auricollis (F1,33 = 0.20, p = 0.66), as was the case for B. alpinus (F1,135 = 1.03, p = 0.31) reported previously (Fig. 2A). The same was true for mean

HWexp (Fig. 2B; statistics not reported). For B. alpinus, neither A nor HWexp was significantly different between sites in genetically differentiated and undifferentiated streams (F1,136 = 2.28, p = 0.13). The inbreeding coefficient, f, was significantly > 0 for R. loyolaea in 4 of 60 instances (~ 7 %; Table 3). For A. auricollis, a significant f was observed in 3 of 36 instances (~ 8 %; Table 4). These values were in contrast to the large number of significant deviations for B. alpinus (28 %; Monaghan et al. 2001; see chapter 2). Of the 375, 110, and 90 pairwise comparisons used to test for linkage disequilibrium of B. alpinus, R. loyolaea, and A. auricollis, respectively, 18, 4, and 4 were significant (p < 0.05); such results are expected through chance alone. 39

A 6 36 Fragmented sites 102 Unfragmented sites 5 35 20 24 12 4

A 3 2 1 0

B 0.8 Figure 2. Genetic diversity measured as (A) mean number 0.6 of alleles per locus and (B)

p mean expected H-W x

e heterozygosities for each 0.4 W species. Error bars indicate +1 H SD. Sample sizes (no. 0.2 populations x no. loci) for both variables are indicated 0.0 above the error bars in panel Rhithrogena Allogamus Baetis A. loyolaea auricollis alpinus

Table 3. Inbreeding coefficient (f) for each locus in each population of Rhithrogena loyolaea. * = significant difference from 0 following Bonferroni correction for the number of tests. -- = locus was monomorphic. Locus names given in Table 2. Location Gda Mpi Pep-D Pgi Pgm Rhine Schwellisee-1 0.538 0.473 0.580* 0.366 0.133 Schwellisee-2 -0.062 -0.041 0.062 0.017 -0.154 Upper Jöri-1 -0.047 -- 0.375 0.205 0.019 Upper Jöri-2 0.280 0.653 0.183 -0.094 -0.093 Lower Jöri-2 -0.036 -0.036 0.745* 0.079 0.105 Arosa-1 -0.079 0.500 -0.141 -0.048 0.183 Arosa-2 -0.048 0.000 0.000 0.028 0.102 Inn Upper Julierpass-1 -0.015 -0.024 0.713 0.252 0.469 Upper Julierpass-2 0.319 -0.076 0.650* 0.081 0.317 Minor-2 0.118 0.063 0.056 0.005 0.079 Ticino Bianco-1 -0.096 0.000 0.618 0.259 0.073 Bianco-2 -0.051 0.999* -0.045 0.077 -0.179

Table 4. Inbreeding coefficient (f) for each locus in each population of Allogamus auricollis. * = significant difference from 0 following Bonferroni correction for the number of tests. Locus names given in Table 2. Location Mpi Pep-a Pep-C-1 Pep-C-2 Pgi Pgm Livigno-1 -0.255 0.228 0.174 -0.036 -0.020 -0.010 Livigno-2 0.223 -0.011 0.000 -0.171 -0.057 -0.067 Marmorera-1 1.000* 0.000 1.000* -0.209 -0.204 -0.048 Marmorera-2 1.000* 0.000 -0.032 0.050 0.000 -0.076 Ticino River-1 0.084 0.481 -0.067 -0.005 -0.057 -0.166 Ticino River-2 -0.096 0.000 0.000 0.098 -0.020 -0.204

q at multiple spatial scales Values of q for R. loyolaea were significant within and among drainages (Table 5). q among major drainages of the Alps was much more pronounced (q = 0.080) than among streams within any drainage (q = 0.026 - 0.032; Table 5). Within streams, q was significant across Upper Jöri Lake but indicated little differentiation (q = 0.030; Fig. 3). We observed no genetic differentiation of R. loyolaea within any of the other streams, including those where relatively large differentiation occurred for B. alpinus (Fig. 3). q for A. auricollis was significant among major drainages but indicated little differentiation (Table 5). Across the reservoir Livigno, q for A. auricollis was significant but also low (q = 0.023); q was not significant across Marmorera or along the unfragmented Ticino River (Fig. 3). The lack of even moderate differentiation at the reservoir sites was similar to that observed for B. alpinus (Fig. 3). Values of q for B. alpinus were significant within and among drainages when all streams were included in the analysis and when only genetically undifferentiated streams were included (Table 5). Considering all streams together, differentiation within drainages (q = 0.064 - 0.089) was much higher than among drainages (q = 0.010). Considering only undifferentiated streams (i.e., with values of q < 0.05, Fig. 3), among- stream q values were lower than those computed using all streams together. Nonetheless, q remained at levels indicative of moderate differentiation. The surprising pattern of greater differentiation within drainages than among drainages was opposite to the pattern observed for R. loyolaea.

0.40 *** 0.35 Rhithrogena loyolaea Allogamus auricollis 0.15 ** Baetis alpinus

θ *** 0.10 ** ** 0.05 ** *** 0.00 Minor Arosa Ritom Muesa Bianco Livigno Cadagno Julierpass Lower Jöri Upper Jöri Silvaplana Marmorera Schwellisee Ticino River Figure 3. Genetic differentiation (q) within fragmented (lake or reservoir) and unfragmented streams for each species. Asterisks indicate significant differences from 0 (* = p < 0.05, ** = p < 0.01, *** = p < 0.001). The dotted line represents levels above which moderate differentiation occurs (Hartl and Clark 1997).

Table 5. Estimates of genetic differentiation (q) jackknifed across loci within and among each of the 3 major drainages (Rhine, Inn, and Ticino rivers). For Baetis alpinus, q-values are presented for all streams and separately for only the undifferentiated streams. Only 1 stream (Muesa) in the Ticino drainage was undifferentiated, so an among-stream q could not be calculated. Rhithrogena loyolaea was not genetically differentiated in any streams and so only a single analysis was performed. Allogamus auricollis was sampled from a single stream in each major drainage so only an among-stream analysis was performed. * = p < 0.05, ** = p < 0.01, *** = p < 0.001. -- = no analysis was performed. Undifferentiated All streams 95% CI Streams 95% CI Baetis alpinus Rhine 0.089*** 0.022 0.052*** 0.022 Inn 0.069*** 0.044 0.049*** 0.024 Ticino 0.064*** 0.040 -- -- Among 0.010*** 0.005 0.015*** 0.006

Rhithrogena loyolaea Rhine 0.032*** 0.021 -- -- Inn 0.026** 0.005 -- -- Ticino 0.026** 0.010 -- -- Among 0.080*** 0.005 -- --

Allogamus auricollis Among 0.042* 0.011 -- --

Multiscale patterns of q are presented graphically for simultaneous comparison of species and spatial scales (Fig. 4). Rhithrogena loyolaea genetic population structure was most pronounced at the largest spatial scale, indicating populations were structured primarily among drainages. Allogamus auricollis exhibited very little structure at the scales of the present study, with only a slight increase in q moving up 2 steps in the spatial hierarchy (Fig. 4). Considering B. alpinus populations undifferentiated within streams (Fig. 4), genetic population structure appeared most pronounced among streams, with lower values of q at both smaller and larger spatial scales. Considering populations with high levels of within-stream differentiation, genetic structure of B. alpinus populations was most pronounced within streams, with subsequent reduction in q moving to larger (among streams) and larger (among drainages) scales (Fig. 4).

0.40 0.35 Rhithrogena loyolaea 0.15 Allogamus auricollis Baetis alpinus - differentiated streams 0.10 Baetis alpinus - undifferentiated streams θ

0.05

0.00 Within Among Among streams streams drainages

Figure 4. Multilocus genetic differentiation (q) of each species at 3 hierarchical spatial scales. Within-stream values are the same as presented in Fig. 3 and are offset along the x- axis for clarity of presentation.

Discussion Genetic diversity in fragmented streams Results for all 3 species suggest that genetic diversity was not reduced by the fragmentation of lotic habitat by lentic water bodies. We observed no difference between fragmented and unfragmented populations using 2 informative estimates of genetic diversity: A and HWexp . In addition, we observed no difference in genetic diversity when populations of B. alpinus from genetically differentiated streams (q > 0.05) were compared with undifferentiated streams. These results contrast with expectations from empirical and theoretical work in fragmented populations, which have often found that genetic diversity is reduced (Lacy 1987, van Dongen et al. 1998, Morden and Loeffler 1999). The maintenance of genetic diversity implies population sizes may be large enough in fragments so that the loss of alleles by genetic drift is minimal, or gene flow among subpopulations sustains genetic diversity (Slatkin 1985).

Gene flow estimates at multiple spatial scales Multilocus estimates of q were distinctly different among species and varied within species depending on the spatial scale considered. We observed little or no genetic differentiation in R. loyolaea except among the major drainages of the study, suggesting that at least a moderate level of dispersal occurs within and among streams. The multiscale pattern of q (Fig. 4) presumably typifies species with a relatively large-scale population structure, an equilibrium between genetic drift and gene flow, and a decreased relatedness with increasing geographic distance (Slatkin 1993). One conclusion is that failure to observe even moderate genetic differentiation across any lakes, including those where B. alpinus was differentiated, results from gene flow among habitat fragments. Thomas (1975) observed R. loyolaea flying upstream and gaining twice as much elevation as B. alpinus, suggesting they are stronger flyers and capable of traveling considerable distances as adults. Flecker and Allan (1988) observed a congener, R. hageni, to fly randomly, including away from the stream, suggesting Rhithrogena may be capable of crossing areas of unsuitable habitat and may not be confined to following the drainage pattern. The low level of differentiation in B. alpinus among the 3 major drainages was surprising, based on our previous conclusion that gene flow was limited over lentic water bodies ~ 300 to 1000 m across (Monaghan et 44

al. 2001; see chapter 2). The homogeneity among major drainages also seems contradictory to among-stream q values in the present study. These values suggest limited dispersal of B. alpinus among different valleys. Small-scale differentiation and large-scale homogeneity is evidence that a species has not yet reached equilibrium between genetic drift and gene flow (Hellberg 1994). We suggest 2 possible mechanisms responsible for the genetic population structure of B. alpinus. Either, or both, could account for the lack of equilibrium between gene flow and genetic drift (as indicated by different estimates of q at different spatial scales), and for the lack of HW equilibrium in populations. One mechanism concerns heterozygote deficiency as evidence for recurring changes in small-scale population structure in the midst of large-scale equilibrium. The other mechanism concerns consistent small-scale patterns of q as evidence for limited present-day dispersal in the midst of large-scale patterns that continue to reflect historical patterns of gene flow. Recurring small-scale changes in genetic sructure - We observed a large number of heterozygote deficiencies for B. alpinus and frequency appeared unrelated to habitat fragmentation, longitudinal position in the stream, or geographical location in the study. Heterozygote deficiency may result from nonrandom mating, the presence of null alleles, mis- scoring of gels, or the presence of multiple species. We observed no null homozygotes in an analysis of > 1000 individuals and, although mis- scoring of gels can never be ruled out, it was an unlikely source of error because of consistent HW equilibrium observed at the same loci for the 2 other species. It is unlikely that multiple species of Baetis were analyzed at any sampling site because of good larval taxonomic descriptions (see Sartori and Landolt 1999), the lack of linkage disequilibrium, and a preliminary examination of DNA fragment length polymorphism for the same individuals (MTM, unpublished data). Heterozygote deficiency has been observed in several studies of aquatic insects, often at levels similar to what we observed for B. alpinus (28 % of possible instances). Schmidt et al. (1995) reported 25 % for Baetis sp.; Wishart and Hughes (2001) reported 30 % for Elporia barnardi (Blephariceridae), and Hughes et al. (1998) reported 23% in Tasiagma ciliata (Trichoptera). Schmidt et al. (1995) proposed and Bunn and Hughes (1997) extended an explanatory mechanism, suggesting that reduced direct-count heterozygosity results from larval populations at any given location being the result of only a few ovipositing females. If B. alpinus populations are the result of a small number of matings, then allele 45 frequencies at any given sample site could fluctuate randomly from one generation to the next. If genetic homogeneity at large scales results from contemporary wide-ranging dispersal ability, then ovipositing females constitute a small but random sample of females drawn from a very large gene pool, allowing genetic differentiation at local scales to arise by chance. Such a mechanism suggests equilibrium has been reached at the scale of the Alps but that random sampling of alleles (bottlenecking) occurring each generation results in changes in allele frequencies too rapid for equilibrium to be reached within streams. One limitation of this mechanism is that it seems unable to account for the consistent pattern of differentiation among streams for B. alpinus, regardless of whether we considered all streams or only undifferentiated streams. In addition, we observed a lack of HW equilibrium at all but 2 sampling sites for B. alpinus, but genetic differentiation was consistently observed only in fragmented streams and only in those streams where fragmentation was comparatively old (Monaghan et al. 2001; see chapter 2). Present-day and historical gene flow - A second possible mechanism is that small-scale population differentiation reflects present-day levels of gene flow and large-scale homogeneity reflects historical processes and a slow rate of approach to equilibrium between genetic drift and gene flow for B. alpinus. Major glacial advances (occurring twice in the last 200,000 y) forced populations downward in river drainages and likely mixed headwater populations below major confluences. During and after glacial retreat, populations dispersed into headwaters and slowly began to diverge genetically because of the tight coupling of downstream drift and upstream flight (Lavandier 1982). Populations have genetically diverged within streams in drainages fragmented by lakes formed during and soon after glacial retreat (Monaghan et al. 2001; see chapter 2). The overall distribution of alleles among the headwaters of the Alps remains similar to its historical configuration, however, because of the slow rate of approach to equilibrium between gene flow and genetic drift at this largest scale. Although neither mechanism can be explicitly ruled out with our data set, we can propose testable hypotheses based on each. Recurring bottlenecks should mean that allele frequencies at any given site change randomly from one generation to the next in a manner similar to a metapopulation (Piertney and Carvalho 1995). If the lack of HW equilibrium results from populations being founded by only a few ovipositing females, then these populations should contain relatively few 46 mtDNA haplotypes. Such a test would require comparison with another species whose populations are in HW equilibrium, presumably because populations are founded by many more females. On the other hand, equilibrium between gene flow and genetic drift should be achieved more rapidly with faster evolving molecular markers (e.g., mtDNA; Brown et al. 1982), thus allowing one to distinguish between historical and present-day gene flow.

Allogamus auricollis and lack of genetic differentiation over reservoirs As with B. alpinus, we observed little or no genetic differentiation of A. auricollis over the reservoirs Livigno and Marmorera. There also was little or no subpopulation structure even among major drainages. Limnephilidae typically are strong flyers (Svensson 1974) and dispersal among major drainages suggests that A. auricollis is able to cross reservoirs. However, the conclusion that gene flow continues over reservoirs can only be tentative because of uncertainty as to how rapidly genetic markers can detect recent fragmentation. Sweeney et al. (1986) observed no genetic differentiation between populations of mayflies (Ephemerella subvaria and Eurylophella verisimilis) above and below reservoirs of the Delaware River, USA, and Stiven and Kreiser (1994) observed no differentiation of stream-dwelling gastropod (Goniobasis proxima) populations separated by a reservoir. Thus, to our knowledge, researchers have never observed genetic isolation of benthic invertebrates, using allozymes, across reservoirs up to 10 km long.

Dispersal modes and genetic population structure of stream insects Using data from the present study and from other published studies of different species of stream insects, we present a synthesis of observed relationships between genetic population structure and dispersal modes. We examined studies that investigated at least 2 spatial scales and we consider those taxa that have a wide enough geographical distribution such that q among drainages can be calculated. Figure 5 depicts 4 different relationships between q and spatial scale for stream macroinvertebrates. Note that the x-axis depicts increasing spatial scales used in our study rather than linear distance as would occur in a strict isolation-by-distance (IBD) model (Slatkin 1993). In general, increasing q from left to right in 47

Stream equilibrium

Reach nonequilibrium

θ Drainage equilibrium

Stream nonequilibrium

Within Among Among streams streams drainages

Figure 5. Relationships between genetic differentiation (θ) and spatial scale for stream benthic macroinvertebrates as a function of the scales of dispersal and the mechanisms responsible for the patterns.

Fig. 5 should be indicative of species in equilibrium between genetic drift and gene flow. The rate of increase in θ depends on whether or not species disperse readily among streams and drainages. A decrease from left to right or a hump-shaped distribution should indicate species that are not in equilibrium. This pattern may be the result of local dynamics that recurringly alter allele frequencies, or may be the result of historical levels of gene flow confounding present-day population genetic signatures. The names of the 4 different curves are defined by the spatial scale at which dispersal becomes limited and by the presence or absence of gene flow - genetic drift equilibrium. Progressively increasing subpopulation differentiation (drainage equilibrium, Fig. 5) is indicative of a relatively widespread species with gene flow occurring within and among streams (i.e. gene flow becomes limited only among drainages) and large-scale equilibrium between genetic drift and gene flow (Slatkin 1993). Such species are likely to exhibit an IBD pattern, and include R. loyolaea in our study and the caddisfly Helicopsyche borealis studied by Jackson and Resh (1992). We note, however, that strong relationships of IBD could result even in the absence of gene flow at smaller spatial scales if recent population differentiation 48

(for example by habitat fragmentation) is not yet manifest in population genetic signatures (e.g., Barber 1999). Stream equilibrium (Fig. 5) includes species that experience gene flow primarily within streams or among reaches of streams, thereby having a more rapid increase in q as one moves to larger spatial scales. A clear example of such a species is the waterstrider Aquarius remigis (Hemiptera: Gerridae) investigated by Preziosi and Fairbairn (1992). They observed orders of magnitude increase in q from within to among streams. The atyid shrimp Paratya australiensis (Decapoda: Atyidae) also exhibited an order of magnitude increase in q from within to among streams (Hughes et al. 1995, Bunn and Hughes 1997). Paratya australiensis does not fly and A. remigis dispersal by flight is very rare (Preziosi and Fairbairn 1992). We suggest that most taxa with limited dispersal abilities would fall into this category. Taxa with very high instream dispersal ability but limited dispersal among streams should display a similar pattern, but with the curve shifted downward. Several taxa exhibit patterns of reduced differentiation at progressively larger spatial scales (reach nonequilibrium, Fig. 5). This pattern indicates large-scale structure is relatively homogeneous but that small-scale (within-stream) structuring exists, thus implying that small- scale structuring forces are more evolutionarily recent events (Hellberg 1994). The populations of B. alpinus that were genetically differentiated within streams in our study exhibited such a pattern, as did the caddisfly T. ciliata (Hughes et al. 1998). We note that B. alpinus genetic structure also fit a second curve when undifferentiated populations were considered (see below). Species that undergo the process of repeated bottlenecks (e.g., oviposition by only a few females) should match this distribution because such a process constitutes small-scale substructuring of each generation. Last, some taxa may structure genetically at intermediate spatial scales, with highest q found among streams. Such a curve (stream nonequilibrium) was the case for B. alpinus in unfragmented streams in the present study. This pattern may be predominant in species that are widespread biogeographically and have colonized areas in evolutionarily recent times from large source populations (e.g., mountain ranges following Pleistocene glaciation), but that maintain relatively low levels of gene flow among streams. Of note, the stonefly Yoroperla brevis (Plecoptera: Peltoperlidae) examined by Hughes et al. (1999) may exhibit a pattern of stream equilibrium (shifted downward as described earlier) or 49 stream nonequilibrium depending on its genetic structure at the largest spatial scale. In conclusion, we observed no reduction in genetic diversity in fragmented streams for any of the 3 species. Populations in fragments may remain large enough that no loss of alleles occurs via genetic drift, or levels of gene flow among fragments may remain high enough to counteract the loss of alleles. For R. loyolaea, the consistent pattern of increasing genetic differentiation with increasing spatial scale suggests populations were in equilibrium between genetic drift and gene flow. We conclude R. loyolaea disperses readily both within and among streams, but less so among major drainages. Allogamus auricollis did not exhibit genetic differentiation at any scale, suggesting that dispersal occurs throughout the geographical range of the study. In contrast, homogeneity at large spatial scales and differentiation at small spatial scales suggest a lack of equilibrium for B. alpinus. Consistent differentiation of B. alpinus between older stream fragments (~ 10,000 y) indicates dispersal is limited among fragments and that large-scale structure reflects historical levels of gene flow. Pronounced heterozygote deficiencies suggest structure at small spatial scales reflects genetic bottlenecks during recruitment. Last, limited dispersal among fragments and demographic processes likely affect small-scale patterns, and historical processes likely affect large-scale patterns. The simultaneous study of multiple spatial scales can help us to determine the relative importance of each process.

Acknowledgements We gratefully acknowledge the assistance of Mäggi Hieber during field sampling and we thank Andreas Frutiger for collecting animals from the Ticino River. Field and laboratory assistance also was provided by Peter Burgherr, Christine Calvino, Christine Dambone-Boesch, Christina Jolidon, Sandra Lass, Florian Malard, Friederike Mösslacher, Marcos de la Puenta Nilsson, Karsten Rinke, Sven Schalla, and Bettina Wagner. Gigi Ostrow assisted with field and laboratory work and commented on an earlier draft of the manuscript. MTM thanks Michel Sartori for inspiration and valuable insights regarding mayfly behavior, and Mike Dobson for conversations and an unpublished manuscript regarding historical patterns of gene flow in aquatic macroinvertebrates. Urs Uehlinger assisted with study site selection and we thank Thomas Scheurer and Flurin Filli for their continued encouragment of sampling in the Swiss National Park. The 50 manuscript benefited from the critical comments of Jane Hughes, Chris Caudill, Dave Strayer, David Rosenberg, and an anonymous reviewer. Research was funded by grant No. 31-50444.97/1 from the Swiss National Science Foundation.

Literature Cited Barber, P. H. 1999. Patterns of gene flow and population genetic structure in the canyon treefrog, Hyla arenicolor (Cope). Molecular Ecology 8:563-576. Breitenmoser-Würsten, C., and M. Sartori. 1995. Distribution, diversity, life cycle and growth of a mayfly community in a prealpine stream system (Insect, Ephemeroptera). Hydrobiologia 308:85-101. Breitenmoser-Würsten, C., and M. Sartori. 1995. Distribution, diversity, life cycle and growth of a mayfly community in a prealpine stream system (Insecta, Ephemeroptera). Hydrobiologia 308:85-101. Brown, V. M., E. M. Prager, A. Wang, and A. C. Wilson. 1982. Mitochondrial DNA sequences of primates: tempo and mode of evolution. Journal of Molecular Evolution 18:225-239. Bunn, S. E., and J. M. Hughes. 1997. Dispersal and recruitment in streams: evidence from genetic studies. Journal of the North American Benthological Society 16:338-346. Crichton, M. I. 1976. The interpretation of light trap catches of Trichoptera from the Rothamsted Insect Survey. Pages 147-157 in H. Malicky (editor). Proceedings of the Second International Symposium on Trichoptera. Junk, The Hague. Flecker, A. S., and J. D. Allan. 1988. Flight direction in some Rocky Mountain mayflies (Ephemeroptera), with observations of parasitism. Aquatic Insects 10:33-42. Graf, W., U. Grassner, and O. Moog. 1993. The role of Allogamus auricollis (Trichoptera: Limnephilidae) larvae in benthic communities of a 4th order crystalline mountain stream with ecological notes. Pages 297-303 in C. Otto (editor). Proceedings of the 7th International Symposium on Trichoptera. Backhuys, Leiden. Hanski, I. 1998. Metapopulation dynamics. Nature 396:41-49. Hartl, D. L., and A. G. Clark. 1997. Principles of population genetics, 3rd edition. Sinauer, Sunderland, Massachusetts. Hebert, P. D. N., and M. J. Beaton. 1989. Methodologies for allozyme analysis using cellulose acetate electrophoresis. Helena Laboratories, Beaumont, Texas. Hellberg, M. E. 1994. Relationships between inferred levels of gene flow and geographic distance in a philopatric coral, Balanophyllia elegans. Evolution 48:1829-1854. Hughes, J. M., S. E. Bunn, D. A. Hurwood, and C. Cleary. 1998. Dispersal and recruitment of Tasiagma ciliata (Trichoptera: Tasimiidae) in rainforest streams, south-eastern Australia. Freshwater Biology 39:117-127. Hughes, J. M., S. E. Bunn, D. M. Kingston, and D. A. Hurwood. 1995. Genetic differentiation and dispersal among populations of Paratya australiensis (Atyidae) in rainforest streams in southeast Queensland, Australia. Journal of the North American Benthological Society 14:158-173. Hughes, J. M., P. B. Mather, A. L. Sheldon, and F. W. Allendorf. 1999. Genetic structure of the stonefly, Yoraperla brevis, populations: The extent of gene flow among adjacent montane streams. Freshwater Biology 41:63-72. Humpesch, U. H. 1979. Life cycles and growth of Baetis spp. (Ephemeroptera: Baetidae) in the laboratory and in two stony streams in Austria. Freshwater Biology 9:467-479. Jackson, J. K., and V. H. Resh. 1992. Variation in genetic structure among populations of the caddisfly Helicopsyche borealis from three streams in northern California, USA. Freshwater Biology 27:29-42. 51

Lacy, R. C. 1987. Loss of genetic diversity from managed populations: interacting effects of drift, mutation, immigration, selection, and population subdivision. Conservation Biology 1:143-158. Lavandier, P. 1981. Cycle biologique, croissance et production de Rhithrogena loyolaea Navas (Ephemeroptera) dans un torrent pyrénéen de haute montagne. Annales de Limnologie 17:163-179. Lavandier, P. 1982. Evidence of upstream migration by female adults of Baetis alpinus Pict. (Ephemeroptera) at high altitude in the Pyrenees. Annales de Limnologie 18:55-59. Lavandier, P. 1988. Semivoltinisme dans des populations de haute montagne de Baetis alpinus Pictet (Ephemeroptera). Bulletin de la Société d'Histoire Naturelle de Toulouse 124:61-64. Lavandier, P. 1991. Movements of Rhithrogena loyolaea Navas and Baetis alpinus Pictet in a high mountain stream in the Pyrenees. Pages 367-376 in J. Alba-Tercedor and A. Sanchez- Ortega (editors). Overview and Strategies of Ephemeroptera and Plecoptera. The Sandhill Crane Press, Gainesville, Florida. Minshall, G. W. 1988. Stream ecosystem theory: a global perspective. Journal of the North American Benthological Society 7:263-288. Monaghan, M. T., P. Spaak, C. T. Robinson, and J. V. Ward. 2001. Genetic differentiation of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in fragmented alpine streams. Heredity 86:395-403. Morden, C. W., and W. Loeffler. 1999. Fragmentation and genetic differentiation among subpopulations of the endangered Hawaiian mint Haplostachys haplostachya (Lamiaceae). Molecular Ecology 8:617-625. Müller, K. 1982. The colonization cycle of freshwater insects. Oecologia 52:202-207. Olechowska, M. 1981. Life cycle of Rhithrogena loyolaea (Navas) (Ephemeroptera, Heptageniidae) in the Stream Strazyski in the Tatra Mts. Acta Hydrobiologica 23:69-76. Piertney, S. B., and G. R. Carvalho. 1995. Microgeographic genetic differentiation in the intertidal isopod Jaera albifrons Leach. II: temporal variation in allele frequencies. Journal of Experimental Marine Biology and Ecology 188:277-288. Preziosi, R. F., and D. J. Fairbairn. 1992. Genetic population structure and levels of gene flow in the stream dwelling waterstrider, Aquarius ( = Gerris) remigis (Hemiptera: Gerridae). Evolution 46:430-444. Richardson, B. J., P. R. Baverstock, and M. Adams. 1986. Allozyme electrophoresis: a handbook for animal systematics and population studies. Academic Press, San Diego. Saccheri, I., M. Kuussaari, M. Kankare, P. Vikman, W. Fortelius, and I. Hanski. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature 392:491-494. Sartori, M., and P. Landolt. 1999. Atlas de distribution des éphémères de Suisse (Insecta, Ephemeroptera). Centre Suisse de Cartographie de la Faune, Neuchâtel. Schmidt, S. K., J. M. Hughes, and S. E. Bunn. 1995. Gene flow among conspecific populations of Baetis sp. (Ephemeroptera): adult flight and larval drift. Journal of the North American Benthological Society 14:147-157. Slatkin, M. 1985. Gene flow in natural populations. Annual Review of Ecology and Systematics 16:393-430. Slatkin, M. 1993. Isolation by distance in equilibrium and nonequilibrium populations. Evolution 47:264-279. Sode, A., and P. Wiberg-Larsen. 1993. Dispersal of adult Trichoptera at a Danish forest brook. Freshwater Biology 30:439-446. Stacey, P. B., and M. Taper. 1992. Environmental variation and the persistence of small populations. Ecological Applications 2:18-29. Stiven, A. E., and B. R. Kreiser. 1994. Ecological and genetic differentiation among populations of the gastropod Goniobasis proxima (Say) in streams separated by a reservoir in the Piedmont of North Carolina. Journal of the Elisha Mitchell Scientific Society 110:53- 67. Svensson, B. W. 1974. Population movements of adult Trichoptera in a South Swedish stream. Oikos 25:157-175. 52

Sweeney, B. W., D. H. Funk, and R. L. Vannote. 1986. Population genetic structure of two mayflies (Ephemerella subvaria, Eurylophella versimilis) in the Delaware River drainage basin (USA). Journal of the North American Benthological Society 5:253-262. Swofford, D. L., and R. B. Selander. 1981. Biosys-1: a FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72:281-283. Thomas, A. G. B. 1975. Ephéméroptères du sud-ouest de la France. I. migrations d'imagos a haute altitude. Annales de Limnologie 11:47-66. van Dongen, S., T. Backeljau, E. Matthysen, and A. A. Dhondt. 1998. Genetic population structure of the winter moth (Operophtera brumata L.) (Lepidoptera, Geometridae) in a fragmented landscape. Heredity 80:92-100. Vincon, G., and A. G. B. Thomas. 1987. Etude hydrobiologique de la vallée d'Ossau (Pyrénées- Atlantiques). I. Répartition et écologie des Ephéméroptères. Annales de Limnologie 23:95- 113. Wade, M. J., and D. E. McCauley. 1988. Extinction and recolonization: their effects on the genetic differentiation of local populations. Evolution 42:995-1005. Ward, J. V. 1994. Ecology of alpine streams. Freshwater Biology 32:277-294. Waringer, J. A. 1986. The abundance and distribution of caddis flies (Insecta: Trichoptera) caught by emergence traps in the "Ritrodat" research area of the Lunzer Seebach (Lower Austria) from 1980-1982. Freshwater Biology 16:49-60. Waringer, J. A. 1989. Life cycle, horizontal microdistribution and current resistance of Allogamus auricollis (Trichoptera: Limnephilidae) in an Austrian mountain brook. Freshwater Biology 22:177-188. Weir, B. S., and C. C. Cockerham. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38:1358-1370. Wishart, M. J., and J. M. Hughes. 2001. Exploring patterns of population subdivision in thew net-winged midge, Elporia barnardi (Diptera: Blephariceridae), in mountain streams of the south-western Cape, South Africa. Freshwater Biology 46:479-490. 53

Appendix. 1. Allele frequencies for Rhithrogena loyolaea at upstream (1) and downstream (2) sites. Alleles were scored by their relative mobility, with A being slowest. Locus names are as in Table 2. Locus Allele Schwellisee U. Jöri L. Jöri Arosa U. Julierpass Minor Bianco 1 2 1 2 2 1 2 1 2 2 1 2 Gda (n) 25 18 16 18 30 25 12 22 25 20 19 24 A 0.00 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 B 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.00 C 0.00 0.00 0.06 0.06 0.00 0.06 0.04 0.05 0.08 0.00 0.08 0.02 D 0.74 0.92 0.91 0.83 0.93 0.88 0.88 0.80 0.80 0.88 0.84 0.92 E 0.02 0.00 0.03 0.03 0.03 0.00 0.04 0.11 0.02 0.00 0.00 0.06 F 0.08 0.08 0.00 0.03 0.03 0.06 0.04 0.05 0.10 0.13 0.03 0.00 G 0.16 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Mpi (n) 25 18 19 18 30 25 9 22 17 20 22 22 A 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 B 0.08 0.06 0.00 0.08 0.05 0.04 0.00 0.00 0.03 0.08 0.02 0.09 C 0.92 0.92 1.00 0.92 0.95 0.76 0.94 0.96 0.77 0.90 0.98 0.91 D 0.00 0.03 0.00 0.00 0.00 0.18 0.06 0.05 0.21 0.03 0.00 0.00

Pep-D (n) 25 18 17 18 30 25 12 22 25 20 22 24 A 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.02 0.00 0.00 0.00 0.00 B 0.14 0.00 0.15 0.00 0.25 0.02 0.00 0.00 0.24 0.00 0.07 0.06 C 0.60 0.92 0.62 0.81 0.65 0.72 1.00 0.82 0.60 0.93 0.80 0.94 D 0.26 0.08 0.24 0.14 0.08 0.26 0.00 0.16 0.16 0.08 0.14 0.00 E 0.00 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Pgi (n) 25 17 19 18 30 25 12 23 24 19 20 23 A 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.00 0.00 B 0.00 0.06 0.03 0.03 0.02 0.00 0.00 0.07 0.08 0.05 0.13 0.22 C 0.26 0.15 0.32 0.39 0.37 0.38 0.38 0.41 0.27 0.13 0.25 0.28 D 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.16 0.00 0.00 E 0.52 0.15 0.18 0.28 0.30 0.30 0.38 0.22 0.29 0.03 0.25 0.33 F 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 G 0.22 0.56 0.47 0.28 0.32 0.20 0.25 0.17 0.35 0.58 0.38 0.17 H 0.00 0.03 0.00 0.03 0.00 0.12 0.00 0.11 0.00 0.03 0.00 0.00

Pgm (n) 25 18 19 18 30 25 12 22 20 20 22 24 A 0.00 0.00 0.00 0.00 0.03 0.16 0.00 0.00 0.03 0.00 0.02 0.00 B 0.40 0.33 0.53 0.31 0.35 0.20 0.13 0.30 0.23 0.30 0.34 0.17 C 0.18 0.17 0.03 0.06 0.08 0.24 0.25 0.32 0.15 0.23 0.64 0.83 D 0.42 0.50 0.45 0.64 0.53 0.40 0.63 0.39 0.60 0.48 0.00 0.00

Appendix 2. Allele frequencies for Allogamus auricollis at upstream (1) and downstream (2) sites. Locus Allele Livigno Marmorera Ticino R. 1 2 1 2 1 2 Mpi (n) 24 24 18 12 15 17 A 0.04 0.15 0.11 0.00 0.10 0.00 B 0.75 0.60 0.89 0.92 0.67 1.00 C 0.21 0.19 0.00 0.08 0.23 0.00 D 0.00 0.06 0.00 0.00 0.00 0.00

Pep-a (n) 26 25 30 17 25 27 A 0.02 0.02 0.00 0.00 0.06 0.00 B 0.87 0.96 1.00 1.00 0.92 1.00 C 0.08 0.02 0.00 0.00 0.02 0.00 D 0.04 0.00 0.00 0.00 0.00 0.00

Pep-C-1 (n) 26 24 28 17 25 27 A 0.02 0.00 0.00 0.00 0.00 0.00 B 0.85 0.98 0.96 0.94 0.92 1.00 C 0.10 0.02 0.04 0.06 0.08 0.00 D 0.04 0.00 0.00 0.00 0.00 0.00

Pep-C-2 (n) 26 25 27 17 25 21 A 0.02 0.00 0.00 0.00 0.02 0.00 B 0.92 0.84 0.59 0.77 0.78 0.81 C 0.02 0.00 0.00 0.00 0.00 0.00 D 0.04 0.16 0.41 0.24 0.20 0.19

Pgi (n) 26 25 30 17 25 26 A 0.00 0.02 0.00 0.00 0.00 0.00 B 0.00 0.00 0.03 0.00 0.02 0.00 C 0.96 0.90 0.95 1.00 0.90 0.96 D 0.00 0.06 0.00 0.00 0.06 0.04 E 0.04 0.02 0.02 0.00 0.02 0.00

Pgm (n) 26 25 30 17 25 27 A 0.00 0.00 0.07 0.00 0.10 0.09 B 0.02 0.00 0.10 0.09 0.10 0.20 C 0.96 0.92 0.80 0.88 0.78 0.70 D 0.02 0.08 0.03 0.03 0.02 0.00

Appendix 3. Allele frequencies for Baetis alpinus at upstream (1) and downstream (2) sites. Locus Allele Schwellisee U. Jöri L. Jöri Marmorera Arosa 1 2 1 2 2 1 2 1 2 Mpi (n) 15 25 21 5 22 16 14 26 26 A 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.02 0.02 B 0.07 0.00 0.00 0.10 0.00 0.00 0.00 0.08 0.06 C 0.13 0.14 0.14 0.10 0.09 0.22 0.29 0.21 0.27 D 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 E 0.77 0.82 0.81 0.80 0.86 0.66 0.21 0.67 0.60 F 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.00 0.00 G 0.03 0.02 0.05 0.00 0.05 0.13 0.29 0.02 0.04 H 0.00 0.00 0.00 0.00 0.00 0.00 0.14 0.00 0.00 I 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Pep-B (n) 25 27 25 25 25 24 24 27 26 A 0.04 0.00 0.06 0.06 0.02 0.00 0.00 0.04 0.02 B 0.54 0.06 0.26 0.10 0.14 0.19 0.10 0.32 0.27 C 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 D 0.40 0.48 0.60 0.76 0.68 0.69 0.63 0.50 0.46 E 0.02 0.39 0.04 0.08 0.14 0.06 0.25 0.11 0.25 F 0.00 0.07 0.04 0.00 0.02 0.06 0.02 0.04 0.00 G 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Pep-C (n) 25 27 25 25 25 25 25 27 26 A 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.02 B 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C 0.00 0.00 0.02 0.02 0.04 0.00 0.08 0.02 0.00 D 1.00 0.98 0.98 0.96 0.96 1.00 0.92 0.96 0.98 E 0.00 0.02 0.00 0.02 0.00 0.00 0.00 0.00 0.00

Pep-D (n) 25 26 23 23 23 20 21 22 26 A 0.00 0.00 0.00 0.00 0.00 0.05 0.00 0.00 0.00 B 0.20 0.00 0.00 0.00 0.00 0.10 0.02 0.21 0.12 C 0.08 0.00 0.20 0.28 0.07 0.05 0.10 0.05 0.02 D 0.68 0.02 0.78 0.54 0.70 0.75 0.83 0.61 0.42 E 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.00 0.00 F 0.04 0.79 0.02 0.13 0.22 0.05 0.00 0.05 0.33 G 0.00 0.19 0.00 0.04 0.02 0.00 0.00 0.09 0.12

Pgi (n) 25 26 21 25 25 24 25 27 25 A 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 B 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.06 C 0.00 0.06 0.00 0.00 0.04 0.08 0.02 0.07 0.04 D 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 E 0.68 0.94 0.81 0.80 0.88 0.79 0.80 0.76 0.80 F 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 G 0.28 0.00 0.19 0.20 0.08 0.10 0.18 0.11 0.10 H 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Pgm (n) 25 27 21 25 16 24 25 27 26 A 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 B 0.00 0.00 0.07 0.12 0.00 0.00 0.04 0.02 0.04 C 0.08 0.17 0.26 0.36 0.31 0.13 0.22 0.28 0.35 D 0.68 0.83 0.00 0.00 0.00 0.29 0.38 0.35 0.39 E 0.24 0.00 0.55 0.42 0.47 0.54 0.34 0.30 0.23 F 0.00 0.00 0.00 0.00 0.00 0.04 0.02 0.06 0.00 G 0.00 0.00 0.12 0.10 0.22 0.00 0.00 0.00 0.00 H 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Appendix 3. Continued. Locus Allele Minor Livigno U. Julierpass L. Julierpass Bianco 1 2 1 2 1 2 2 1 2 Mpi (n) 15 10 24 19 27 16 24 20 17 A 0.03 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.00 B 0.07 0.00 0.00 0.16 0.02 0.03 0.04 0.10 0.09 C 0.17 0.60 0.06 0.13 0.11 0.25 0.23 0.25 0.38 D 0.00 0.00 0.00 0.16 0.00 0.00 0.00 0.00 0.00 E 0.67 0.20 0.46 0.18 0.80 0.63 0.44 0.53 0.38 F 0.00 0.00 0.00 0.05 0.00 0.00 0.04 0.00 0.00 G 0.07 0.00 0.40 0.24 0.07 0.09 0.25 0.13 0.15 H 0.00 0.00 0.04 0.08 0.00 0.00 0.00 0.00 0.00 I 0.00 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00

Pep-B (n) 26 18 25 25 28 24 26 23 25 A 0.00 0.00 0.04 0.00 0.02 0.08 0.08 0.00 0.04 B 0.10 0.14 0.08 0.02 0.14 0.23 0.14 0.11 0.04 C 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 D 0.50 0.47 0.74 0.76 0.75 0.65 0.58 0.61 0.78 E 0.37 0.33 0.12 0.22 0.09 0.04 0.21 0.28 0.14 F 0.04 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.00 G 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Pep-C (n) 26 18 26 25 28 26 26 24 26 A 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 B 0.00 0.03 0.02 0.00 0.00 0.00 0.00 0.00 0.00 C 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.02 D 1.00 0.97 0.98 1.00 1.00 1.00 1.00 0.98 0.98 E 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Pep-D (n) 16 16 26 25 26 25 22 22 25 A 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 B 0.00 0.00 0.02 0.04 0.00 0.08 0.05 0.00 0.00 C 0.00 0.00 0.04 0.00 0.08 0.18 0.05 0.02 0.06 D 0.56 0.50 0.89 0.90 0.65 0.64 0.52 0.39 0.66 E 0.00 0.00 0.04 0.04 0.04 0.00 0.00 0.00 0.00 F 0.31 0.28 0.02 0.02 0.08 0.10 0.25 0.41 0.24 G 0.13 0.22 0.00 0.00 0.15 0.00 0.14 0.18 0.04

Pgi (n) 26 18 25 25 28 26 26 24 25 A 0.00 0.03 0.00 0.00 0.02 0.00 0.00 0.00 0.00 B 0.06 0.03 0.00 0.00 0.02 0.00 0.00 0.00 0.00 C 0.04 0.00 0.06 0.02 0.00 0.02 0.00 0.02 0.00 D 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 E 0.89 0.94 0.82 0.78 0.73 0.87 0.94 0.92 0.82 F 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 G 0.02 0.00 0.12 0.20 0.21 0.12 0.06 0.06 0.16 H 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02

Pgm (n) 26 18 26 23 28 26 26 24 26 A 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.12 B 0.02 0.08 0.04 0.07 0.00 0.00 0.02 0.38 0.14 C 0.06 0.14 0.17 0.30 0.07 0.19 0.12 0.31 0.08 D 0.31 0.64 0.33 0.17 0.59 0.50 0.37 0.15 0.37 E 0.62 0.08 0.42 0.35 0.32 0.31 0.46 0.17 0.31 F 0.00 0.00 0.04 0.09 0.02 0.00 0.04 0.00 0.00 G 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 H 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00

Appendix 3. Continued. Locus Allele Cadagno Ritom Muesa 1 2 2 1 2 Mpi (n) 15 20 19 25 25 A 0.00 0.00 0.00 0.00 0.00 B 0.00 0.05 0.05 0.08 0.02 C 0.13 0.10 0.24 0.12 0.20 D 0.00 0.00 0.00 0.00 0.00 E 0.80 0.33 0.45 0.70 0.64 F 0.00 0.00 0.00 0.00 0.00 G 0.07 0.40 0.18 0.10 0.14 H 0.00 0.13 0.08 0.00 0.00 I 0.00 0.00 0.00 0.00 0.00

Pep-B (n) 25 24 24 25 25 A 0.10 0.17 0.13 0.02 0.02 B 0.22 0.52 0.27 0.16 0.26 C 0.00 0.00 0.00 0.00 0.00 D 0.44 0.31 0.54 0.58 0.32 E 0.16 0.00 0.06 0.24 0.40 F 0.02 0.00 0.00 0.00 0.00 G 0.06 0.00 0.00 0.00 0.00

Pep-C (n) 25 25 25 26 25 A 0.00 0.00 0.00 0.00 0.00 B 0.00 0.00 0.00 0.00 0.00 C 0.08 0.00 0.04 0.00 0.00 D 0.92 1.00 0.90 1.00 1.00 E 0.00 0.00 0.06 0.00 0.00

Pep-D (n) 25 24 25 24 25 A 0.00 0.00 0.00 0.00 0.00 B 0.08 0.00 0.02 0.00 0.06 C 0.14 0.38 0.12 0.04 0.20 D 0.64 0.52 0.70 0.73 0.70 E 0.00 0.00 0.00 0.00 0.00 F 0.08 0.10 0.16 0.23 0.04 G 0.06 0.00 0.00 0.00 0.00

Pgi (n) 25 23 25 26 25 A 0.00 0.00 0.00 0.00 0.00 B 0.00 0.04 0.00 0.00 0.02 C 0.00 0.04 0.06 0.04 0.12 D 0.00 0.00 0.00 0.00 0.00 E 0.82 0.72 0.74 0.90 0.66 F 0.00 0.00 0.00 0.02 0.00 G 0.18 0.20 0.20 0.04 0.20 H 0.00 0.00 0.00 0.00 0.00

Pgm (n) 22 25 25 26 25 A 0.07 0.00 0.00 0.08 0.02 B 0.11 0.00 0.00 0.12 0.04 C 0.18 0.08 0.00 0.17 0.12 D 0.43 0.70 0.46 0.29 0.36 E 0.21 0.22 0.54 0.25 0.46 F 0.00 0.00 0.00 0.08 0.00 G 0.00 0.00 0.00 0.02 0.00 H 0.00 0.00 0.00 0.00 0.00

CHAPTER 4

LOCAL DIFFERENTIATION AND REGIONAL

HOMOGENEITY: LACK OF EQUILIBRIUM IN THE

POPULATION GENETIC STRUCTURE OF AN

AQUATIC INSECT

with P. Spaak, C.T. Robinson, and J.V. Ward

Abstract Discordant patterns of genetic differentiation (q) at different spatial scales, e.g., local differentiation and regional homogeneity, may be evidence for a lack of equilibrium between gene flow and genetic drift. This lack of equilibrium limits our ability to estimate gene flow of organisms at any spatial scale. Two processes may cause the nonequilibrium pattern: (1) present-day gene flow may be widespread, but founder events and population growth that occur more rapidly than gene flow may cause large genetic differences among populations, or (2) present-day gene flow may be limited, but widespread gene flow in the past may continue to obscure present-day patterns. Hutchison and Templeton (1999) suggested that, on a plot of isolation by distance (IBD) correlating pairwise genetic difference with pairwise geographic distance, large and small variance in q should indicate limited and widespread dispersal, respectively. In a study of the mayfly, Baetis alpinus Pictet (Ephemeroptera: Baetidae) in the Swiss Alps, we observed a large variance in q with AFLP and a small variance with allozymes. We hypothesize that this reflects two stages of range expansion following recolonization of the Alps after the retreat of glaciers. The less sensitive allozymes reflect rapid upstream 60

gene flow into alpine streams from a few glacial refugia, while AFLP markers reflect small-scale local differentiation within and among streams that has occurred more recently. We conclude that we can observe two successive stages of colonization history using 2 different genetic markers.

Introduction Inferring the extent and magnitude of gene flow from the spatial distribution of neutral genetic markers requires the assumption that genetic drift and migration are in equilibrium among spatially separated populations. It is becoming increasingly clear that the assumption of drift- migration equilibrium often may be unfounded (Boileau et al. 1992, Ibrahim et al. 1996, Barber 1999, Pogson et al. 2001, Tsutsui and Case 2001, Monaghan et al. 2001; see chapter 2). Two primary hypotheses have been proposed for the lack of equilibrium. The first is that present-day gene flow may be widespread, but past founder events and rapid population growth obscure present-day patterns, resulting in high levels of genetic differentiation (q) despite high levels of gene flow (Boileau et al. 1992, Ibrahim et al. 1996). Alternatively, present-day gene flow may be limited, but past gene flow that was much more widespread may continue to obscure present-day patterns (Barber 1999, Pogson et al. 2001, Monaghan et al. 2001; see chapter 2). Both mechanisms of nonequilibrium represent historical shifts in the extent of gene flow that have not yet equilibrated with the ongoing process of genetic drift. Values of q represent an average over many generations and recent reductions or expansions in gene flow may have minimal effects on the overall distribution of genetic diversity. Hutchison and Templeton (1999) provided a possible means of detecting the effects of range expansion on population genetic signatures. They examined the variance of q on an isolation by distance (IBD; Wright 1943, Slatkin 1993) plot correlating pairwise genetic differences with pairwise geographic distances. They hypothesized that in the absence of IBD, a large variance of q should indicate reduced dispersal at all spatial scales and a small variance in q should indicate high levels of dispersal at all scales. The alternate mechanisms causing the different degrees of variance correspond to the fact that a lack of IBD can result from two different processes: panmyxia, or low levels of dispersal at all spatial scales (e.g., Peterson and Denno 1998). Panmyxia, or random mating 61 throughout the region from which samples are drawn, breaks down IBD because all populations are equally likely to exchange migrants. This should lead to low levels of variance in q regardless of geographic distance. Very limited dispersal breaks down IBD because few if any populations are likely to exchange migrants. This should lead to high levels of variance regardless geographic distances. In a previous study, we observed what appeared to be a lack of drift- migration equilibrium in larval populations of a stream-dwelling mayfly, Baetis alpinus Pictet (Insecta, Baetidae). Using allozymes, we found a large amount of differentiation between populations separated by standing water bodies only 300-1000 m across, but near homogeneity among headwater catchments in the Swiss Alps (among sites > 100 km distant; Monaghan et al. in press; see chapter 3). We concluded that regional genetic structure reflected historical gene flow i.e., the colonization of the Alps from a few large source populations, and that local genetic structure reflected reduced present-day dispersal. Additionally, allozyme differentiation was evident only across lakes that were geologically old (~10,000 y); not across lakes formed more recently (~1000 y) or across reservoirs (~100 y). We suggested this was because the resolution of allozyme markers was insufficient to detect the more recent reductions in gene flow (Monaghan et al. 2001; see chapter 2). For the present study, we expected that by examining populations using a genetic marker with higher resolution we could detect the effects of genetic drift more rapidly. For example, Berry and Kreitman (1993) found 113 RFLP haplotypes among more than 1000 individuals of Drosophila for which only 2 alleles were detected for the allozyme locus alcohol dehydrogenase (Adh). This suggests that DNA restriction site polymorphism examined with AFLP may detect the effects of genetic drift undetected by protein polymorphism. Therefore, genetic differentiation across the reservoirs may be detected by AFLP. The present study had two major objectives. The first was to test the hypothesis that q from allozymes reflected historical gene flow and that AFLP would show genetic differentiation across geologically younger lakes and across reservoirs by detecting a finer resolution of genetic variation. The second objective was to determine whether the nonequilibrium pattern of genetic structure was consistent when examined with a second marker and to evaluate the degree of scatter in plots correlating pairwise q with pairwise geographic distance. We studied the genetic structure of 15 B. alpinus subpopulations in the Swiss Alps. These 62 subpopulations were a subset of the earlier allozyme study (Monaghan et al. 2001; see chapter 2) and the animals examined with AFLP were a subset of the individuals examined for allozyme variation.

Methods

Study animal Baetis alpinus is a widespread and abundant alpine species, occurring between 200 and 2600 m in streams that do not exceed 20 °C (Sartori and Landolt 1999). Larvae live in streams for a few months to one year depending on water temperature (Humpesch 1979, Lavandier 1988). Periods of emergence from the stream, mating, adult flight, and oviposition typically are asynchronous and extend over several months (Humpesch 1979, Kukula 1997, Céréghino and Lavandier 1998), although individual adults live only a few hours. Upstream bias in adult flight has been observed in B. alpinus (Thomas 1975, Lavandier 1982) and Hershey et al. (1993) reported flight distances of 1.6 to 1.9 km for Baetis in Alaska.

Field collection, allozyme electrophoresis and AFLP fingerprinting We collected animals from 15 study sites located in the headwaters of 3 major river drainages (Rhine, Inn, Ticino) in the Swiss Alps (Fig. 1, Table 1). Field collection of larval B. alpinus and allozyme electrophoresis were described by Monaghan et al. (2001; see chapter 2). We obtained DNA for AFLP fingerprinting from head capsules that were removed prior to allozyme electrophoresis and stored in 96% ethanol. Sample sizes for allozyme and AFLP analysis are listed in Table 1. DNA was extracted and isolated using a procedure modified from Sunnucks and Hales (1996). Head capsules were ground and incubated with 150 mg proteinase K in 200 ml TNES buffer (50 mM Tris, 20 mM EDTA, 400 mM NaCl, 0.1% SDS) at 55 °C for 2-3 hr. After incubation, 25 ml of 5 M NaCl was added, followed by centrifugation at 3000 rpm for 10 min. The supernatant was transferred to new tubes and an equal volume of chloroform was added. Following centrifugation at 8000 rpm for 5 min, the aqueous phase was transferred to new tubes and DNA was precipitated with an equal volume of 100% ethanol (-20°C). DNA was pelleted by 63

Switzerland 1 Arosa Julierpass . e R hin A R C 2 . R D n In 1 G B 2 T Upper Jörisee F ic E 1 in o R Bianco . Cadagno Marmorera Lower Jörisee N Livigno 2 50 km 2

Figure 1. Map of the study sites in the Swiss Alps. A-Upper and Lower Jörisee, B- Marmorera, C-Arosa, D-Livigno, E-Julierpass, F-Bianco, G-Cadagno. Schematics on the right depict the arrangement of sampling locations at each study site. centrifugation (14000 rpm, 20 min) before being air dried and resuspended in 30 µl autoclaved nanopure water. AFLP procedures generally followed those of Vos and Kuiper (1997). An aliquot of total DNA (5 µl) was digested using the restriction enzymes MseI (5 U) and EcoRI (5 U) together in NEB buffer No. 2 (New England Biolabs) with 1x BSA in a 20-µl reaction at 37°C for 2-3 hr. Ligation was started immediately afterward by adding 5 µl of a mix of MseI (27.5 pmol) and EcoRI (2.75 pmol) adapters, T4 DNA Ligase, and 1x ligation buffer. The ligation was carried out overnight at 12°C. Restricted/ ligated DNA was diluted 10-fold in TE0.1 buffer and used as template for the preamplification reaction. Preamplification PCR was performed in 20-µl volumes, with 2 µL diluted restricted/ligated DNA, 1x PCR buffer, 1.5 mM MgCl2, 0.2 mmol of each dNTP, 30 ng MseI +C primer and 30 ng EcoRI primer. PCR was carried out with a Primus 96 Plus thermocycler (MWG Biotech) with 1- min initial denaturation at 94° C followed by 25 cycles of denaturing (94°C 30 sec), annealing (60°C 30 sec), and extension (72°C 1 min). Selective amplification was performed using an MseI+CAG and EcoRI+AC primer combination. PCR was performed in 20-µl volumes, with 1 µL preamplification product as template, 1x PCR buffer with 1.5 mM MgCl2, 0.2 mmol of each dNTP, 30 ng 64

Table 1. Sample size from upstream (-1) and downstream (-2) sampling sites in the 7 study streams. Jörisee had 2 lakes and therefore 3 sampling sites (see Fig. 1). Allozyme n-size is the mean n-size of the 6 loci examined (from Monaghan et al. 2001; see chapter 2). Major Stream Sampling site Dispersal Allozyme n AFLP n drainage barrier Rhine Jörisee Upper Jörisee-1 Lakes (2) 23 11 Upper Jörisee-2 22 5 Lower Jörisee-2 23 7 Marmorera Marmorera-1 Reservoir 22 7 Marmorera-2 23 7 Arosa Arosa-1 None 26 5 Arosa-2 26 5

Inn Livigon Livigno-1 Reservoir 25 12 Livigno-2 24 11 Julierpass Julierpass-1 None 28 12 Julierpass-2 24 6

Ticino Bianco Bianco-1 Lake 23 8 Bianco-2 24 14 Cadagno Cadagno-1 Lake 23 8 Cadagno-2 24 9

MseI primer and 10 ng EcoRI primer. EcoRI+AC primers were labeled (3' end) with IRD700 (MWG Biotech) infrared dye. PCR consisted of 2 min denaturation (94°C), 13 cycles of 30 s 94°C, 30 s 65°C, and 1 min 72°C, with the annealing temperature decreasing from 65°C by 0.7°C increments in cycles 2-13. This was followed by 26 cycles of 30 s 94°C, 30 s 56°C, and 1 min 72°C. All adapter and primer sequences were taken from Vos and Kuiper (1997). Following selective amplification, PCR products were denatured by adding 10 mL microSTOP loading buffer (Microzone Limited, East Sussex, U.K.), incubated at 90° C for 5 minutes, and immediately placed on ice. Fragments were separated by loading 0.5 mL PCR product onto 6% denaturing polyacrylamide gels. Electrophoresis and fragment detection was carried out on a Li-Cor 4200 LongReadIR automatic sequencer using 41-cm plates run at 45 W (~1500 V), 40 mA, and 45 °C. Fragment sizes were determined using microSTEP 24a (Microzone Limited) fluorescent marker and RFLPscan version 3.1 software.

Data analysis We scored 34 AFLP bands that were consistent and repeatable in individuals across PCR reactions and gels. Bands ranged in size from 50 65 to 281 bp and each gel that was scored for data analysis contained 5 marker individuals that, in combination, displayed all 34 bands. Presence or absence of bands was scored in a binary data matrix. Allozyme loci were scored as described previously (Monaghan et al. 2001; see chapter 2).

We calculated q for both AFLP (qAFLP) and allozymes (qallozyme) at three hierarchical levels in a manner analogous to Rank (1992). Within-stream q was calculated for each pair of adjacent sampling sites in each stream. Five streams were fragmented and two were not (Table 1), and so this first level of the hierarchy examined whether genetic differentiation occurred between sites separated by standing water bodies. The next level of the hierarchy was among-stream q, calculated by pooling sampling sites within each stream. Three values were calculated at this level, one for each major drainage. The highest hierarchical level, q among major drainages, was calculated by pooling all sampling sites within each major drainage. We also used analysis of molecular variance (AMOVA; Excoffier et al. 1992) to estimate the partitioning of AFLP variation within and among the major drainages. Calculations of qAFLP and AMOVA were carried out using the ARLEQUIN program version 2.000 (Schneider et al.

2000). Values of qallozyme were calculated using FSTAT (version 2.9.3.1, J. Goudet, University of Lausanne, Switzerland); values here are taken from a previous study (Monaghan et al. in press; see chapter 3). The presence or absence of IBD was examined separately for AFLP and allozyme data by constructing scatterplots of pairwise q against pairwise geographic distance. Correlation coefficients were calculated using Mantel tests, carried out using ARLEQUIN and FSTAT for AFLP and allozyme data, respectively. For both Mantel tests, 1000 permutations were performed. Geographic distances were determined using a 1:300,000 map of Switzerland. For all analyses, 1000 permutations were used to estimate p-values

Results

Measures of qAFLP within streams ranged from 0.00 to 0.24 and generally were higher than values of q from allozymes (qallozyme; Table 2). For both allozyme and AFLP markers, only the lakes Bianco and Cadagno exhibited significant levels of genetic differentiation. The value of qallozyme was significantly different from 0 across the reservoir Livigno, but was low, indicating little or no differentiation (0.02; Table 2). 66

In the Rhine and Inn drainages (Fig. 2, open circles), q was greater among different streams than within single streams for both AFLP and allozymes. This was not the case for the Ticino drainage, where both streams (Bianco and Cadagno) were differentiated, resulting in similar within- and among-stream q (Fig. 2, cross-hatches). Regardless of whether the 2 Ticino streams were included, q among drainages was lower than q among streams. This was the case for both AFLP and allozyme data. AMOVA within and among major drainages partitioned 15.8 % of molecular variance within major drainages, and partitioned none of the variation among major drainages (the value of 1.8 % was not significant; Table 3).

Table 2. Genetic differentiation (q) between populations in the same stream, among streams within the same drainage, and among drainages, estimated with AFLP and allozymes. * = p < 0.05. AFLP q Allozyme q Within streams Bianco 0.156* 0.092* Cadagno 0.240* 0.145* Upper Jörisee 0.087 0.012 Lower Jörisee 0.007 0.007 Livigno 0.024 0.025* Marmorera 0.067 0.025 Arosa -0.041 0.005 Julierpass -0.009 0.004

Within drainages Rhine 0.190* 0.089* Inn 0.103* 0.069* Ticino 0.203* 0.064*

Among drainages 0.051* 0.010*

A 0.25

0.20 Figure 2. Population genetic structure at 3 spatial scales

P 0.15 L

F measured as (A) AFLP and A θ 0.10 (B) allozyme θ. All streams for which within-stream θ 0.05 was not significant (see Table 0.00 2) are plotted here as θ = 0.00. All other points on the B plot represent significant θ (p 0.25 < 0.05). Significant θ 0.20 occurred in both Ticino streams but in none of the 0.15 Rhine or Inn streams. e

m Therefore, among-stream y z 0.10 o l

l estimates for Ticino (+) are a

θ displayed separately from 0.05 Rhine and Inn (O) streams, 0.00 and among-drainage θ was Within Among Among calculated with (+) and streams streams drainages without (O) the Ticino data.

The scatterplot comparison of pairwise θAFLP with pairwise geographic distance exhibited a significant correlation coefficient (r2 = 0.11; Fig. 3), but the trendline slope was near zero (0.0009), indicating the relationship between genetic differentiation and geographic distance was weak. Results for allozyme data were very similar (r2 = 0.11; trendline slope = 0.0004; Fig. 3). The plots differed in the degree of scatter; the variance in θAFLP (0.0126) was greater than in θallozyme (0.0021). The residuals from the regression of θ vs. geographic distance were not related to geographic distance for either marker (data not shown), i.e. scatter did not increase with geographic distance.

Discussion

Genetic differentiation at multiple spatial scales We observed nearly complete agreement between AFLP and allozyme estimates of genetic differentiation (θ) for all spatial scales of interest in this study of Baetis alpinus. At the smallest scale, populations were genetically differentiated across lakes Bianco and Cadagno based on 68

Table 3. Nested analysis of molecular variance (AMOVA) for 127 individuals of Baetis alpinus distributed among 15 sampling sites, 7 streams, and the headwaters of 3 major drainages (see Table 1). Statistics include sums of squared deviations, variance component estimates, the percentage of total variance contributed by each component, and the p-values based on 1000 permutations. Source of variation d.f. SSD Variance % of p-value components total Among drainages 2 38.21 0.12 1.8 0.080 Within drainages 12 155.79 0.97 15.8 <0.001 Within sites 111 563.00 5.07 82.4 <0.001

qAFLP. This supports the allozyme data, thereby providing evidence from two genetic markers that a large amount of genetic structure exists at small spatial scales for B. alpinus (sampling sites were 525 and 1075 m apart at

Bianco and Cadagno, respectively). However, qAFLP indicated no significant differentiation across the Jöri lakes or across the reservoirs Livigno and Marmorera. We suggested that adult flight bahavior caused the genetic differentiation across lakes, and therefore, the lack of differentiation across younger lakes and reservoirs observed with allozyme data was because the fragmentation was too recent (Monaghan et al. 2001; see chapter 2). We can not rule out that the reduction of gene flow is too recent to detect with molecular markers, but we also can not rule out the possibility that dispersal and gene flow continue to occur across the Jöri lakes and the 2 reservoirs. Hierarchical values of q were consistent for both markers, providing further evidence for nonequilibrium between genetic drift and gene flow. For most streams, q was highest among streams and actually decreased among the major drainages (open circles in Fig. 2). AMOVA showed the same pattern, partitioning nearly 16 % of variation within drainages (i.e., among streams), but no variation among drainages. In the fragmented streams (Bianco and Cadagno) q was greatest at the local scale and decreased at progressively larger scales (cross-hatches in Fig. 2). The patterns suggest that most dispersal occurs within streams rather than between different streams and that dispersal can even be limited to discrete reaches (e.g., Bianco, Cadagno, and 3 other lakes examined by Monaghan et al. 2001; see chapter 2). Some study sites were geographically close together despite being in different major drainages, but the IBD relationship was weak for both markers. Thus, limited local dispersal

A 0.6 0.5 0.4 θ 0.3 0.2 AFLP 0.1 0.0

B 0.6 0.5 Figure 3. Scatterplots of q θ 0.4 against geographic distance 0.3 for (A) AFLP (Mantel test, r = 0.322; p = 0.013) and 0.2 (B) allozymes (r = 0.324; p

Allozyme 0.1 = 0.002). Filled circles 0.0 represent the within-stream pairwise q-values 0 20 40 60 80 100 120 140 presented in Table 2. Geographic distance (km)

should result in a q-value among drainages being higher on the y-axis in Fig. 2.

Relative influences of gene flow and drift Pairwise q for all sampling sites plotted against geographic distance failed to detect an IBD pattern for either marker, indicating genetic drift and gene flow are not in equilibrium (Slatkin 1993, Barber 1999) for B. alpinus. Other workers have reached similar conclusions of nonequilibrium for the coral Balanophyllia elegans (Hellberg 1995), the treefrog Hyla arenicolor (Barber 1999), the brook charr Salvelinus fontinalis (Castric et al. 2001), and the ant Linepithema humile (Tsutsui and Case 2001). These authors have suggested the lack of equilibrium results from (1) large-scale environmental changes that have caused recent reductions in gene flow (Hellberg 1995, Barber 1999), or (2) rapid range expansion into new areas (Castric et al. 2001, Tsutsui and Case 2001). The mechanism is essentially the same, in that large shifts in the extent of gene flow require time for drift to equilibrate. The rate of approach to equilibrium largely depends on migration rate and population size, assuming the rate of genetic drift is constant (Crow and Aoki 1984). With 70 large population size (i.e., the inverse of population size is much less than zero), estimates of q approach equilibrium in a number of generations approximately equal to the inverse of migration rate. An interesting characteristic of the two IBD plots of the present study was the greater scatter of qAFLP compared to qallozyme. The large amount of scatter in qAFLP at all geographic distances quite closely resembles what Hutchison and Templeton (1999) termed a Case III relationship, while the reduced scatter in qallozyme resembles a Case II relationship. Their empirical data came from microsatellite analysis of eastern collared lizards (Crotaphytus collaris collaris) in two different geographic regions. Both regions have been recolonized by the species following the retreat of Wisconsin glaciers. They observed the Case II relationship (no IBD, low scatter) in a region with fewer barriers to dispersal and the Case III relationship (no IBD, high scatter) in a region with more dispersal barriers due to habitat fragmentation (Hutchison and Templeton 1999). In the present study, we observed a difference in scatter with the same individuals and populations examined using two different genetic markers, rather than in two different regions with the same marker. Because restriction site variation is greater than electrophoretic variation in proteins (Berry and Kreitman 1993), AFLP markers presumably should exhibit more subtle (i.e., more recent) effects of genetic drift than allozymes. Thus, the two different relationships may well represent the progression of B. alpinus from a Case II to a Case III relationship following range expansion in the deglaciated Alps. Recolonization of Alpine headwaters after glacial retreat may have been rapid; Baetis spp. typically are early colonizers (Mackay 1992) and Milner (1987) observed Baetis to colonize streams within 15 years after glacial retreat. Movement from downstream refugia upward into headwaters probably was much more prevalent than movement among different headwater streams because of strong upstream flight bias for this species (Thomas 1975, Lavandier 1982). Both markers employed in the present study indicated that movement among streams is limited and probably reflect the pattern of post-glacial colonization. Pairwise q at the smallest spatial scale (within streams, 100s - 1000s m apart) revealed that local differentiation can be pronounced and that dispersal probably is quite limited (Monaghan et al. 2001; see chapter 2). We conclude that the lack of equilibrium in B. alpinus results from local differentiation and large scale homogeneity, caused by a combination of 71 localized, present-day gene flow and regional, historical colonization of the Alps after the retreat of Pleistocene glaciers.

Acknowledgements

Mäggi Hieber helped with the field collection of nearly every single animal and her help is very much appreciated. We also thank Peter Burgherr, Christine Calvino, Christine Dambone, Massimiliano Gili, Christa Jolidon, Sandra Lass, Florian Malard, Friederike Mösslacher, Marcos de la Puenta Nilsson, Karsten Rinke, Sven Schalla, and Bettina Wagner for their help in the field and laboratory. Christoph Werlen, Kirsten Lawlor, Vladimir Sentchilo, Jan Roelof van der Meer, and Rik Eggen in the MIX molecular microbiology group at EAWAG provided help and support with sequencing gels. Part of the research was conducted in the Zoology Department at Otago University, Dunedin, New Zealand, and author MTM thanks Mike Roy, Graham Wallis, Carolyn Burns, Patrick Dwyer, Karen Judge, Gigi Ostrow, Cécile Perrin, and John Waters for their hospitality, help, and encouragement. Sophie Karrenberg van der Nat and Sandra Lass provided critical reviews of an earlier draft. Research was funded by grant No. 31-50444.97/1 from the Swiss National Science Foundation.

Literature Cited

Excoffier, L., P. Smouse, and J. Quattro. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131:479-491. Hellberg, M. E. 1995. Stepping-stone gene flow in the solitary coral Balanophyllia elegans: equilibrium and nonequilibrium at different spatial scales. Marine Biology 123:573-581. Hershey, A. E., J. Pastor, B. J. Peterson, and G. W. Kling. 1993. Stable isotopes resolve the drift paradox for Baetis mayflies in an Arctic River. Ecology 74:2315-2325. Humpesch, U. H. 1979. Life cycles and growth of Baetis spp. (Ephemeroptera: Baetidae) in the laboratory and in two stony streams in Austria. Freshwater Biology 9:467-479. Hutchison, D. W., and A. R. Templeton. 1999. Correlation of pairwise genetic and geographic distance measures: inferring the relative influences of gene flow and drift on the distribution of genetic variability. Evolution 53:1898-1914. Ibrahim, K. M., R. A. Nichols, and G. M. Hewitt. 1996. Spatial patterns of genetic variation generated by different forms of dispersal during range expansion. Heredity 77:282-291. Kukula, K. 1997. The life cycles of three species of Ephemeroptera in two streams in Poland. Hydrobiologia 353:193-198. Lavandier, P. 1982. Evidence of upstream migration by female adults of Baetis alpinus Pict. (Ephemeroptera) at high altitude in the Pyrenees. Annales de Limnologie 18:55-59. Lavandier, P. 1988. Semivoltinisme dans des populations de haute montagne de Baetis alpinus Pictet (Ephemeroptera). Bulletin de la Société d'Histoire Naturelle de Toulouse 124:61-64. Mackay, R. J. 1992. Colonization by lotic macroinvertebrates: a review of processes and patterns. Canadian Journal of Fisheries and Aquatic Sciences 49:617-628. Milner, A. M. 1987. Colonization and ecological development of new streams in Glacier Bay National Park, Alaska. Freshwater Biology 18:53-70. Monaghan, M. T., P. Spaak, C. T. Robinson, and J. V. Ward. 2001. Genetic differentiation of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in fragmented alpine streams. Heredity 86:395-403. Monaghan, M. T., P. Spaak, C. T. Robinson, and J. V. Ward. in press. Population genetic structure of 3 Alpine stream insects: influences of gene flow, demographics, and habitat fragmentation. Journal of the North American Benthological Society. Peterson, M. A., and R. F. Denno. 1998. The influence of dispersal and diet breadth on patterns of genetic isolation by distance in phytophagus insects. American Naturalist 152:428-446. Pogson, G. H., C. T. Taggart, K. A. Mesa, and R. G. Boutilier. 2001. Isolation by distance in the Atlantic cod, Gadus morhua, at large and small geographic scales. Evolution 55:131- 146. Rank, N. E. 1992. A hierarchical analysis of genetic differentiation in a montane leaf beetle Chrysomela aeneicollis (Coleoptera: Chrysomelidae). Evolution 46:1097-1111. Sartori, M., and P. Landolt. 1999. Atlas de distribution des éphémères de Suisse (Insecta, Ephemeroptera). Centre Suisse de Cartographie de la Faune, Neuchâtel. Schneider, S., D. Roessli, and L. Excoffier. 2000. ARLEQUIN version 2.000: A software for population genetics data analysis. Genetics and Biomery Laboratory, University of Geneva, Switzerland, Geneva, Switzerland. Slatkin, M. 1993. Isolation by distance in equilibrium and nonequilibrium populations. Evolution 47:264-279. Sunnucks, P., and D. F. Hales. 1996. Numerous transposed sequences of mitochondrial cytochrome oxidase I-II in aphids of the genus Sitobion (Hemiptera: Aphididae). Molecular Biology and Evolution 13:510-524. Thomas, A. G. B. 1975. Ephéméroptères du sud-ouest de la France. I. migrations d'imagos a haute altitude. Annales de Limnologie 11:47-66. Tsutsui, N. D., and T. J. Case. 2001. Population genetics and colony structure of the Argentine ant (Linepithema humile) in its native and introduced ranges. Evolution 55:976-985. Vos, P., and M. Kuiper. 1997. AFLP Analysis. Pages 115-131 in G. Caetano-Annoles and P. M. Gresshoff (editors). DNA Markers: protocols, applications, and overviews. Wiley-VCH, New York. Wright, S. 1943. Isolation by distance. Genetics 28:114-138. 73

CHAPTER 5

TAXON RICHNESS AND NESTEDNESS OF BENTHIC

MACROINVERTEBRATES IN FRAGMENTED ALPINE

STREAMS: IMPLICATIONS FOR FRESHWATER

CONSERVATION

with C.T. Robinson, J.V. Ward, and P. Spaak.

Abstract Habitat fragmentation results in the loss of habitat area, an increase in edge, and a loss of connectivity among fragments. Consequences include a reduction in species richness and changes in assemblage structure caused by habitat alteration, the isolation of subpopulations, or both. In streams, standing water bodies (natural lakes and reservoirs) can disrupt longitudinal connectivity and fragment flowing water habitats into discreet reaches. We investigated the effects of habitat fragmentation by collecting and identifying benthic macroinvertebrates from 22 sites in 10 alpine streams. We used presence-absence data to test whether taxon richness was reduced in fragments and whether taxon assemblages in fragments were nested within assemblages of unfragmented sites. Total taxon richness, all sites combined, was 69, ranging from 6 to 27 for individual sites. While total richness was not significantly different between fragmented and unfragmented sites, both Ephemeroptera and Diptera (excluding Simuliidae) richness was significantly reduced in stream fragments. Assemblages in fragments were not nested subsets of unfragmented sites, indicating that site- and stream-specific habitat conditions, rather than extinction and lack of subsequent recolonization, were primary determinants of assemblage 74

structure in all sites. Collectively, our results suggest taxon turnover among sites and among streams (beta-diversity) is an important component of biodiversity in alpine streams. We also suggest that our ability to understand the relative importance of colonization in structuring macroinvertebrate assemblages would be enhanced by more studies of dispersal at both local and regional scales.

Introduction Habitat fragmentation is considered a major threat to the preservation of biodiversity (Wilcox and Murphy 1985). Habitat fragmentation can reduce species diversity (Klein 1989, Newmark 1991, Kattan et al. 1994) and change assemblage composition (Margules et al. 1994, Davies and Margules 1998) through a variety of inter-related processes (Robinson et al. 1992). The primary physical effects of fragmentation include the loss of habitat area, an increase in the proportion of edge habitat, and the isolation of fragments (Saunders et al. 1991, Andren 1994). The reduction of total habitat area often results in reduced species richness, primarily as a function of species-area relationships and the fact that larger areas contain more kinds of habitat (Rosenzweig 1995, Bender et al. 1998). An increase in the amount of edge relative to the amount of interior habitat can reduce species richness because of "edge effect" alterations to habitat (e.g., greater fluxuations of light, temperature, wind; Lovejoy et al. 1986). Isolation of fragments also can result in the formation of subpopulations which may have a greater probability of extinction because of environmental or demographic stochasticity (Pimm et al. 1995). The resulting metapopulation may become more dependent on dispersal among subpopulations for persistence (Hanski 1998). Like most ecosystems, streams are subject to habitat fragmentation which threatens to alter their natural patterns of biodiversity (Zwick 1992, Allan and Flecker 1993, Dynesius and Nilsson 1994, Ward and Tockner 2001). Lentic water bodies along a stream course represent chemically (Kling et al. 2000) and biologically (Breitenmoser-Würsten and Sartori 1995, Willis and Magnuson 2000) distinct habitats and act to fragment streams into distinct flowing reaches. Fragmentation can result from natural lakes or from anthropogenic reservoirs. Both disrupt connectivity between reaches, but differ in their specific effects on downstream habitat. Dispersal by some benthic macroinvertebrates may be quite limited (Bunn 75 and Hughes 1997) and lentic habitat can act as a barrier to dispersal (Monaghan et al. 2001; see chapter 2). Reservoirs also tend to be much larger than lakes (Ryding and Rast 1989), potentially isolating upstream and downstream reaches more than natural lakes. The different effects of lakes and reservoirs on downstream habitat result, in part, from the differences in their effect on temperature and flow regimes. Lakes can act to buffer fluctuations in flow, stabilize temperature regime, alter food sources, and stabilize habitat, all of which can result in changes to assemblage structure (Richardson and Mackay 1991). The presence and operation of reservoirs can have much larger effects on flow and temperature regimes (Ward and Stanford 1979) that may extend several km downstream (Ward 1976, Vinson 2001). Our first hypothesis was that taxon richness (a) would be reduced in fragmented sites relative to sites in unfragmented streams. We also tested whether taxon richness in fragmented streams (g, determined by pooling taxa from multiple sites in a stream) was reduced relative to taxon richness in unfragmented streams. Our second hypothesis examined whether reduced a or g resulted from habitat alteration or from local extinction and a lack of recolonization. We sought to assess the relative importance of each mechanism by testing whether or not taxon assemblages among study sites were "nested." Sites are considered to be nested when the species assemblages of depauperate sites comprise a subset of species assemblages in progressively richer sites (Patterson and Atmar 1986, Patterson 1987). The idea of nestedness first was applied to land-bridge archipelagos, where nested structure can result from “faunal relaxation,” or the loss of species from extinction following the isolation of islands by Pleistocene sea-level rise (Patterson and Atmar 1986, Wright et al. 1998). Subsequent theoretical and empirical work suggests that nested assemblages observed in more recently fragmented habitats also result from extinction (Blake 1991, Taylor 1997, Taylor and Warren 2001). We hypothesized that if extinction is an important determinant of assemblage structure, taxon assemblages in fragments would be nested subsets of those in unfragmented sites. Thus, the occurrence of nestedness would imply that recolonization ability (and dispersal) is limited and, therefore, is an important process determining assemblage composition in fragmented streams. Alternatively, if fragments are not nested subsets of unfragmented sites, it is an indication taxon turnover has occurred, that dispersal is widespread, and that habitat availability is an important factor structuring assemblages. 76

We conducted the study in the headwaters of three major drainages (Rhine, Inn, and Ticino rivers) of the Swiss Alps, where an abundance of streams, lakes, and reservoirs provides the opportunity to test predictions about the ecological effects of habitat fragmentation in aquatic ecosystems. We examined taxon richness and nestedness of benthic macroinvertebrates from 22 sampling sites in 10 streams. Sixteen sites were in fragmented streams and 6 sites were in unfragmented streams. The goal of the study was to gain an understanding of how stream habitat fragmentation by lakes and reservoirs affects benthic marcoinvertebrate assemblages.

Switzerland 1 H. Arosa Rhine R. H G I. Julierpass A J. Muesa D Inn R. 2 F E Ticino R. 1 B J I C 2 1 F. Cadagno / Ritom A. Schwellisee G. Upper / Lower Jöri N B. Bianco C. Minor 50 km D. Livigno E. Marmorera 2 2

Figure 1. Map of the 10 study streams in the Swiss Alps (left) and schematic representations of the organization of sampling sites in each stream. Geographic distances between sites are reported in Table 1.

Methods Sample sites Eight of the study streams had 2 sampling sites, located above and below a lake (Bianco, Schwellisee, Minor) or reservoir (Livigno, Marmorera), or along an unfragmented length of stream of comparable distance (Arosa, Julierpass, Muesa; Fig. 1; Table 1). Two of the streams had 3 sampling sites. Jörisee had sites located above, between, and below two lakes (Upper and Lower Jörisee; Table 1). Cadagno/Ritom had 1 site above Lake Cadagno, one site located between Lake Cadagno and Lake 77

Table 1. The 10 streams in which the 22 study sites were located (including codes used for Table 2 and Fig. 2) and fragmenting feature, sample site elevation, and geographic distance between sampling sites. Stream Code Fragmenting Upper / lower Distance between feature elevation sampling sites (m) (m) Schwellisee SCH Lake 1935 / 1930 350 Bianco BIA Lake 2080 / 2076 525 Minor VLM Lake 2340 / 2325 375 Cadagno/Ritom Cadagno CAD Lake 1940 / 1900 1075 Ritom RITO Reservoira 1900 / 1780 4500 Jörisee Upper Jöri JUP Lake 2525 / 2495 550 Lower Jöri JLO Lake 2495 / 2320 975 Livigno LIVN Reservoir 1910 / 1660 10000 Marmorera MARM Reservoir 1700 / 1450 7750 Arosa A None 1940 / 1930 375 Muesa M None 2225 / 2200 280 Julierpass J None 2310 / 2205 625 a Ritom is a reservoir that is an enlarged natural lake. For the analyses it was considered a reservoir.

Ritom (a reservoir that is an enlarged lake), and one site located below Lake Ritom. There were no tributaries entering any of the study streams between sample sites. All sampling sites were in alpine tundra above tree line, except for Marmorera-1 and -2, Livigno-2, and Ritom-2. Elevation ranged from 1450 to 2525 m a.s.l. and altitudinal differences between upper and lower sites ranged from 4 to 250 m (Table 1).

Macroinvertebrate collection and identification Benthic macroinvertebrates were sampled semi-quantitatively at each site using a 100-mm mesh kick net. Sampling was standardized to 5 minutes each time and all habitat and substrate types within a 50-m reach of stream were sampled. Habitat types typically encompassed riffles, runs, and pools, while substrata included moss, organic detritus, and sediment ranging from sand to large boulders. Samples were preserved in the field with formalin or alcohol. Each site was sampled at least 3 times, once each in spring, summer, and autumn in either 1998 or 1999. Eleven of the 22 sites were sampled once in mid-winter: both sites at Schwellisee, Bianco, Cadagno, Marmorera, and Arosa; the upper site of Livigno; and lower sites of Moesa and Julierpass. At 4 of those sites we observed some taxa only in 78 the winter samples (see Results). The other sites were inaccessible in winter because of snow cover, avalanche danger, and road closure. In the laboratory, samples were sorted using a dissecting microscope (10X). Taxon richness was determined for each sampling site (a) and for each stream (g) based on the combined occurrences from all samples. Taxonomic resolution varied among groups. Ephemeroptera, Plecoptera, Trichoptera, and Blephariceridae and Simuliidae (Diptera) were identified to species with few exceptions. Within the Plecoptera we identified most Isoperla, Leuctra, Nemoura, and Protonemura to genus because of taxonomic uncertainty in the larvae (Aubert 1959, V. Lubini pers. comm.). Within the Trichoptera we identified 6 species of the genus Rhyacophila but a number of other Rhyacophila were grouped sensu stricto (Waringer and Graf 1997). Sericostoma (Trichoptera) also was identified only to genus. Most other Diptera were identified to family, with the exceptions of Atherix, Dicranota, Hexatoma, Rhabdomastix, and Rhypholophus. We grouped Oligochaeta (Annelida), Elmidae (Coleoptera), and Hydracharina, and we observed only one species of Gammaridae (Amphipoda) and Planariidae (Turbellaria).

Data analysis The first step in data analysis was to determine if there was an effect of drainage basin (Rhine, Inn, Ticino) on assemblage structure. We used cluster analysis to group sampling sites to test whether sites clustered randomly or clustered according to drainage. Our assumption was that a random clustering of sites would mean that assemblage structure was not influenced by drainage, and therefore that sites could be considered independent. We calculated all pairwise Euclidean linkage distances from the presence-absence data matrix and then created a clustering tree using Ward's (1963) method. Distance calculation and clustering were performed using SATISTICA version 5.1. Because no major drainage effect was apparent (see Results), subsequent analyses were performed using all sites. We used ANOVA on log-transformed data to compare a among unfragmented sites (n = 6), lake sites (n = 11), and reservoir sites (n = 5). Post-hoc testing was carried out using Tukey's honest significant difference test. We also analyzed each major taxonomic group (Ephemeroptera, Plecoptera, Trichoptera, and Diptera) separately using the same analysis. We performed a 2nd analysis of the Diptera that excluded the Simuliidae. This was done to more closely examine differences among sites that may arise because of Simuliidae, a group closely associated with lake outlets 79

(Ciborowski and Adler 1990, Wotton 1992, Malmqvist et al. 1999). We also used ANOVA to compare between fragmented (n = 7) and unfragmented (n = 3) streams. For this analysis, the upper and lower reaches of Jörisee and Cadagno/Ritom were considered as separate streams, thus from the middle sites of these streams (see Fig. 1) were used for 2 calculations of . We examined the degree of nestedness using the Temperature Calculator of Atmar and Patterson (1993, 1995). The calculated system "temperature" (T) reflects the degree of order present in presence-absence matrices and ranges from T = 0L (perfectly nested) to T = 100L (random). The program packs the data into rows (sampling sites) and columns (species) such that nestedness is maximized and Monte Carlo randomization (we used 100 permutations) is used to test whether T resulting from the program's packing of the matrix is significantly lower (more nested) than if the matrix was packed randomly. We examined the packed ordering of sites with the hypothesis that unfragmented sites would comprise the upper, more taxa-rich rows, and with fragmented sites nested below them in the matrix. We examined taxa turnover (-diversity) among sample sites in each stream using the Sorenson index C: C = 2j/(a+b), where j = the number of taxa found in both sites, a = the number of taxa in site a, and b= the number of taxa in site b (Magurran 1988). Values of C range from 0 to 1, with a value of 1.0 indicating identical taxa composition and a value of 0.0 indicating that sites share no taxa in common. An additional measure of taxa turnover along each stream and among different streams was

Whittaker’s , calculated as W = /-1, where = the total number of taxa in the system (i.e. within a stream or over the whole study area) and = mean taxon richness at each site (Whittaker 1960). The lower the similarity of taxa assemblages, the higher the value of W. We compared C and βW among lake, reservoir, and unfragmented sites using ANOVA on transformed (log arcsine square root for C; log X+1 for W) data (Zar 1984).

Results We collected 69 taxa, made up of 12 Ephemeroptera, 12 Plecoptera, 20 Trichoptera, 20 Diptera, and 5 others: Crenobia alpina (Planaria), Gammarus fossarum (Amphipoda), Elmidae (Coleoptera), Hydracarina (Acarina), and Oligochaeta (Table 2). Sampling sites within streams Table 2. Presence (*) and absence (blank) of the 69 taxa at each study site. Codes for sites are from Table 1, with 1 and 2 designating upstream and downstream sites, respectively. Lake-fragmented streams Reservoir-fragmented streams Unfragmented streams

SCH BIA VLM CAD JUP JLO LIVN MARM RITO A M J 1 2 1 2 1 2 1 2 1 2 2 1 2 1 2 2 1 2 1 2 1 2 EPHEMEROPTERA Baetis alpinus * * * * * * * * * * * * * * * * * * * * * * Baetis melanonyx * Baetis rhodani * * * * * * Ecdyonurus alpinus * * * * * Ecdyonurus helveticus * * Ecdyonurus pitceti * * * * * * * * * * Ecdyonurus venosus * * * * Epeorus alpicola * * * * * Rhithrogena alpestris * Rhithrogena endenensis * * * * * * Rhithrogena loyolaea * * * * * * * * * * * * * * * * Rhithrogena nivata * * *

PLECOPTERA 80 Dictyogenus alpinum * * * * * * * * * * * * * * Dinocras ferreri * Isoperla carbonaria/rivulorum * Isoperla sp. * * * * * * * Luectra sp. * * * * * * * * * * * * * * * * * * * * * Nemoura mortoni * * * * * * * * * * Nemoura sp. * * * * Perla grandis * Perlodes intricatus * * * * * * Protonemura sp. * * * * * * * * * * * * * * * * * * Rhabdiopteryx alpina * * * * * Siphonoperla montana * TRICHOPTERA Acrophylax zerberus * * * * Allogamus auricollis * * * * Chaetopteryx sp. * Consorophylax * * * consors/styriacus Drusus biguttatus * * * Drusus destitutus * * * Drusus discolor * * * * * * * * * * * * * * * Drusus mixtus * * * * Drusus monticola * * * * * Helesus rubricollis * * *

Table 2. Continued. Lake-fragmented streams Reservoir-fragmented streams Unfragmented streams

SCH BIA VLM CAD JUP JLO LIVN MARM RITO A M J 1 2 1 2 1 2 1 2 1 2 2 1 2 1 2 2 1 2 1 2 1 2 Philopotamus ludificatus * * * Plectrocnemia conspersa * * * * * Rhyachophila glareosa * * * * * Rhyachophila intermedia * * * * * * * * * * Rhyachophila laevis/producta * Rhyachophila s.s. * Rhyachophila torrentium * Rhyachophila tristis * * * * Rhyachophila vulgaris * * * * * * * * * * * Sericostoma sp. * DIPTERA Atherix * * * CHIRONOMIDAE * * * * * * * * * * * * * * * * * * * * * * Dicranota * * * * * * * * * * * * EMPIDIDAE * * * * * Hexatoma * * * * * LIMONIIDAE * * * * * * * * * Liponeura decipiens * * * * * * * * PSYCHODIDAE * * * * * * * * * 81 Rhabdomastix * * Rhypholophus * STRATIOMYIDAE * * TIPULIDAE * * * * * SIMULIIDAE Prosimulium latimucro * * * * Prosimulium rufipes * * * * * * * * * * * * * * * Simulium argyreatum * * * * * * * * * Simulium carthusiensis * * * * Simulium cryophilum * * Simulium monticola * Simulium nölleri * * * * Simulium tuberosum grp * * * * * OTHERS Crenobia alpina * * * * * * * * * * * * * * * * * * * * * * Oligochaeta * * * * * * * * * * * * * * * * * * * * * * Elmidae * * Gammarus fossarum * Hydracarina * * * TOTAL α 18 20 14 16 6 25 24 20 13 14 27 19 17 20 23 26 19 21 25 25 22 20

8 Figure 2. Cluster diagram showing grouping of the 7 study sites by Euclidean 6 distance between 5 assemblage structures. Site codes are from Table 1, 4 with 1 and 2 designating 3 upstream and downstream 2 sites, respectively. Major drainage is indicated by R 1 (Rhine), I (Inn), or T (Ticino). Asterix (*)

*J-2 (I) indicates sites in *J-1 (I) LIVN-2 (I) LIVN-2 (I) LIVN-1 BIA-2 (T) BIA-2 (T) BIA-1 *M-2 (T) (R) *A-2 (R) *A-1 *M-1 (T) VLM-1 (I) VLM-2 (I) RITO-2 (T) RITO-2 JLO-2 (R)

JUP-2 (R) JUP-1 (R) unfragmented streams. CAD-2 (T) CAD-1 (T) SCH-2 (R) SCH-2 SCH-1 (R) SCH-1 Euclidian linkage distance MARM-2 (R) MARM-2 (R) MARM-1

clustered together in 6 of 10 cases and the presence of a fragmenting feature did not appear to influence whether or not sites clustered together (Fig. 2). The 3 sites of Cadagno (Cadagno -1 and -2, Ritom-2) clustered together, as did the 2 sites in Marmorera, Livigno, Bianco, Arosa, and Julierpass. Neither sites nor streams were clustered by major drainage, indicating no drainage effect on assemblage structure. Subsequent analyses were therefore carried out on the combined data set. Taxon richness of individual sites () ranged from 6 to 27 (Table 2). There were no significant differences in total -richness among stream types: lake-fragmented sites (17.9 M 6.2), reservoir-fragmented sites (21.0 M 3.5), and unfragmented sites (22.0 M 2.5; mean M 1 SD); however, Ephemeroptera was significantly different among stream types (p = 0.010; Table 3). Post-hoc tests indicated Ephemeroptera was reduced in lake-fragmented sites (2.6 M 1.4) relative to reservoir-fragmented sites (5.0 M 0.7; p = 0.019) and unfragmented streams (4.5 M 1.9; p = 0.05; Table 3). Total Diptera richness, i.e. when Simuliidae were included in the analysis, was not significantly different among site types (Table 3); however, Diptera was significantly different among stream types when Simuliidae were excluded from the analysis (p = 0.009; Table 3). Post-hoc tests showed lake (2.8 M 1.0) was significantly lower than unfragmented (4.3 M 0.8; p = 0.007). Total ranged from 18 to 33 and was not significantly different among streams (F2,9 = 0.385; p = 0.691; Table 4), except for the Ephemeroptera and Diptera (excluding Simuliidae), as above (statistics not presented). 83

Table 3. Mean (SD) taxon richness among sites and ANOVA results for all taxa combined and for the 4 major groups (Diptera were analyzed with and without Simuliidae). Shared superscript letters indicate a lack of significant difference (Tukey's post-hoc test p < 0.05). Taxon Mean taxon richness of sites ANOVA p F2,19 Lake- Reservoir- fragmented fragmented Unfragmented All 17.9 (6.2) 21.0 (3.5) 22.0 (2.5) 1.55 0.240 Ephemeroptera 2.6 (1.4)a 5.0 (0.7)b 4.5 (1.9)b 5.87 0.010 Plecoptera 3.7 (1.7) 3.8 (0.8) 4.8 (0.8) 1.36 0.280 Trichoptera 3.9 (2.7) 3.8 (2.9) 4.2 (1.9) 0.03 0.970 Diptera 5.2 (1.6) 5.2 (2.3) 6.0 (0.9) 0.54 0.590 excluding 2.8 (1.0)a 3.6 (0.5)ab 4.3 (0.8)b 6.13 0.009 Simuliidae

Assemblage structure was nested among the 22 study sites (T = 31.4 °; p < 0.001) but fragmented sites did not form nested subsets of unfragmented sites (Fig. 3). For example, the first 2 rows of the packed data set were fragmented sites (Lower Jöri-2, Minor-2). The frequency distribution of taxa among sites generated by the Temperature Calculator showed that 46 of the 69 taxa (Fig. 4) were "rare," occurring in less than 25% of sample sites (Gaston 1994). Twenty taxa (29 %) occurred in less than 10 % of sites (Fig. 4).

Lower Jöri - 2 Minor - 2 *Julierpass - 1 *Julierpass - 2 Cadagno - 1 Marmorera - 2 Ritom - 2 Schwellisee - 2 *Moesa - 1 *Arosa - 2 Cadagno - 2 Marmorera - 1 *Moesa - 2 *Arosa - 1 Livigno - 1 Schwellisee - 1 Livigno - 2 Upper Jöri - 2 Bianco - 2 Upper Jöri - 1 Bianco - 1 Minor - 1

Figure 3. The packed data matrix of nested (p < 0.001) taxon assemblages in the 22 study sites. An asterix (*) indicates sites that were in unfragmented streams. Four taxa were found in all sites (Baetis alpinus, Crenobia alpina, Chironomidae, and Oligochaeta) and therefore are compressed into a single (left-most) column. 84

Taxon turnover within streams (b-diversity) was not significantly different among stream types, measured either as C or bW (Table 4). We observed a higher degree of similarity in Bianco than all other fragmented sites, and

ANOVA excluding Bianco indicated significant differences in bW among stream types (p = 0.043), but not in C. Post-hoc testing showed b W was lower in lake-fragmented sites than unfragmented sites (p = 0.042). Taxon turnover among streams (bW for the whole study) was 3.70. A single taxon, Rhyachophila sensu stricto, was found in only unfragmented streams, while 27 taxa (~ 40 %) were observed only in fragments.

10 Figure 4. Frequency distribution of the 69 taxa 5 among the 22 sampling sites.

Nr. of sites occupied The taxa to the right of the 0 arrow were defined as rare 0 10 20 30 40 50 60 70 by their occurrence in fewer Taxon than 25 % of sites.

Discussion Reduced taxon richness of Ephemeroptera and Diptera Our first hypothesis was that taxon richness was reduced in fragmented sites. While total taxon richness was not significantly influenced by habitat fragmentation, two taxonomic groups exhibited marked reduction in richness: Ephemeroptera a was reduced in lake- fragmented sites compared with the reservoir-fragmented sites and with the unfragmented sites, and Diptera a (excluding Simuliidae) was reduced in lake-fragmented sites compared to unfragmented sites. Ephemeroptera species largely missing from lake-fragmented sites included Ecdyonurus alpinus, E. picteti, and Rhithrogena endenensis. The similar taxon richness between reservoir-fragmented sites and unfragmented sites results, in part, from the fact that reservoir sites tended to be at lower elevations. Baetis melanonyx, B. rhodani, and E. venosus are subalpine 85

Table 4. Total taxon richness in streams (g) and hierarchical b-diversity measured as assemblage similarity (C) and turnover (bW) within streams. Total bW was calculated for all study streams. Stream g C bW Lake Schwellisee 27 0.58 1.50 Bianco 18 0.80 1.29 Minor 26 0.32 1.79 Cadagno 30 0.64 1.43 Upper Jöri 18 0.67 1.44 Lower Jöri 31 0.49 1.59 Mean 25.0 0.61 1.51

Reservoir Livigno 23 0.72 1.35 Marmorera 29 0.65 1.41 Ritom 31 0.65 1.41 Mean 27.7 0.67 1.39

Unfragmented Arosa 23 0.85 1.21 Julierpass 27 0.71 1.35 Moesa 33 0.68 1.38 Mean 27.7 0.75 1.31

TOTAL 69 3.70

species, occurring below 2000 m (Sartori and Landolt 1999), and therefore their distribution reflects altitudinal limits. Ephemeroptera are known to have a low tolerance to habitat alteration (Brittain 1982, Bauernfeind and Moog 2000) and one possibility for the reduced a in lake-fragmented sites is the result of habitat change or degradation. Most research on the influence of lakes on lotic habitat and communities has focused on the lake outlet itself rather than on downstream reaches (Richardson and Mackay 1991). In contrast, a large amount of research has demonstrated the significant effects that reservoirs can have on Ephemeroptera communities (Brittain and Saltveit 1989) and the persistence of such effects downstream (e.g., Ward 1976). Habitat changes result from alterations in flow pattern and discharge that lead to changes in current, substrate, water quality, and thermal regimes (Ward and Stanford 1979). In our study, alpine Ephemeroptera appeared to be more affected by lake-fragmentation, and generally were well represented in the 86 reservoir-fragmented sites. This may result from the hypolimnetic release from these reservoirs maintaining a thermal regime similar to unregulated alpine streams. It is doubtful that habitat alteration alone was responsible for the reduced Ephemeroptera richness in lake-fragmented sites, because study sites were located both above and below lakes. Thus, we further examined a second consequence of habitat fragmentation of streams: the loss of physical habitat connectivity. A disruption of flow continuity may affect the ability of organisms to disperse among fragments, particularly if dispersal ability is limited. Hershey et al. (1993) reported downstream drift and upstream flight distances of 2.1 km and 1.6 - 1.9 km, respectively, for Baetis along an arctic river. Other studies, using population genetic techniques, have indicated Ephemeroptera dispersal within streams can be quite limited (Schmidt et al. 1995, Bunn and Hughes 1997). In an earlier study we found large genetic differences between populations of B. alpinus separated by 100s of m of lentic habitat (Monaghan et al. 2001; see chapter 2). If this results from a behavior of halted flight when animals reach standing water bodies, dispersal may be limited among habitat fragments. We investigated the relative importance of habitat alteration and dispersal with our second hypotheses (see Nestedness, below). We also observed reduced Diptera in lake-fragmented sites compared to unfragmented sites when the Simuliidae were excluded. The observation that the Simuliidae data obscure pattern in lake-fragmented streams could be because the proximity to lake outlets is an important determinant of their distribution (Ciborowski and Adler 1990, Malmqvist et al. 1999). Alpine lakes rarely occur singly within a basin, but rather in clusters or in chains (e.g., paternoster lakes) because of the influence of glacial processes on their formation (Ritter et al. 1995). Thus, nearly all of the lake-fragmented sites had one or more lakes higher in the basin and this likely accounts for Simuliidae richness in our upstream as well as downstream study sites. The reduced Diptera primarily was the result of a decrease in the occurrence of Limoniidae, Psychodidae, and Tipulidae in lake-fragmented sites, although Hexatoma was more commonly observed in fragments. It is interesting that the first 3 taxa generally feed on large (> 1 mm) detritus and that Hexatoma primarily preys on living macroinvertebrates (Tachet et al. 2000). Such a pattern suggests a change in resource availability may be responsible for the change in Diptera assemblage composition, although we note that similar changes in trophic guild did not appear predominant in 87 other taxa, such as the Plecoptera and Trichoptera. Diptera dispersal has been examined primarily in the Simuliidae, with mixed (i.e., species- specific) results, suggesting that dispersal can be both limited or widespread (Snyder and Linton 1984, Crosskey 1990). Delettre and Morvan (2000) observed limited dispersal distances in Chironomidae, but it is not clear how important dispersal ability among the Diptera may be in determining whether they occur in fragmented streams.

Nestedness We used nestedness analysis to test our second hypothesis, that taxa assemblages in fragmented sites would be nested subsets of assemblages in unfragmented sites. Such a nested pattern would imply that extinction from fragments and a lack of recolonization is an important determinant of macroinvertebrate assemblage structure (Blake 1991, Newmark 1991), thereby implicating dispersal as an important process maintaining subpopulations within fragments (Stacey and Taper 1992, Hastings and Harrison 1994, Hanski 1998). The fact that the assemblages were nested was somewhat expected. Although nestedness is a common attribute of nearly all species assemblages (Wright et al. 1998), aquatic invertebrates have been seen as an exception to this pattern (Boecklen 1997, Malmqvist et al. 1999). Boecklen (1997) considered the fact that "aquatic invertebrates" includes a group of organisms so diverse (e.g., insects, mollusks, crustaceans) as to obscure pattern. In fact, Malmqvist and Hoffsten (2000) found nested distributions using separate analyses of Ephemeroptera, Plecoptera, Trichoptera, and Simuliidae. Unexpected was the observation that fragmented sites were not nested subsets of unfragmented sites. Instead, the uppermost 2 sites in the packed matrix (Fig. 2) were lake-fragmented sites, and unfragmented sites were distributed throughout the packed matrix. The observation that taxon assemblages in fragments were not nested subsets of those in unfragmented sites indicates a large amount of taxa turnover in the study sites. The estimates of b-diversity do show large amounts of taxa turnover: mean Sorrenson C was 0.65 across the whole study (ANOVA indicated no significant differences among stream types), a value that suggests paired sites within streams often shared 50 % or fewer taxa. The large number of "rare" taxa (46 out of 69) and the occurrence of 27 taxa found only in fragmented streams also provides evidence that species turnover was common and that assemblage structure was highly variable. 88

Together, high turnover and the lack of nestedness indicate that, for most taxa, dispersal is probably widespread and occurs among fragments. To state that the dispersal of most stream benthic macroinvertebrates is widespread (i.e. within and among streams and drainage basins) overlooks many species-specific examples of limited dispersal (Preziosi and Fairbairn 1992, Bunn and Hughes 1997, Monaghan et al. 2001, see also chapter 2), but it may be a reasonable general hypothesis to explain the data from the present study of 69 taxa. Studies of large-scale, particularly inter-basin dispersal are limited in number (e.g., Resh 1982), but the broad geographic range of many taxa also suggests widespread dispersal (Bilton et al. 2001). Many studies have concluded that stream macroinvertebrate assemblage structure is closely related to local-scale habitat characteristics (Douglas and Lake 1994, Richards et al. 1997, Downes et al. 1998, Minshall and Robinson 1998, Voelz and McArthur 2000, Burgherr and Ward 2001, Melo and Froehlich 2001, Sponseller et al. 2001). The overriding assumption is that most or all members of the regional species pool are able to disperse over large areas and that more local processes determine their presence or absence in particular habitats (habitat filters, sensu Poff 1997).

Implications for freshwater conservation The high degree of taxa turnover and the large number of rare taxa suggest that b-diversity is an important component of biodiversity in these alpine streams (see also Ward et al. 1999, Ward and Tockner 2001). Even the most taxa-rich site contained less than 50 % of all observed taxa, indicating that a large proportion of richness lies among different sites. Cluster analysis also suggested a large amount of taxon richness was found among the 10 different streams; multiple sites within streams were grouped together in 6 of 10 possible cases. Fragmentation appeared to have little effect, as 4 of the clustered streams were fragmented and 2 were not. Forty six of the taxa we observed in the study were "rare" in that they were observed in fewer than 6 (25 %) of the study sites. The prevalence of rare taxa may be a common attribute of stream benthic macroinvertebrate communities (Malmqvist and Hoffsten 2000, Robinson et al. 2000). This not only affects ecological patterns of species assemblages (Robinson et al. 2000), but also could result in stream biodiversity being particularly susceptible to habitat fragmentation. Rare taxa may be more sensitive than common taxa to environmental change (Gaston 1994, Cao et al. 1998). Lastly, the findings illustrate the need for more information regarding dispersal distance and ability of aquatic macroinvertebrates. There have 89 been many direct studies of dispersal in adult aquatic insects, but most have focussed on flight direction rather than distance (Bilton et al. 2001). There certainly are exceptions (e.g., Jackson and Fisher 1986, Collier and Smith 1998, Petersen et al. 1999), but knowledge of the dispersal abilities of a wide range of taxa may be necessary to understand how connectivity and dispersal interact with habitat alteration to influence taxon richness in fragmented streams.

Acknowledgements Many thanks to M. Hieber, P. Burgherr, C. Jolidon, M. Locke, D. Locke, F. Malard, F. Moesslacher, M. de la Puenta Nilsson, K. Rinke, S. Schalla, and M. Winder for assistance with field sampling. We thank R. Glatthaar for identification of the Simuliidae specimens and C. Jolidon for identification of the Blepharaceridae specimens. Author MTM thanks M. Sartori, V. Lubini, W. Graf, and J. Waringer for their instruction and help with macroinvertebrate taxonomy, and D. Arscott for an introduction to nestedness. Research was funded by Swiss National Science Foundation grants 31-50440.97 and 31-50444.97.

Literature Cited Allan, J. D., and A. S. Flecker. 1993. Biodiversity conservation in running waters. BioScience 43:32-43. Andren, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: a review. Oikos 71:355-366. Atmar, W., and B. D. Patterson. 1993. The measure of order and disorder in the distribution of species in fragmented habitat. Oecologia 96:373-382. Atmar, W., and B. D. Patterson. 1995. The nestedness temperature calculator: a visual basic program, including 294 presence-absence matrices. AICS Research, Inc. University Park, New Mexico, USA, and The Field Museum, Chicago, USA. Aubert, J. 1959. Plecoptera. Imprimerie La Concorde, Lausanne. Bauernfeind, E., and O. Moog. 2000. Mayflies (Insecta: Ephemeroptera) and the assessment of ecological integrity : a methodological approach. Hydrobiologia 423:71-83. Bender, D. J., T. A. Contreras, and L. Fahrig. 1998. Habitat loss and population decline: A meta- analysis of the patch size effect. Ecology 79:517-533. Bilton, D. T., J. R. Freeland, and B. Okamura. 2001. Dispersal in freshwater invertebrates. Annual Review of Ecology and Systematics 32:159-181. Blake, J. G. 1991. Nested subsets and the distribution of birds in isolated woodlots. Conservation Biology 5:58-66. Boecklen, W. J. 1997. Nestedness, biogeographic theory, and the design of nature reserves. Oecologia 112:123-142. 90

Breitenmoser-Würsten, C., and M. Sartori. 1995. Distribution, diversity, life cycle and growth of a mayfly community in a prealpine stream system (Insecta, Ephemeroptera). Hydrobiologia 308:85-101. Brittain, J. E. 1982. Biology of mayflies. Annual Review of Entomology 27:119-147. Brittain, J. E., and S. J. Saltveit. 1989. A review of the effects of river regulation on mayflies (Ephemeroptera). Regulated Rivers: Research and Management 3:191-204. Brown, J. H., and A. Kodrick-Brown. 1977. Turnover rates in insular biogeography: effect of immigration on extinction. Ecology 58:445-449. Bunn, S. E., and J. M. Hughes. 1997. Dispersal and recruitment in streams: evidence from genetic studies. Journal of the North American Benthological Society 16:338-346. Cao, Y., D. D. Williams, and N. E. Williams. 1998. How important are rare species in aquatic community ecology and bioassessment? Limnology and Oceanography 43:1403-1409. Ciborowski, J. J. H., and P. H. Adler. 1990. Ecological segregation of larval black flies (Diptera: Simuliidae) in northern Saskatchewan, Canada. Canadian Journal of Zoology 68:2113-2122. Collier, K. J., and B. J. Smith. 1998. Dispersal of adult caddisflies (Trichoptera) into forests alongside three New Zealand streams. Hydrobiologia 361:53-65. Crosskey, R. W. 1990. The natural history of blackflies. John Wiley and Sons, Chichester, U.K. Davies, K. F., and C. R. Margules. 1998. Effects of habitat fragmentation on carabid beetles: experimental evidence. Journal of Animal Ecology 67:460-471. Delettre, Y. R., and N. Morvan. 2000. Dispersal of adult aquatic Chironomidae (Diptera) in agricultural landscapes. Freshwater Biology. Jul 44:399-411. Douglas, M., and P. S. Lake. 1994. Species richness of stream stones: an investigation of the mechanisms generating the species-area relationship. Oikos 69:387-396. Downes, B. J., P. S. Lake, E. S. G. Schreiber, and A. Glaister. 1998. Habitat structure and regulation of local species diversity in a stony, upland stream. Ecological Monographs 68:237- 257. Dynesius, M., and C. Nilsson. 1994. Fragmentation and flow regulation of river systems in the northern third of the world. Science 266:753-762. Gaston, K. J. 1994. Rarity. Chapman and Hall, London. Hanski, I. 1998. Metapopulation dynamics. Nature 396:41-49. Hastings, A., and S. Harrison. 1994. Metapopulation dynamics and genetics. Annual Review of Ecology and Systematics 25:167-188. Hershey, A. E., J. Pastor, B. J. Peterson, and G. W. Kling. 1993. Stable isotopes resolve the drift paradox for Baetis mayflies in an Arctic River. Ecology 74:2315-2325. Jackson, J. K., and S. G. Fisher. 1986. Secondary production, emergence, and export of aquatic insects of a Sonoran desert stream. Ecology 67:629-638. Kattan, G. H., H. Alvarez-Lopez, and M. Giraldo. 1994. Forest fragmentation and bird extinctions: San Antonio eighty years later. Conservation Biology 8:138-146. Klein, B. C. 1989. Effects of forest fragmentation on dung and carrion beetle communities in central Amazonia. Ecology 70:1715-1725. Kling, G. W., G. W. Kipphut, M. M. Miller, and W. J. O' Brien. 2000. Integration of lakes and streams in a landscape perspective: The importance of material processing on spatial patterns and temporal coherence. Freshwater Biology 43:477-497. Lovejoy, T. E., A. B. Bierregaard, A. B. Rylands, J. R. Malcolm, C. E. Quintela, L. H. Harper, K. S. Brown, A. H. Powell, G. V. N. Powell, H. O. R. Schubart, and M. B. Hays. 1986. Edge and other effects of isolation on Amazon forest fragments. Pages 257-285 in M. E. Soulé (editor). Conservation biology: the science of scarcity and diversity. Sinauer Associates, Sunderland, Massachusetts. Magurran, A. E. 1988. Ecological diversity and its measurement. Croom Helm, London. Malmqvist, B., and P. O. Hoffsten. 2000. Macroinvertebrate taxonomic richness, community structure and nestedness in Swedish streams. Archiv für Hydrobiologie 150:29-54. Malmqvist, B., Y. X. Zhang, and P. H. Adler. 1999. Diversity, distribution and larval habitats of North Swedish blackflies (Diptera : Simuliidae). Freshwater Biology 42:301-314. 91

Margules, C. R., G. A. Milkovits, and G. T. Smith. 1994. Contrasting effects of habitat fragmentation on the scorpion Cercophonius squama and an amphipod. Ecology 75:2033- 2042. Melo, A. S., and C. G. Froehlich. 2001. Macroinvertebrates in neotropical streams: richness patterns along a catchment and assemblage structure between 2 seasons. Journal of the North American Benthological Society 20:1-16. Minshall, G. W., and C. T. Robinson. 1998. Macroinvertebrate community structure in relation to measures of habitat heterogeneity. Archiv für Hydrobiologie 141:129-151. Newmark, W. D. 1991. Tropical forest fragmentation and the local extinction of understory birds in the eastern Usambara Mountains, Tanzania. Conservation Biology 5:67-78. Patterson, B. D. 1987. The principle of nested subsets and its implications for biological conservation. Conservation Biology 1:323-334. Patterson, B. D., and W. Atmar. 1986. Nested subsets and the structure of insular mammalian faunas and archipeligos. Biological Journal of the Linnean Society 28:65-82. Petersen, I., J. H. Winterbottom, S. Orton, N. Friberg, A. G. Hildrew, D. C. Spiers, and W. S. C. Gurney. 1999. Emergence and lateral dispersal of adult Plecoptera and Trichoptera from Broadstone Stream, U.K. Freshwater Biology 42:401-416. Pimm, S. L., G. J. Russell, J. L. Gittleman, and T. M. Brooks. 1995. The future of biodiversity. Science 269:347-350. Poff, N. L. 1997. Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16:391- 409. Preziosi, R. F., and D. J. Fairbairn. 1992. Genetic population structure and levels of gene flow in the stream dwelling waterstrider, Aquarius ( = Gerris) remigis (Hemiptera: Gerridae). Evolution 46:430-444. Resh, V. H. 1982. Age structure alteration in a caddisfly population after habitat loss and recovery. Oikos 38. Richards, C., R. J. Haro, L. B. Johnson, and G. E. Host. 1997. Catchment and reach-scale properties as indicators of macroinvertebrate species traits. Freshwater Biology 37:219-230. Richardson, J. S., and R. J. Mackay. 1991. Lake outlets and the distribution of filter-feeders: an assessment of hypotheses. Oikos 62:370-380. Ritter, D. F., R. C. Kochel, and J. R. Miller. 1995. Process geomorphology, 3rd edition. W.C. Brown Publishers, Dubuque, Iowa. Robinson, C. T., G. W. Minshall, and T. V. Royer. 2000. Inter-annual patterns in macroinvertebrate communities of wilderness streams in Idaho, U.S.A. Hydrobiologia 421:187-198. Robinson, G. R., R. D. Holt, M. S. Gaines, S. P. Hamburg, M. L. Johnson, H. S. Fitch, and E. A. Martinko. 1992. Diverse and contrasting effects of habitat fragmentation. Science 257:524- 526. Rosenzweig, M. L. 1995. Species diversity in space and time. Cambridge University, Cambridge. Ryding, S. O., and W. Rast. 1989. The control of eutrophication of lakes and reservoirs. UNESCO, Paris. Sartori, M., and P. Landolt. 1999. Atlas de distribution des éphémères de Suisse (Insecta, Ephemeroptera). Centre Suisse de Cartographie de la Faune, Neuchâtel. Saunders, D. A., R. J. Hobbs, and C. R. Margules. 1991. Biological consequences of ecosystem fragmentation: a review. Conservation Biology 5:18-32. Schmidt, S. K., J. M. Hughes, and S. E. Bunn. 1995. Gene flow among conspecific populations of Baetis sp. (Ephemeroptera): adult flight and larval drift. Journal of the North American Benthological Society 14:147-157. Snyder, T. P., and M. C. Linton. 1984. Population structure in black flies: allozymic and morphological estimates for Prosimulium mixtum and P. fuscum (Diptera: Simuliidae). Evolution 38:942-956. Sponseller, R. A., E. F. Benfield, and H. M. Valett. 2001. Relationships between land use, spatial scale and stream macroinvertebrate communities. Freshwater Biology 46:1409-1424. 92

Stacey, P. B., and M. Taper. 1992. Environmental variation and the persistence of small populations. Ecological Applications 2:18-29. Tachet, H., P. Richoux, M. Bournaud, and P. Usseglio-Polatera. 2000. Invertébrés d'eau douce. CNRS Éditions, Paris. Taylor, C. M. 1997. Fish species richness and incidence patterns in isolated and connected stream pools: Effects of pool volume and spatial position. Oecologia 110:560-566. Taylor, C. M., and M. L. Warren, Jr. 2001. Dynamics in species composition of stream fish assemblages: environmental variability and nested subsets. Ecology 82:2320-2330. Vinson, M. R. 2001. Long-term dynamics of an invertebrate assemblage downstream from a large dam. Ecological Applications 11:711-730. Voelz, N. J., and J. V. McArthur. 2000. An exploration of factors influencing lotic insect species richness. Biodiversity and Conservation 9:1543-1570. Ward, J. H. 1963. Hierarchical grouping to optimize an objective function. Journal of the American Statistical Association 58:236. Ward, J. V. 1976. Comparative limnology of differentially regulated sections of a Colorado mountain river. Archiv für Hydrobiologie 78:319-342. Ward, J. V., and J. A. Stanford, editors. 1979. The ecology of regulated streams. Plenum, New York. Ward, J. V., and K. Tockner. 2001. Biodiversity: towards a unifying theme for river ecology. Freshwater Biology 46:807-819. Ward, J. V., K. Tockner, and F. Schiemer. 1999. Biodiverstity of floodplain river ecosystems: ecotones and connectivity. Regulated Rivers: Research and Management 15:125-139. Ward, J. V., and N. J. Voelz. 1988. Downstream effetcs of a large, deep-release, high mountain reservoir on lotic zoobenthos. Vereinigung für Theoretisch und Angewandte Limnologie 23:1174-1178. Waringer, J. A., and W. Graf. 1997. Atlas der österreichischen Köcherfliegenlarven: unter Einschluss der angrenzenden Gebiete. Facultas Universitätsverlag, Wien (Vienna). Whittaker, R. M. 1960. Vegetation of the Siskiyou Mountains, Oregon and California. Ecological Monographs 30:279-338. Wilcox, B. A., and D. D. Murphy. 1985. Conservation strategy: the effects of fragmentation on extinction. American Naturalist 125:879-887. Willis, T. V., and J. J. Magnuson. 2000. Patterns in fish species composition across the interface between streams and lakes. Canadian Journal of Fisheries and Aquatic Sciences 57:1042- 1052. Wotton, R. S. 1992. Feeding by blackfly larvae (Diptera: Simuliidae) forming dense aggregations at lake outlets. Freshwater Biology 27:139-149. Wright, D. H., B. D. Patterson, G. M. Mikkelson, A. Cutler, and W. Atmar. 1998. A comparative analysis of nested subset patterns of species composition. Oecologia 113:1-20. Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall, Englewood Cliffs, New Jersey. Zwick, P. 1992. Stream habitat fragmentation - a threat to biodiversity. Biodiversity and Conservation 1:80-97.

CHAPTER 6

SPATIAL PATTERNS OF EPHEMEROPTERA,

PLECOPTERA, AND TRICHOPTERA DIVERSITY IN

FRAGMENTED ALPINE STREAMS

with M. Hieber, C.T. Robinson, P. Spaak, and J.V. Ward, (in press) Verhandlungen Internationale Vereinigung für Theoretisch und Angewandte Limnologie. 28.

Abstract We examined how lakes within a drainage system may fragment flowing water habitats by assessing changes in taxon richness and composition of larval Ephemeroptera, Plecoptera, and Trichoptera in three fragmented alpine streams. Richness (a-diversity) and turnover (b-diversity) generally increased along longitudinal gradients of 200-300 m that encompassed sites above lakes, lake outlet sites, and sites farther downstream. Downstream sites were compositionally different from upstream sites, suggesting that effects of habitat fragmentation were evident beyond lake outlets. Site-specific differences among the three streams indicate that alteration of local habitat is a critical factor determining species response to stream habitat fragmentation.

Introduction Habitat fragmentation can affect species diversity by reducing total habitat area, increasing the amount of habitat “edge,” and by isolating fragments. The relative importance of each mechanism can be difficult to assess because of their confounding nature (Andren 1994). The fragmentation of streams by a lentic (lake) habitat that is chemically (Kling 94 et al. 2000) and biologically (Breitenmoser-Würsten and Sartori 1995) distinct, results in a loss of connectivity among “patches” of stream thus increasing the number of isolated fragments. The total loss of lotic habitat within a drainage usually is minimal and creation of “edge” (e.g. lake outlets) relatively local, as shown by studies of lake outlets and associated species assemblages (Richardson and Mackay 1991). At the scale of individual patches, fragmentation by lakes can reduce connectivity of discreet reaches (e.g. Monaghan et al. 2001; see chapter 2), which potentially can affect species richness. However, along a stream, increased spatial heterogeneity may lead to greater overall richness, manifested in species turnover among relatively short longitudinal distances. Thus, two components of species diversity relative to habitat fragmentation studies are species richness (-diversity) and species turnover (-diversity). Alpha-diversity () represents the number of species present at any given location in a stream, while beta-diversity () measures how rapidly species composition changes from one location to another (Magurran 1988). Taken together, these two measures indicate the spatial scale at which biodiversity is structured. For example, low at multiple sites in a stream but high between those sites indicates that spatial heterogeneity contributes to biodiversity. High at multiple sites and low between them indicates biodiversity lies within any given site. To examine how habitat fragmentation may affect species richness in patches and influence spatial heterogeneity in streams, we examined - and -diversity of Ephemeroptera, Plecoptera, and Trichoptera (EPT) assemblages in fragmented and nonfragmented alpine streams in Switzerland. Our first hypothesis was that at any given sampling site in fragmented streams would be reduced relative to sites in a nonfragmented stream, primarily because of reduced connectivity of fragments. Our second hypothesis was that increased in fragmented streams would result from the presence of a lake outlet and its characteristic species assemblage. This would result in higher total richness () for the whole stream, despite reduced at any given site.

Methods We conducted the study in 4 alpine streams in the Swiss Alps (Fig. 1, Table 1). Three streams were fragmented by lakes: Lago Bianco, Val 95

S witzerland

. R Upstream ine Rh C D . R Arosa n Inlet In B Lago Bianco Downstream T Val Minor A ic in Inlet o Outlet 1 R Outlet .

N J öri Downstream 50 km Outlet 2

Downstream Figure 1. Map of the four study streams, indicating sampling sites within each stream.

Minor, and Jörisee.The fourth stream, Arosa (9º39’2”E 46º46’21”), was nonfragmented. Bianco and Minor each contained a single lake (surface area 3.9 and 0.7 ha, respectively) and sampling sites were located at the inlet, immediately at the lake outlet, and about 150 m downstream of the outlet (Fig. 1). Jöri was fragmented by two lakes about 200 m apart. Thus, the outlet of the upper lake (5.8 ha) also served as the inlet of the lower lake (9.6 ha) and only the lower lake had a downstream sampling site (Fig. 1). The stream at Arosa was sampled at 2 locations approximately 300 m apart. There were no tributaries entering any of the study streams between sample sites. Larval zoobenthos were sampled semi-quantitatively at each site using a 100-µm mesh kick net. Sampling was conducted for 5 minutes each time to standardize sampling effort; all habitat types within a 50-m reach of stream were sampled. Habitat types typically encompassed riffles, runs, and pools, while substrata included moss, organic detritus, and sediment ranging from sand to large boulders. Samples were preserved in the field and sorted in the laboratory using a dissecting microscope (10X). Each site was sampled at least 3 times, once each in spring, summer, and autumn in either 1998 or 1999. Bianco, Minor, and Jöri outlet-2 also were sampled once in winter (the other sites were inaccessible in winter). For each site, species were placed into abundance categories of low (no more than 1-5 animals in any sample), moderate (6 to 96

Table 1. Physical characteristics of the 12 study sites. Median substrate size (n=50), mean (1 SD) width, and slope were measured along a 50-m reach at each site in Autumn. Stream Site Substrate Wetted Bankful Slope Altitude size width width (%) (m a.s.l.) (cm) (m) (m) Bianco inlet 9 6.0 (1.6) 12.5 (2.0) 2 2080 outlet 35 6.7 (2.5) 8.6 (1.5) 4 2076 downstream 20 4.2 (0.4) 9.7 (2.2) 4 2072

Minor inlet 17 8.6 (2.9) 10.3 (2.0) 7 2340 outlet 21 3.6 (0.4) 5.0 (0.6) 9 2336 downstream 11 5.0 (1.5) 6.4 (1.2) 7 2315

Jöri inlet 31 3.1 (1.1) 3.9 (1.5) 17 2525 outlet 1 37 6.8 (4.5) 8.8 (4.0) 4 2515 outlet 2 18 8.3 (2.4) 10.3 (2.0) 5 2489 downstream 17 4.7 (1.9) 5.3 (0.6) 8 2320

Arosa upstream 12 1.2 (0.5) 1.2 (0.6) 3 1940 downstream 14 1.5 (0.3) 1.5 (0.2) 7 1930

50 animals in any sample), and high (more than 50 animals). These categories are designated by one, two, or three asterices in Table 2. Species richness was determined for each sampling site (a) and for each stream (g) by combining samples from the different seasons. We examined species turnover (b) between sample site pairs in each stream using the Sorenson index C: C = 2j/(a+b), where j = the number of species found in both sites, a = the number of species in site a, and b = the number of species in site b. The Sorenson index has been found to be most appropriate using presence-absence data (Magurran 1988). Values of C range from 0 to 1, with a value of 1.0 indicating identical species composition and a value of 0.0 indicating that sites share no species in common. We calculated 3 values of C for each stream: turnover between the inlet and outlet (Cil-ol), between the outlet and downstream (Col-ds), and between the inlet and downstream (Cil-ds). Taken together, this allowed us to compare the lake outlet with upstream and downstream sites, and 97 allowed us to assess whether assemblages “recovered” to resemble those at the inlet after a given distance downstream of the outlet. To measure species turnover along each stream and among different streams, we calculated Whittaker’s b as bW = S/a-1, where S = the total number of species in the system (i.e. within a stream or over the whole study area) and a = mean species richness at each site (Magurran 1988). The lower the similarity of species assemblages, the higher the value of bW. To assess changes in chemical conditions along longitudinal gradients, 1-L water samples were collected at the same time as zoobenthos from each sample site and returned to the laboratory for analysis (see Tockner et al. 1997 for methods). Samples were collected in spring, summer, and autumn of 1998 and 1999 (n = 6); lake outlets were sampled more frequently in both summers (n = 10). Owing to the variance, likely the result of seasonal effects, we report these only as trends in mean values and made no statistical analysis. In autumn, we measured physical habitat characteristics: benthic substrate size (a-axis of the first particle encountered in 50 haphazardly chosen locations in a 50-m reach); wetted and bankfull width at 5 transects using a tape measure; and reach slope using a hand-held clinometer.

Results Water chemistry parameters showed a general pattern of elevated mean concentrations for dissolved and particulate N and P and particulate organic carbon (POC) at lake outlets (Fig. 2). Exceptions were at Minor, where particulate N and POC were highest at the upstream site and nitrate-

N (NO3) was highest downstream. Specific conductance was similar at all sites. In the nonfragmented stream, mean concentrations for all parameters were higher downstream, although mean specific conductance appeared to slightly decrease (Fig. 2). Median substrate size decreased from lake outlets to downstream sites but increased at the downstream site in the reference stream (Table 1). Downstream changes in wetted width, bankfull width, and slope varied among sites and no consistent patterns were evident. For example, lake outlets were wider than inlets and downstream sites at Jöri and Bianco but more narrow at Minor. None of 98

) Stream -1 225 20 Lake Outlet

150 ) -1 15 S cm

m 75 g ml 10 m

50 ( 4 5

25 NH

0 0 Spec. cond. ( 500 8

400 6

g/ml) 300 g/ml) m m 4 ( ( 3 200 4 2 PO NO 100

0 0

) 50 0.5 -1 ) g L 40 0.4 -1 m

30 g L 0.3 m 20 0.2

10 POC ( 0.1

Particulate N ( 0 0.0 inlet inlet inlet inlet inlet inlet outlet outlet outlet outlet outlet-1 outlet-2 outlet-1 outlet-2 upstream upstream downstream downstream downstream downstream downstream downstream downstream downstream Bianco Minor Jöri Arosa Bianco Minor Jöri Arosa

Figure 2. Chemical characteristics for each study site (mean + 1 SD) as measured over the course of one year. Table 2. Species composition at each site. Asterices indicate species was present in low (*), moderate (**), or high (***) abundance. Bianco Bianco Bianco Minor Minor Minor Jöri Jöri Jöri Jöri Arosa Arosa Inlet outlet Lower Inlet outlet Lower Inlet outlet 1 outlet 2 Lower Upper lower EPHEMEROPTERA Baetis alpinus ** * *** * *** ** *** ** * *** ** ** Ecdyonurus alpinus * Ecdyonurus pitceti ** *** Epeorus alpicola ** ** * Rhithrogena endenensis *** Rhithrogena loyolaea ** ** *** ** ** ** ** ** Rhithrogena nivata ** PLECOPTERA Dictyogenus alpinum ** ** ** ** ** ** ** ** Isoperla * Leuctra sp. ** ** ** ** * * ** ** ** Nemoura mortoni ** ** * ** Nemoura sp. *** Perlodes intricatus ** * ** * * Protonemura sp. * * ** * ** ** * **

Rhabdiopteryx alpina ** ** 99 Siphonoperla montana ** TRICHOPTERA Acrophylax zerberus *** Anitella obs./Melamophylax mel. * Consorophylax sp. ** ** * * Drusus biguttatus ** Drusus destitutus *** Drusus discolor *** *** ****** Drusus mixtus ** ** * ** Drusus monticola ** Halesus rubricollis * Rhyacophila glareosa ** * Rhyacophila intermedia ** ** * ** Rhyacophila vulgaris ** ** ** *

α diversity 9 6 10 1 4 14 7 7 8 18 11 12 100 the physical or chemical characteristics were correlated with the values of a-diversity reported below (Pearson correlation, all r-values < 0.6). A total of 28 Ephemeroptera, Plecoptera, and Trichoptera taxa (hereafter referred to as species) were collected among the 12 study sites (Table 2). The total number of species may in fact be higher because of taxonomic uncertainty in the Plecoptera genera Leuctra, Nemoura, and Protonemura (Aubert 1959,V. Lubini, pers. comm.) and in the Trichoptera genus Rhyacophila (Waringer and Graf 1997). Alpha (a) diversity ranged from 1 to 18 species at individual sites, with the highest richness found at downstream sites in two fragmented streams (Table 2). Longitudinal changes in a varied among the streams, with lowest a at the outlet at Bianco and at the inlet at Minor; however, the most downstream sampling site always had the highest a-value (Table 2). Similarity (C) of site pairs within each stream, used to evaluate the uniqueness of lake outlet species composition and “recovery” of downstream sites, ranged from 0.13 to 0.86 (Table 3). In Minor and Jöri, the inlet and outlet sites were most similar, followed by the outlet and downstream (Cil-ol > Col-ds); lowest similarity was between the inlet and downstream (Table 3). In contrast, highest similarity was between the inlet and downstream in Bianco, followed by inlet - outlet and outlet - downstream (Table 3). The two sites in the nonfragmented stream (Arosa) had the highest similarity in any stream (Table 3). Evaluating the role of whole-stream spatial heterogeneity, species turnover (bW) was greater in the three fragmented streams (1.91 to 2.81) than the reference stream (1.2; Table 3). Gamma (g) diversity also was higher in fragmented streams (14-

22) than in the reference stream (13; Table 3). bW among all sampling sites in the study (3.54 was higher than bW within any single stream (maximum of 2.81).

Discussion Habitat fragmentation did not lead to EPT-depauperate patches of stream habitat relative to a nonfragmented stream, and in fact the two highest a values occurred at fragmented sites. However, fragmentation did result in a shift in the spatial pattern of species richness in the fragmented streams. In the nonfragmented stream, Arosa, most of the species present were found at both sites and therefore species turnover (b- diversity) was low. We can therefore conclude that the majority of species in this stream should be found at any given site, at least within the 100s of 101

Table 3. Species similarity between site pairs (C) and total species richness (g) and turnover (b-diversity) in each stream. Similarity between site pairs within each stream is expressed as a Sorenson index for inlet-outlet (Cil-ol; 2 lakes occurred at the Jöri site), outlet-downstream (Col-ds), and inlet- downstream (Cil-ds) comparisons. Whittaker’s b (MAGURRAN 1988) was calculated for each stream and for the entire study.

Study site Cil-ol Col-ds Cil-ds g bW Bianco 0.53 0.38 0.74 14 1.91 Minor 0.40 0.33 0.13 15 2.81 Jöri 0.57/0.38 0.30 0.31 22 2.40 Arosa - - 0.86 13 1.20 total 28 3.54

m of similar reach. In contrast, there was more species turnover among sites within each fragmented stream. This turnover resulted in higher total species richness (g) in fragmented streams, despite the fact that two to three sites in each fragmented stream had lower a than the reference stream. Thus, high b-diversity resulted in increased species richness in fragmented streams when a larger spatial scale was considered (although only 100s of m of stream) that incorporated the spatial heterogeneity of fragments. This finding supports the contention that turnover at multiple spatial scales (b-diversity) is an important measure of biodiversity in aquatic systems (Ward et al. 1999). Additionally, whether richness increases or decreases from environmental heterogeneity depends in part on the spatial scale considered. Somewhat surprisingly, the high species turnover in fragmented streams did not result from lake outlets harboring different EPT species. Only 2 of 28 species identified (Nemoura sp. and Anitella/ Melamophylax) were found solely in lake outlets. This is, in part, because many lotic or lentic species typical of lake outlets (e.g., Simuliidae) were not investigated in the present study, although our lake outlets appeared to be chemically richer and perhaps distinct from the stream segments. We also observed little of the expected “recovery” of downstream sites because the inlet and downstream sites were most similar only at Bianco. This further supports our findings that lake outlets were not taxonomically distinct and indicates that other habitat characteristics of downstream sites may be responsible for the greater observed EPT richness or that the taxonomic 102

‘distinctness’ of lake outlet communities declines with increasing elevation. We did not find any explanatory relationships between physical characteristics and a-diversity, although a lack of strong relationships was not surprising given the small number of sampling sites. Species ability to disperse among disconnected patches appeared to have minimal influence on richness, as inlets and outlets generally were highly similar. Thus, local habitat characteristics probably are primary determinants of EPT response to habitat fragmentation. High diversity downstream of lake outlets has been observed in other, lower elevation streams. The high diversity has been attributed to increased transported organic matter (Robinson and Minshall 1990) and to stream size (Malmqvist 1999); parameters that were similar among our study sites. In summary, we examined how lakes within a drainage system may fragment flowing water habitats by assessing changes in species richness and composition of larval Ephemeroptera, Plecoptera, and Trichoptera (EPT) in four different alpine streams. Species richness (a-diversity) and turnover (b-diversity) generally increased along longitudinal gradients of 200-300 m that encompassed lake inlets, lake outlets, and sites 100-300 m downstream. Eight of ten fragmented sites had reduced EPT a-richness, although g-richness was greater in fragmented streams than in the nonfragmented stream. This was due to high EPT richness at downstream sites rather than to these high elevation lake outlets harboring distinct EPT assemblages.

Acknowledgements The authors thank P. Burgherr, C. Jolidon, F. Malard, F. Moesslacher, M. de la Puenta Nilsson, K. Rinke, S. Schalla, and M. Winder for assistance with field sampling. M. Sartori, W. Graf, J. Waringer, and V. Lubini assisted with taxonomy. Research was funded by Swiss National Science Foundation grants No. 31-50440.97 and 31-50444.97.

Literature Cited

Andren, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: a review. Oikos 71:355-366. Aubert, J. 1959. Plecoptera. Imprimerie La Concorde, Lausanne. Breitenmoser-Würsten, C., and M. Sartori. 1995. Distribution, diversity, life cycle and growth of a mayfly community in a prealpine stream system (Insecta, Ephemeroptera). Hydrobiologia 308:85- 101. 103

Kling, G. W., G. W. Kipphut, M. M. Miller, and W. J. O' Brien. 2000. Integration of lakes and streams in a landscape perspective: The importance of material processing on spatial patterns and temporal coherence. Freshwater Biology 43:477-497. Magurran, A. E. 1988. Ecological diversity and its measurement. Croom Helm, London. Malmqvist, B. 1999. Lotic stoneflies (Plecoptera) in northern Sweden: patterns in species richness and assemblage structure. Pages 63-72 in N. Friberg and J. D. Carl (editors). Biodiversity in benthic ecology: Proceedings from the Nordic Benthological Meeting in Silkeborg. National Environmental Research Institute, Denmark. Monaghan, M. T., P. Spaak, C. T. Robinson, and J. V. Ward. 2001. Genetic differentiation of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in fragmented alpine streams. Heredity 86:395-403. Richardson, J. S., and R. J. Mackay. 1991. Lake outlets and the distribution of filter-feeders: an assessment of hypotheses. Oikos 62:370-380. Robinson, C. T., and G. W. Minshall. 1990. Longitudinal development of macroinvertebrate communities below oligotrophic lake outlets. Great Basin Naturalist 50:303-311. Tockner, K., F. Malard, P. Burgherr, C. T. Robinson, U. Uehlinger, R. Zah, and J. V. Ward. 1997. Physico-chemical characterization of channel types in a glacial floodplain ecosystem (Val Roseg, Switzerland). Archiv für Hydrobiologie 140:433-463. Ward, J. V., K. Tockner, and F. Schiemer. 1999. Biodiverstity of floodplain river ecosystems: ecotones and connectivity. Regulated Rivers: Research and Management 15:125-139. Waringer, J. A., and W. Graf. 1997. Atlas der österreichischen Köcherfliegenlarven: unter Einschluss der angrenzenden Gebiete. Facultas Universitätsverlag, Wien (Vienna). 104 105

CHAPTER 7

THESIS CONCLUSIONS WITH RECOMMENDATIONS

FOR FUTURE RESEARCH

'…the wide ranging power of fresh-water productions can, I think, in most cases be explained by their having become fitted, in a manner highly useful to them, for short and frequent migrations from pond to pond, or from stream to stream, within their own countries; and liability to wide dispersal would follow from this capacity as an almost necessary consequence.' C. Darwin (1859) The Origin of Species (cited from the Modern Library Paperback Edition, 1998, New York, p. 522).

Regional and local effects of habitat alteration One of the goals of this thesis was to better understand the processes leading to species loss in fragmented habitats. In one sense, the thesis examined processes at 2 different spatial scales. At the regional scale was the question of how well species are able to disperse among patches. This was addressed with the population genetics data, where inferred levels of gene flow allow one to estimate how likely a species would be able to recolonize a fragment from which it has become extinct. At the local scale was the question of how assemblages differed among different habitat fragments. This was addressed with species assemblage data, which provide estimates of taxon turnover among fragments. The conclusion I draw from the data is that although dispersal ability (vagility) varies among species, most stream macroinvertebrates exhibit high vagility. The presence of a species in scattered fragments indicates that it was able to disperse there (unless it is endemic to the fragment). The large number of species found only in a few widely separated fragments suggests they are capable of dispersal and that the suitability of local habitat conditions 106 probably is the primary reason a species is present or absent in a habitat fragment. The vast majority of studies examining colonization processes of stream benthic macroinvertebrates make this assumption (see Mackay 1992). It is an explicit part of Poff's (1997) formulation of multi-scale habitat filters that macroinvertebrate species must "pass through" to be present at a given locale. I conclude that the statement "Given long time periods, all species are assumed capable of dispersing to all locales in the region" (Poff 1997) is an appropriate assumption. My conclusion is based on the data in this thesis and from a large amount of empirical evidence on colonization (Sheldon 1984, Mackay 1992), dispersal (Bilton et al. 2001), and habitat fidelity (e.g., Minshall 1984, Douglas and Lake 1994, Richards et al. 1997, Stazner 1997, Voelz and McArthur 2000, Burgherr and Ward 2001). The important "filters" that constrain species occurrence and richness include (but are not limited to) flow and temperature regimes, disturbance, sediment structure, food resources, water velocity, water chemistry, competition, and predation. An important point of the assumption is the qualifier "Given long time periods..." It is clear from some studies of colonization that aerial dispersers may take time to colonize defaunated areas from distant habitats (Mackay 1992). Minshall et al. (1983) describe how this process can take weeks to more than a year, and that distance to a source of colonists is an important determinant of the rate of recolonization. Hierarchy theory suggests that two processes must occur at relatively similar rates if they are to interact to create pattern (O'Neill et al. 1986, O'Neill and King 1998). I conclude that the dispersal of benthic macroinvertebrates occurs among stream fragments often enough that it is not a critical determinant of species composition in fragments. This is because local extinction from habitat alteration, while it may occur rapidly during the process of fragmentation, is overall a much slower process. When these two processes, dispersal and local extinction, operate at similar rates, their interaction may be much more important for determining distribution patterns. Such an interaction can result in metapopulation structure (sensu Hanski 1998). Unfortunately, there is a limited amount of information regarding temporal stability of macroinvertebrate assemblages, with much of it pertaining to relatively pristine streams (see Robinson et al. 2000) and streams at lower elevations. These conclusions lead to several questions that could be addressed with future research on alpine streams. A considerable amount of research 107 has been conducted on alpine streams since the review by Ward (1994), predominantly focussing on how different habitat types support different macroinvertebrate species assemblages and how seasonal changes in habitat types influence assemblage structure (e.g., papers in Brittain and Milner 2001). Important question arising from the thesis conclusions stated above are: How persistent are macroinvertebrate assemblages over time (scales of years) in alpine streams? Is local population extinction common in the alpine environment, which often is considered to be harsh? Related questions concern dispersal: what is the spatial extent of dispersal and how important is dispersal relative to habitat alteration in determining assemblage structure? Most study and speculation about the importance of dispersal to alpine stream assemblages have considered the effects of dispersal on the available species pools in biogeographical regions (reviewed by Ward 1994, see Milner et al. 2001). The thesis raised the question of whether dispersal may be more locally limiting. Another question asks what generalizations can be made regarding dispersal ability and how much is truly species-specific? Finally, the conclusion that species turnover is high and that it probably results from habitat turnover suggests spatial and temporal habitat heterogeneity is an important driver of biodiversity in streams (Ward and Tockner 2001). An important consideration is whether the rate of change in habitat characteristics occurs at a rate similar to or slower to the rate at which macroinvertebrates can respond by colonization or by life cycle adaptations. Another aspect of this thesis addressed differences between natural and anthropogenic fragmentation in streams. The results suggest that lakes and reservoirs affect different aspects of total biodiversity. They had similar effects on taxon richness but the effects on genetic differentiation were more pronounced in lakes. This suggests that the reduction of taxon diversity is a more immediate threat of reservoir construction. An important management and conservation question arises from this conclusion: In reservoir-fragmented systems, how can managers mimic the changes in habitat (e.g., the flow regime) at the appropriate rates so that natural biodiversity is preserved?

Future research on habitat fragmentation using population genetics Population genetics presents a means to study not only gene flow among habitat fragments, but also the processes of population dynamics 108 that are otherwise difficult to study using traditional ecological methodologies. Adult oviposition and population recruitment are two examples of population dynamics that have important consequences for the distribution and abundance of benthic macroinvertebrates, including their persistence in disturbed and fragmented habitats. In a broader perspective, separating historical patterns from present-day processes is an area of population genetics that requires much work regardless of the organism (Larson et al. 1984, Boileau et al. 1992, Barber 1999, Freeland et al. 2000, Bilton et al. 2001, chapters 2,3,4 of this thesis). If we are to use estimates of gene flow to study the dispersal of organisms, it is critical that we understand how much of the genetic signature is due to (1) historical levels of gene flow, (2) ongoing population dynamics, and (3) present-day gene flow. Understanding the rates of molecular evolution, e.g., how quickly certain markers may mutate or how much resolution is provided by different markers, is critical for understanding the temporal resolution of genetic markers. For example, Berry and Kreitman (1993) observed 113 RFLP haplotypes (genotypes designated by the presence and absence of DNA restriction sites) of a gene that expressed only 2 allozyme alleles in 1533 individuals, indicating that fewer than 1 out of 50 changes to the DNA sequence were detected with allozyme electrophoresis. Freeland et al. (2000) suggested the rapid mutation rate of microsatellites showed more recent patterns than 16S mtDNA. One interesting possibility for studying the effects of ongoing population dynamics on population genetic signatures involves the use of microsatellites. Patchy recruitment (i.e., the founding of populations by only a few females each generation) effectively represents a genetic bottleneck each generation. Previous tests for bottlenecks required extensive long-term data and large numbers of allozyme loci (Waples 1989). A lack of long-term data can be overcome using microsatellites rather than allozymes. Like allozymes, microsatellites provide information about the number of alleles (k). In addition, microsatellites have information about the range in allele size (r). Rather than examining k over time (loss of alleles from bottlenecks), microsatellite data allow for the examination of the ratio of k/r, which is reduced more quickly than k (Garza and Williamson 2001). Directly examining sequence variation is the most sensitive method for examining genetic variation, and the use of sequencing methods for population-level studies is increasingly common (McGlashan and Hughes 109

2000, Myers et al. 2000). The ability to sequence much larger and more functionally diverse regions of the genome may be the most common method in the future. Another possibility is the growing interest in single nucleotide polymorphisms (SNPs). The human genome project found 1.4 million such polymorphisms, 96 % of which are found in non-coding regions (Sachidanandam et al. 2001). Their utility in population genetic studies of neutral variation have yet to be fully evaluated (pers. obs., Conservation Genetics Meeting, September 2001, Lausanne). Ultimately, it must always be remembered that studies of the spatial distribution of molecular markers are used to estimate gene flow. This is only one means of studying the important process of dispersal. Ecological investigations continue to require field studies and direct measures of dispersal to compliment genetic studies. An additional consideration is that the study design itself can help to disentangle the potentially confounding effects of historical gene flow, ongoing population dynamics, and present-day gene flow. Hierarchy theory, discussed earlier, suggests that processes that occur on short and long temporal scales (i.e., have comparatively rapid and slow rates), should manifest in patterns at comparatively small and large spatial scales (O'Neill et al. 1986). Thus, by examining multiple spatial scales, we may be able to examine multiple temporal scales as well. Specifically, we may be able to examine important processes that occur too slowly to investigate over the time scale of a doctoral thesis.

Literature Cited

Barber, P. H. 1999. Patterns of gene flow and population genetic structure in the canyon treefrog, Hyla arenicolor (Cope). Molecular Ecology 8:563-576. Berry, A., and M. Kreitman. 1993. Molecular analysis of an allozyme cline: alcohol dehydrogenase in Drosophila melanogaster on the east coast of North America. Genetics 134:869-893. Bilton, D. T., J. R. Freeland, and B. Okamura. 2001. Dispersal in freshwater invertebrates. Annual Review of Ecology and Systematics 32:159-181. Boileau, M. G., P. D. N. Hebert, and S. S. Schwartz. 1992. Non-equilibrium gene frequency divergence: persistent founder effects in natural populations. Journal of Evolutionary Biology 5:25-39. Brittain, J. E., and A. M. Milner. 2001. Ecology of glacier-fed rivers: current status and concepts. Freshwater Biology 46:1571-1578. Burgherr, P., and J. V. Ward. 2001. Longitudinal and seasonal distribution patterns of the benthic fauna in a glacier-fed stream. Freshwater Biology 46:1705-1721. Darwin, C. 1859. The origin of species. Modern Library Paperbacks edition (1998), New York. Douglas, M., and P. S. Lake. 1994. Species richness of stream stones: an investigation of the mechanisms generating the species-area relationship. Oikos 69:387-396. Freeland, J. R., C. Romualdi, and B. Okamura. 2000. Gene flow and genetic diversity: a comparison of freshwater bryozoan populations in Europe and North America. Heredity 85:498-508. Garza, J. C., and E. G. Williamson. 2001. Detection of reduction in population size using data from microsatellite loci. Molecular Ecology 10:305-318. Hanski, I. 1998. Metapopulation dynamics. Nature 396:41-49. 110

Larson, A., D. B. Wake, and K. P. Yanev. 1984. Measuring gene flow among populations having high levels of genetic fragmentation. Genetics 106:293-308. Mackay, R. J. 1992. Colonization by lotic macroinvertebrates: a review of processes and patterns. Canadian Journal of Fisheries and Aquatic Sciences 49:617-628. McGlashan, D. J., and J. M. Hughes. 2000. Reconciling patterns of genetic variation with stream structure, earth history and biology in the Australian freshwater fish Craterocephalus stercusmuscarum (Atherinidae). Molecular Ecology 9:1737-1751. Milner, A. M., J. E. Brittain, E. Castellas, and G. E. Petts. 2001. Trends of macroinvertebrate community structure in glacier-fed rivers in relation to environmental conditions: a synthesis. Freshwater Biology 46:1833-1847. Minshall, G. W. 1984. Aquatic insect-substratum relationships. Pages 358-400 in V. H. Resh and D. M. Rosenberg (editors). The ecology of aquatic insects. Praeger, New York. Minshall, G. W., D. A. Andrews, and C. Y. Manuel-Faler. 1983. Application of island biogeographic theory to streams: macroinvertebrate recolonization of the Teton River, Idaho. Pages 279-297 in J. R. Barnes and G. W. Minshall (editors). Stream ecology: application and testing of general ecological theory. Plenum, New York. Myers, M. J., C. P. Meyer, and V. H. Resh. 2000. Neritid and thiarid gastropods from French Polynesian streams: how reproduction (sexual, parthenogenetic) and dispersal (active, passive) affect population structure. Freshwater Biology 44:535-545. O'Neill, R. V., D. L. DeAngelis, J. B. Waide, and T. F. H. Allen. 1986. A hierarchical concept of ecosystems. Princeton University Press, Princeton. NJ. O'Neill, R. V., and A. W. King. 1998. Homage to St. Michael; or, why are there so many books on scale? Pages 3-15 in D. L. Peterson and V. T. Parker (editors). Ecological scale: theory and applications. Columbia University Press, New York. Poff, N. L. 1997. Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16:391-409. Richards, C., R. J. Haro, L. B. Johnson, and G. E. Host. 1997. Catchment and reach-scale properties as indicators of macroinvertebrate species traits. Freshwater Biology 37:219-230. Robinson, C. T., G. W. Minshall, and T. V. Royer. 2000. Inter-annual patterns in macroinvertebrate communities of wilderness streams in Idaho, U.S.A. Hydrobiologia 421:187-198. Sachidanandam, R., D. Weissman, S. C. Schmidt, J. M. Kakol, L. D. Stein, G. Marth, S. Sherry, J. C. Mullikin, B. J. Mortimore, D. L. Willey, S. E. Hunt, C. G. Cole, P. C. Coggill, C. M. Rice, Z. M. Ning, J. Rogers, D. R. Bentley, P. Y. Kwok, E. R. Mardis, R. T. Yeh, B. Schultz, L. Cook, R. Davenport, M. Dante, L. Fulton, L. Hillier, R. H. Waterston, J. D. McPherson, B. Gilman, S. Schaffner, W. J. Van Etten, D. Reich, J. Higgins, M. J. Daly, B. Blumenstiel, J. Baldwin, N. S. Stange-Thomann, M. C. Zody, L. Linton, E. S. Lander, and D. Altshuler. 2001. A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature 409:928-933. Sheldon, A. L. 1984. Colonization dynamics of aquatic insects. Pages 401-429 in V. H. Resh and D. M. Rosenberg (editors). The ecology of aquatic insects. Praeger, New York. Stazner, B. 1997. Complexity of theoretical concepts in ecology and predictive power: patterns observed in stream organisms. Pages 211-218 in P. Landolt and M. Sartori (editors). Ephemeroptera and Plecoptera: Biology-Ecology-Systematics. MTL, Fribourg. Voelz, N. J., and J. V. McArthur. 2000. An exploration of factors influencing lotic insect species richness. Biodiversity and Conservation 9:1543-1570. Waples, R. S. 1989. A generalized approach for estimating effective population size from temporal changes in allele frequency. Genetics 121:379-391. Ward, J. V. 1994. Ecology of alpine streams. Freshwater Biology 32:277-294. Ward, J. V., and K. Tockner. 2001. Biodiversity: towards a unifying theme for river ecology. Freshwater Biology 46:807-819. 111

Curriculum Vitae

Michael T. Monaghan Deptartment of Limnology, EAWAG, Postfach 611, 8600 Dübendorf, Switzerland ph +41.1.823.51.78, fx +41.1.823.53.15, [email protected] Winterthurerstrasse 394, 8051 Zürich, Switzerland ph +41.1.321.01.87

Citizenship U.S.A., born 13.10.1971 Columbus, Ohio.

Education 1998-2002 Dr. Nat. Sci., ETH Zürich (Swiss Federal Institute of Technology) 1995-1998 M.S. Biology, Idaho State University, USA 1990-1995 B.S. Environmental Science, The Ohio State University, USA

Professional positions 1998-present PhD Student Researcher, Department of Limnology, Swiss Federal Institute for Environmental Science and Technology (EAWAG) 2000-2001 Visiting PhD Student, Evolutionary Genetics Laboratory, Department of Zoology, Otago University, New Zealand (Nov 2000-Feb 2001) 1995-1998 Graduate Research Assistant, Stream Ecology Center, Idaho State University 1994 Laboratory Assistant, F. T. Stone Laboratory, The Ohio State University Biological Field Station, Put-In-Bay, Ohio

Research experience PhD Thesis: "Habitat fragmentation of alpine streams: implications for genetic structure and species diversity of aquatic insects." MS Thesis: "Mechanisms determining filter-feeder distributions in lake-outlet streams." Population genetics: aquatic insects and crustaceans (gene flow, hybridization): allozyme electrophoresis, PCR, AFLP, microsatellites, RFLP (Switzerland, New Zealand). Stream ecology: field sampling of remote wilderness streams and large rivers, including extended backpacking trips, back-country skiing, helicopters, boats, small aircraft (Idaho, Alaska, Switzerland, Italy); leaf litter decomposition (Idaho); stream ecosystem metabolism using open-system and chamber methods (Idaho); transported organic matter sampling (Idaho, Switzerland); field release and laboratory analysis of radiolabeled (14C) organic material, conservative tracer injection and analysis, transient storage modeling (Idaho). Taxonomy: Alpine Ephemeroptera, Plecoptera, Trichoptera (Switzerland).

Teaching experience Guest Lecturer: Ecological Genetics of Aquatic Organisms (ETH Zürich). Population substructure, F-statistics, Gene flow. Laboratory instructor: Ecological Genetics of Aquatic Organisms (ETH Zürich) Laboratory instructor: Limnology Practical (ETH Zürich). 112

Field/Laboratory assistant: Ecology and Systematics of Macroinvertebrates (ETH Zürich). Diploma student supervision: Tamara Barthelmes (Dipl. Biology, ETH Zürich) "Räumliche und zeitliche Muster in der genetischen Populationsstruktur von Acrophylax zerberus (Trichoptera) in alpinen Schwemmebenen."

Awards and funding North American Benthological Society Endowment Committee, Graduate Student Research Award 2002. US$ 500 for the project, "Using microsatellite markers to investigate recruitment bottlenecks in Baetis alpinus." Idaho State University, Graduate Studies Research and Scholarship Committee, research and travel awards 1996, 1997 US $1000. Petersen Fund travel award, North American Benthological Society, for attendance at the 2002 Annual Meeting, Pittsburg, USA. Teaching Assistantship, ETH Zürich, 5of 7semesters, 1999-2002. Idaho State University Tuition and Fee Waiver Scholarship 1996-98. Dean's List, College of Agriculture, School of Natural Resources, The Ohio State University 1994-95 Oak Park Conservation Club Scholarship 1994 Mars. G. Fontana Fund Academic Scholarship 1992-93 Procter and Gamble Fund Scholarship 1990-95

Professional experience Manuscript referee: Journal of the North American Benthological Society, Freshwater Biology, Archiv für Hydrobiologie, Aquatic Sciences, Canadian Journal of Fisheries and Aquatic Sciences, Hydrobiologia. Grant proposal reviewer: National Environment Research Council (NERC) U.K. Session Chair, Sixth Annual Meeting of European PhD Students in Evolutionary Biology. Leuven, Belgium, February 2000. Special Session Co-organizer, North American Benthological Society Meeting, Pittsburg, USA, May 2002.

Professional memberships European Society for Evolutionary Biology North American Benthological Society American Society of Limnology and Oceanography International Society of Theoretical and Applied Limnology (SIL)

Publications Monaghan, M.T., P. Spaak, C.T. Robinson, and J.V. Ward. (2002). Population genetic structure of 3 Alpine stream insects: influences of gene flow, demographics, and habitat fragmentation. Journal of the North American Benthological Society 21:114– 131. Monaghan, M.T., P. Spaak, C.T. Robinson, and J.V. Ward. (2001). Genetic differentiation of Baetis alpinus Pictet (Ephemeroptera: Baetidae) in fragmented alpine streams. Heredity 86:395-403. Monaghan, M.T., S.A. Thomas, G.W. Minshall, J.D. Newbold, and C.E. Cushing (2001). The influence of filter-feeding benthic macroinvertebrates on the transport 113

and deposition of particulate organic matter and diatoms in two streams. Limnology and Oceanography 46:1091-1099. Monaghan, M.T., M. Hieber, C.T. Robinson, P. Spaak, and J.V. Ward. (In Press). Spatial patterns of Ephemeroptera, Plectopera, and Trichoptera species diversity in fragmented alpine streams. Verh. Internat. Verein. Lim. 28. Winder, M., M.T. Monaghan, and P. Spaak. (2001). Have human impacts changed alpine zooplankton diversity over the past 100 years? Arctic, Antarctic, and Alpine Research 33:467-475. Thomas, S.A., J.D. Newbold, M.T. Monaghan, G.W. Minshall, T. Georgian, and C.E. Cushing. (2001). The influence of particle size on the deposition of seston in streams. Limnology and Oceanography 46:1415-1424 Minshall, G.W., S.A. Thomas, J.D. Newbold, M.T. Monaghan and C.E. Cushing. (2000). Physical influences on organic particle transport and deposition in streams. Journal of the North American Benthological Society 19:1-16. Royer, T.V., M.T. Monaghan, and G.W. Minshall. (1999). Processing of native and exotic leaf litter in two Idaho (USA) streams. Hydrobiologia 400:123-128.

Presentations Gene flow in fragmented streams: influences of dispersal, recruitment, and population history on the genetic structure of Alpine stream insects. Troisième Cycle Romand en Sciences Biologiques: Conservation Genetics Meeting, Lausanne, Switzerland, September 2001. Spatial patterns of Ephemeroptera, Plectopera, and Trichoptera species diversity in fragmented alpine streams. International Society of Theoretical and Applied Limnology (SIL), Melbourne, Australia, February 2001. Genetic variation of benthic macroinvertebrates in fragmented alpine streams: Baetis alpinus. North American Benthological Society, Colorado, USA, June 2000 Does Holocene glacial advance and retreat influence the genetic structure of benthic insect populations? International Symposium on High Mountain Lakes and Streams: Indicators of a Changing World, Innsbruck, Austria, September 2000 Genetic structure of two insect taxa in fragmented stream ecosystems. "Aquatic Habitats as Ecological Islands Symposium" Joint Meeting of the British Ecological Society and Freshwater and Marine Biological Associations, Plymouth, UK, September 2000. Genetic diversity of Baetis alpinus (Ephemeroptera) in a naturally fragmented alpine stream. Symposium for European Freshwater Sciences, Antwerp, Belgium, June 1999. Downstream transport of particulate and dissolved organic matter in a lake outlet stream. North American Benthological Society, Prince Edward Island, Canada, June 1998. Breakdown of leaf litter from introduced and native riparian trees. North American Benthological Society, Texas, USA, June 1997.

16 additional co-authored presentations, 1997-2002. 114 115

ACKNOWLEDGEMENTS

The list is a long one. It reflects how much help and support I have received from so many people (and I am sure I have missed some), and it also represents the many interactions that make science such an enjoyable and rewarding pursuit. First I thank the three from whom I have learned the most. I thank Piet Spaak, Chris Robinson, and Prof. J.V. Ward for their encouragement, guidance, patience, and support. I thank them for the opportunity to do the work, for allowing me to let my own interests guide the focus of the research, and for their faith and support for my work in New Zealand. I thank Michel Sartori for the conversations that have reminded me just how amazing these animals really are, and for serving as my co-examiner. To my colleagues at EAWAG Limnology; to Mäggi Hieber who helped sample nearly every single animal in the study and accompanied me on many alpine adventures; to Dave Arscott, who was always willing to discuss just about anything and always had something valuable to say, and to Sandra Lass for her constant encouragement and ability to make me laugh. Saying "thank you" seems completely inadequate. There are many others that I thank for their help, support, and encouragement. They are Urs Uehlinger, Peter Burgherr, Friederika Mösslacher (I use the coffee mug every day), Florian Malard, Barbara Keller, Christine Boesch, Diana Soldo, Rainer Zah, Monika Winder, Nanna Büsing, Edith Kaiser, Luana Bottinelli, Andreas Frutiger, Klement Tockner, Gabriella Meier, Mark Gessner, Massimiliano Gili, Tamara Barthelmes, Bettina Wagner, Frank Sunder, Richard Illi, Christine Rapin, Christa Jolidon, Aachim Pätzold, Cécile Claret, Ute Karaus, Laurance Meunier, Fransiska Pfister, Christian Ohlendorf, Gerhard Mohler, Phil Shenkel, Christian Rust, Bergit Klein, Uli Donath, Karsten Rinke, Robert Skvarc, Michael Döring (mmm.... Bizoccals and Kübels), Christian Jakob, Sven Schalla, Marcos de la Puenta, Richard Glatthar, Yeda Arscott, Peggy Mittaud, Holger Bertram, and Nanna Büsing. At mol in EAWAG-MIX, many thanks to Christoph Werlen, Vladimir Sentchilo, Kirsten Lawlor, Jan Roelof van der Meer, and Rik Eggen. I thank Thomas Scheurer and Florin Filli at the Swiss National Park for the moonlight barbeque at Zernez during the controlled floods and for their enthusiasm. Lastly, thanks to Heidi Zimmerman, mountain guide and chain smoker extraordinaire. At Otago University, I am grateful to Mike Roy, Graham Wallis, Gigi Ostrow, Cécile Perrin, Patrick Dwyer, Makoto, Karen Judge, Carolyn 116

Burns, De-Arn Buchholz, Renata Sponer, John Waters, and Ken Stewart. Thank you all for your help, hospitality, and answers to my many questions. Finally, and most of all, I thank the Schwamendingen WG. Where to begin?... Marie for being my closest friend, Dimitry and Sophie for all the leftovers on the stove while I was writing; Michael for being the rock; Elke for being one of the coolest people I've ever met; and Anna for tolerating my long-out-of-practice violin playing enough to endure the occasional duet. Thank you all for giving me a home.