Swiss Federal Institute for Forest, Snow and Landscape Research WSL (2012). ENHANCE. Enhancing ecosystem connectivity through intervention - benefits for nature and society? Final Report. Swiss Federal Research Institute WSL.

(1)

ENHANCE

Enhancing ecosystem connectivity through intervention – benefits for nature and society?

Final Report

(2)

(3)

Enhancing ecosystem connectivity through intervention – benefits for nature and society?

Final Report

Swiss Federal Institue for Forest, Snow and Landscape Research WSL, Birmensdorf

(4)

Zürcherstrasse 111 8903 Birmensdorf Switzerland

Recommended form of citation

Swiss Federal Institute for Forest, Snow and Landscape Research WSL (ed) 2012: ENHANCE. Enhancing ecosystem connectivity through intervention – benefits for nature and society? Final Report. Birmensdorf, Swiss Federal Research Institute WSL, 81 pp.

PDF-Download: http://www.wsl.ch/publikationen/pdf/11617.pdf

Daniela Keller Matthias Buchecker Gwenaëlle Le Lay Tsipe Aavik Maarten van Strien Armin Peter

(5)

Contens

ENHANCE: Enhancing ecosystem connectivity through intervention – benefits for 5 nature and society?

Janine Bolliger

Landscape genetics: key contributions from ENHANCE 11

Janine Bolliger, Rolf Holderegger

The role of management intervention in enhancing breeding-habitat networks 13 in the European tree frog

Gwenaëlle Le Lay, Sonia AngeloneRolf Holderegger, Christoph Flory, Felix Kienast, Janine Bolliger

The potential genetic consequences of seed mixtures 17

Tsipe Aavik, Peter Edwards, Rolf Holderegger, Regula Billeter

Landscape genetics of insects in intensive agriculture: new ecological insights 27 Daniela Keller, Maarten J. van Strien, Jaboury Ghazoul, Rolf Holderegger

Connectivity of river habitats: population genetic survey on the effect of river 37 fragmentation on Swiss brown trout

Hitoshi Araki, Armin Peter, Julian Junker, Laura Langeloh

Effects of increased landscape connectivity on specialized solitary bee species 47 in agricultural habitats

Silvia Dorn

Longitudinal connectivity of river systems 53

Armin Peter, Anton Schleiss

Urban connectivity 57

Sonja Braaker, Martin K. Obrist, Fabio Bontadina, Marco Moretti

Understanding the societal and economic significance of ecosystem qualities and 63 their enhancement for people and stakeholders

Robert Home, Marcel Hunziker

Participatory planning of river revitalization projects: what can it contribute 71 to the transformation towards more adaptive socio-ecological systems?

Matthias Buchecker, Susanne Menzel

Management-intervention costs for damselfly Coenagrion mercuriale in the Oberaargau 79 Irmi Seidl

(6)

(7)

ENHANCE: Enhancing ecosystem connectivity through intervention – benefits for nature and society?

Janine Bolliger

Swiss Federal Research Institute WSL, Zürcherstrasse 111, 8903 Birmensdorf

Motivation for the ENHANCE project

In many parts of the world, human land use is the most important factor shaping landscapes including various socio-economic, policy and management driven processes (Bürgi et al. 2007). Many studies have shown a negative effect of intensive use and management, low landscape heterogeneity and high frag- mentation on species diversity (Purtauf et al. 2005; Vanbergen et al. 2005) with likely detrimental conse- quences for ecosystem services (Kremen 2005). Thus, conservation management has recognized landscape and habitat fragmentation as a key topic (Fahrig 2003; Lindenmayer and Fischer 2006; Baguette and Van Dyck 2007). Habitat fragmentation may lead to loss of habitat area and an increased isolation of remaining populations. Due to geographic distances exceeding movement capacities or high costs in overcoming landscape obstacles, habitat fragmentation may cause lowered movement of individuals and genes among populations (Baguette and Van Dyck 2007). Consequences of this decrease in connectivity between populations include negative impacts on the genetic diversity of species (e.g., Frankham 2003;

Bolliger et al. 2010), a process which may ultimately question the persistence of local populations (Fischer and Lindenmayer 2007).

Connectivity may be represented structurally and functionally (Lindenmayer and Fischer 2006). Structural connectivity refers to landscape structure only. For instance, suitable habitats such as dry meadows may be structurally connected by old field strips. In contrast, functional connectivity refers to the effective biotic movement (migration, dispersal, gene flow; (Tischendorf 2001; Tischendorf and Fahrig 2001). Depending on the organism, structurally connected landscapes or habitats may be functionally connected or not as structural connectivity is differently perceived depending on the organism (Van Dyck and Baquette 2005).

Conservation-management actions may mitigate decreasing habitat connectivity by providing stepping- stones or other de-fragmentation measures to re-connect (remaining) populations to a functional network (Van Dyck and Baquette 2005). Measures to increase connectivity are manifold and range from building individual stepping-stone ponds locally to large-scale measures which are also anchored legally (e.g., agri-environmental schemes). All measures have in common that they pursue the maintenance or restoration of functional connectivity within intensively managed systems. In aquatic habitats, measures such as river widening, gravel bank forming, installation of fish ramps etc. are taken. In terrestrial habitats, the provisioning of stepping-stone habitat to increase connectivity in landscapes includes for example setting aside agricultural land. In urban habitats, the fastest growing land-use type worldwide, green spaces are important stepping stones and green roofs may act as key de-fragmentation measures in densely settled areas.

Despite the large sums of money that authorities spend on management interventions to increase habitat connectivity, it remains unclear, whether the implemented measures to enhance the structural connectivity also serves the purpose to increase functional connectivity. Additional empirical knowledge is thus urgently needed (Stöcklin et al. 2007; Lesbarrères et al. 2010).

As connectivity is a central feature of ecosystem capacity, it is a societal need that scientific understanding of connectivity is improved and integrated into management action. ENHANCE tackles this challenge in a concerted action of different research fields (terrestrial and aquatic population ecology, population genetics, landscape ecology, socio-economics) of institutions of the ETH domain (WSL, ETHZ, EAWAG, EPFL). ENHANCE goals were:

(8)

1. To quantify the effects of increased landscape, cityscape, and riverscape connectivity (de-fragmentation) on various biotic levels (genes, species, communities) prior and after management intervention 2. To test fundamental questions of species movement in relation to de-fragmentation interventions in

landscapes, cityscapes and riverscapes

3. To provide societal and economic assessments of the role of connectivity in the people-landscape relationships since successful development and implementation of management options require knowledge of people’s attitudes towards ecosystem enhancement, of economic consequences, and of the availability of effective (participatory) planning measures.

The ENHANCE project – a summary

The interdisciplinary CCES project ENHANCE was conducted between 2008 and 2012. The project tested and evaluated ecosystem connectivity with state-of-the art molecular genetic experiments and population-dynamic analyses. ENHANCE used the expertise of WSL, ETHZ, EAWAG and EPFL teams to quantify species-specific population viability in aquatic and terrestrial habitats prior and after structural connectivity was enhanced with management or experimental interventions. The findings were used to link measures of structural connectivity in landscapes with species-specific responses. Since enhancing ecosystem connectivity is a conservation and management issue of highest priority, we provided a societal and economic assessment of recently performed and future interventions which aimed at increasing structural connectivity. Special emphasis was given to evaluating people’s attitudes and perception towards ecosystem enhancement as well as the involved costs which were required to build or maintain enhancement structures.

Particularly innovative biological aspects in ENHANCE are:

1. Assessments on the relationship between land- river- and cityscape structure and their func- tionality for a variety of organisms in different habitats (Aavik et al. 2012; Araki et al. 2012; Braaker et al. 2012; Dorn 2012; Keller et al. 2012; Le Lay et al. 2012; Peter and Schleiss 2012)

2. Landscape genetics: the combination of population-genetic methods with landscape- ecological and modeling approaches offer a unique and new setting to test emerging questions of species movement and gene flow in relation to land-use and intervention (Aavik et al. 2012; Araki et al. 2012; Bolliger 2012; Keller et al. 2012; Le Lay et al. 2012); including a winter school on the topic of landscape genet- ics (Bolliger et al. 2010; Bolliger et al. 2012).

3. Experimental set-ups at the landscape, cityscape and riverscape scale which assessed the movement pattern of different animal species (insects (Braaker et al. 2012; Dorn 2012), hedgehogs (Braaker et al. 2012), damselfly (Keller et al. 2012), fish (Araki et al. 2012; Peter and Schleiss 2012) and a plant species (Aavik et al. 2012).

Particularly innovative from a socio-economic perspective was the provisioning of:

1. People’s attitudes towards ecosystem enhancement (Home and Hunziker 2012) 2. Financial costs of ecosystem enhancement (Seidl 2012)

3. Effective and efficient (participatory) planning measures and corresponding institutional settings for implementation of emerging land management options (Buchecker and Menzel 2012)

(9)

ENHANCE addressed the following questions:

Q1: What are effects of land structures¹ on biotic interactions?

Q2: Do structural management interventions² cause ecological benefits ?

Q3: Does the broad public care about structural management interventions for ecosystem enhancement?

Q4: Do the ecological benefits of structural management interventions justify their costs?

1 landscape elements which may represent barriers/corridors for biotic interaction;

2 ecological compensation areas

Topics and methods in ENHANCE:

The project identified the relationship between structural and functional connectivity at various spatial scales and biotic levels for three habitats (agricultural, urban, river) and was structured into four modules according thematic and methodological similarities (Fig. 1): Genetics (Module M1): this module provided molecular genetic analyses of species to identify recent gene flow across landscapes with or without management interventions using microsatellites or AFLPs. The module referred to project questions Q1-2.

Population biology (Module M2): this module used population-biological parameters to identify func- tional connectivity in landscapes and riverscapes with and without management intervention. The module tackled the project questions Q1-2. Landscape analysis (Module M3): this module provided quantitative assessments of landscapes and links functionally assessed connectivity using genetic methods (M1) to landscape structure using spatial models (questions Q1-2). Socio-economy (Module M4): this module identified the relevance of ecosystem enhancement for society (Q3) and provided a cost assessment of intervention measures (Q4).

Fig. 1: ENHANCE modules, selected habitats, intervention types, and investigated species

(10)

Ecologically investigated types of management measures and interventions:

Agricultural habitats in M2 (lead: Silvia Dorn, ETH): We compared pre- and post-interventional condi- tions for agricultural areas in the Grenchner Witi (SO) and Berlingen (TG), i.e. model situations where extensive meadows have been artificially reconnected (de-fragmentation) and model situations where meadows were structurally disconnected (fragmentation). Model organisms were bees and flowering plants (Fig. 1).

Agricultural habitats in M1 and M3 (lead: Rolf Holderegger, Janine Bolliger, WSL): We identified the role of ecological compensation areas as structural management measure on the functional connectivity for different grasshopper species and an endangered damselfly in an intensively managed agricultural setting (Oberaargau; Fig. 1).

River habitats in M2 (lead: Armin Peter, EAWAG; Anton Schleiss, EPFL): The aim of river restoration projects (river widening, gravel bank forming, installation of fish ramps etc.) is to increase the lateral and longitudinal connectivity along catchments. We worked with pre- and post-interventional conditions, i.e.

rivers where restoration interventions increased longitudinal and latitudinal connectivity versus channeled situations. The model organism was brown trout (Fig. 1).

Urban habitat in M2 (lead: Marco Moretti and Martin Obrist, WSL): The relationship between structural and functional connectivity in an urban setting (Zürich) was assessed using field experiments for insects and carabids (traps) and GPS tracking for hedgehogs (Fig. 1). Green roofs, as third dimension of connectivity, were particularly focused on for flying species.

Socio-economically investigated types of management measures and interventions (M4) (lead: Marcel Hunziker, Matthias Buchecker, Irmi Seidl):

Agricultural habitats (ecological compensation areas, ponds, ditches), urban habitats, river habitats (river restoration)

References

Aavik, T., P. Edwards, R. Holderegger, and R. Billeter. 2012. The potential genetic consequences of seed mixtures. EN- HANCE: Enhancing ecosystem connectivity through intervention – benefits for nature and society? Final Report. 17–26.

Araki, H., A. Peter, J. Junker, and L. Langeloh. 2012. Connectivity of river habitats: population genetic survey on the effect of river fragmentation on Swiss brown trout. ENHANCE: enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 37–46.

Baguette, M., and H. Van Dyck. 2007. Landscape connectivity and animal behaviour: functional grain as key determinant of dispersal. Landscape Ecology 22: 117–129.

Bolliger, J. 2012. Landscape genetics: key contributions from ENHANCE. ENHANCE: Enhancing ecosystem connectivity through intervention – benefits for nature and society? Final Report. 11–12.

Bolliger, J., R. Holderegger, and F. Gugerli. 2012. Winter school on landscape genetics. ProClim Flash, April 2012.

Bolliger, J., G. Le Lay, and R. Holderegger. 2010. Landscape genetics – how landscapes affect ecological processes. Gaia 19: 238–240.

Braaker, S., M. Obrist, F. Bontadina, and M. Moretti. 2012. Urban connectivity. ENHANCE: Enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 57–62.

Buchecker, M., and S. Menzel. 2012. Participatory planning of river revitalization projects: What can it contribute to the transformation towards more adaptive socio-ecological systems? ENHANCE: enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 71–78.

Bürgi, M., A. Hersperger, M. Hall, E. W. B. Russell, and N. Schneeberger. 2007. Using the past to understand the present land-use and land-cover. Pages 133-145 in F. Kienast, S. Ghosh, and O. Wildi, editors. A changing world: challenges for landscape research. Kluwer Academic Publisher.

Dorn, S. 2012. Effects of increased landscape connectivity on specialized solitary bee species in agricultural habitats EN- HANCE: Enhancing ecosystem connectivity through intervention – benefits for nature and society? Final Report. 47–52.

Fahrig, L. 2003. Effects of habitat fragmentation on biodiversity. Annual Review of Ecology and Evolution 34: 487–515.

Fischer, J., and D. B. Lindenmayer. 2007. Landscape modification and habitat fragmentation: a synthesis. Global Ecology and Biogeography 16: 265–280.

(11)

Frankham, R. 2003. Genetics and conservation biology. Conservation Biology 326: 22–29.

Home, R., and M. Hunziker. 2012. Understanding the societal and economic significance of ecosystem qualities and their enhancement for people and stakeholders. ENHANCE: Enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 63–70.

Keller, D., M. J. Van Strien, J. Ghazoul, and R. Holderegger. 2012. Landscape genetics of insects in intensive agriculture:

new ecological insights. ENHANCE: Enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 27–36.

Kremen, C. 2005. Managing ecosystem services: what do we need to know about their ecology? Ecology Letters 8: 468–479.

Le Lay, G., S. Angelone, C. Flory, R. Holderegger, and J. Bolliger. 2012. The role of management intervention in enhancing breeding-habitat networks in the European tree frog. ENHANCE: Enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 13–16.

Lesbarrères, D., M. S. Fowler, A. Pagano, and T. Lodé. 2010. Recovery of anuran community diversity following habitat replacement. Journal of Applied Ecology 47: 148–156.

Lindenmayer, D. B., and J. Fischer 2006. Habitat fragmentation and landscape change. An ecological and conservation synthesis, Washington.

Peter, A., and A. Schleiss. 2012. Longitudinal connectivity of river systems. ENHANCE: enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 53–56.

Purtauf, T., J. Dauber, and V. Wolters. 2005. The response of carabids to landscape simplification differs between trophic groups. Oecologia 142: 458–464.

Seidl, I. 2012. Management-intervention costs for the damselfly Coenagrion mercuriale in the Oberaargau. ENHANCE:

enhancing ecosystem connectivity through managemetn intervention – benefit for nature and society? Final Report. 79–81.

Stöcklin, J., A. Bosshard, G. Klaus, K. Rudmann-Maurer, and M. Fischer 2007. Synthesebericht NFP 48. Landnutzung und biologische Vielfalt in den Alpen. Fakten, Perspektiven, Empfehlungen. vdf, Zürich.

Tischendorf, L. 2001. Can landscape indices predict ecological processes consistently? Landscape Ecology 16: 235–254.

Tischendorf, L., and L. Fahrig. 2001. On the usage and measurement of landscape connectivity. Oikos 90: 7–19.

Van Dyck, H., and M. Baquette. 2005. Dispersal behaviour in fragmented landscapes: routine or special movement? Basic and Applied Ecology 6: 535–545.

Vanbergen, A. J., B. A. Woodcock, A. Watt, and J. Niemelä. 2005. Effect of land-use heterogeneity on carabid communities at the landscape scale. Ecography 28: 3–16.

(12)

The ENHANCE team (current addresses)

Aavik Tsipe, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, tsipe.aavik@wsl.ch Araki Hitoshi, EAWAG, Seestrasse 79, CH-6047Kastanienbaum, hitoshi.araki@eawag.ch

Billeter Regula, Institute of Integrative Biology, ETH, Universitätsstrasse 16, CH-8092 Zürich, regula.billeter@env.ethz.ch Bolliger Janine, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, janine.bolliger@wsl.ch Braaker Sonja, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, sonja.braaker@wsl.ch Buchecker Matthias, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

matthias.buchecker@wsl.ch

Buser Tobias, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, tobias.buser@wsl.ch Di Giulio Manuela, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

manuela.digiulio@wsl.ch

Dorn Silvia, Institute of Plant Sciences, ETH, Schmelzbergstrasse 9, CH8092 Zürich, silvia.dorn@ipw.agrl.ethz.ch Edwards Peter, Institute of Integrative Biology, ETH, Universitätsstrasse 16, CH-8092 Zürich, peter.edwards@env.ethz.ch Ghazoul Jaboury, Ecosystem Management, ETH, Universitätsstrasse 16, CH-8092 Zürich, jaboury.ghazoul@env.ethz.ch Holderegger Rolf, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

rolf.holderegger@wsl.ch

Home Robert, Research Institute of Organic Agriculture FiBL, Ackerstrasse, CH-5070 Frick, robert.home@fibl.org Hunziker Marcel, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

marcel.hunziker@wsl.ch

Jaeger Jochen, Department of Geography,Planning and Environment, Concordia University, jjaeger@alcor.concordia.ca Keller Daniela, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, daniela.keller@wsl.ch Kienast Felix, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, felix.kienast@wsl.ch Le Lay Gwenaëlle, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

gwenaelle.lelay@gmail.com

Longatti Peter, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, peter.longatti@wsl.ch Menzel Susanne, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

susanne.menzel@wsl.ch

Moretti Marco, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, marco.moretti@wsl.ch Obrist Martin, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, martin.obrist@wsl.ch Peter Armin, EAWAG, Seestrasse 79, CH-6047Kastanienbaum, armin.peter@eawag.ch

Schleiss Anton, Laboratoires de Constructions Hydrauliques, EPFL, CH-1015 Lausanne, anton.schleiss@epfl.ch Seidl Irmi, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, irmi.seidl@wsl.ch Van Strien Maarten, Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf,

maarten.vanstrien@wsl.ch

Zurbuchen Antonia, ProNatura, Lehnstrasse 35, CH-9014 St. Gallen, antonia.zurbuchen@pronatura.ch

(13)

Landscape genetics: key contributions from ENHANCE

Janine Bolliger, Rolf Holderegger

Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland The innovative field of landscape genetics has been promoted in various review articles (Storfer et al.

2007; Holderegger and Wagner 2008; Segelbacher et al. 2010), international workshops (Balkenhol et al. 2009) and (international) teaching activities (ETHZ: Bolliger et al. (2010), DGS: Wagner et al. (2012)).

Landscape genetics amalgamates population genetics and landscape ecology by combining theory, concepts and methods of population genetics with the spatially dynamic framework of landscape ecol- ogy, spatial statistics and modelling (Manel et al. 2003). Landscape ecology assesses the relationship between spatially dynamic patterns and processes in landscapes by using statistical models including the characterization of landscape structure (e.g., suitable versus unsuitable habitats, barriers, corridors etc.), while population genetics uses a variety of genetic techniques (e.g. AFLPs, microsatellites or SNPs) to describe the genetic structure of populations. Landscape genetics thus provides a powerful tool for explaining genetic structure and gene flow based on spatially dynamic patterns and processes (Fahrig 2003; Cushman 2006; Sander et al. 2006) and allows to identify the effects of explicit landscape proper- ties on ecological processes.

Major scientific benefits of landscape genetics are that it provides a powerful conceptual framework for identifying and explaining dispersal, migration, and gene flow based on spatial (and dynamic) landscape patterns (Fahrig 2003; Cushman 2006; Sander et al. 2006). The combination of molecular genetic and landscape or environment data with modeling approaches also offers a unique setting to test emerging questions of species behavior land-use change. For example, landscape genetics allows identifying the degree of individual or gene exchange as well as its directionality among populations or individuals, and therefore provides direct tests of functional connectivity in relation to landscape structure. This is in distinct contrast to much ecological research, but also practical management interventions, where simple structural connectivity is used as an indicator of functional connectivity. It is thus possible that entire scientific chapters relying on traditional population biological or ecological approaches (e.g. habitat fragmentation) have to be re-written based on insights emerging from landscape genetics (Epperson 2003). What we learned from landscape genetics so far shows that dispersal, migration and gene flow are often more abundant and occur over larger distances than had hitherto been suggested based on ecological studies (Holderegger and Wagner 2008; Bolliger et al. 2011). Thus, landscape genetics provides concepts and methods to contribute significantly to basic and applied research, especially conservation management, the evaluation of connectivity measures and landscape planning.

ENHANCE contributed to the advancement of landscape genetics with the following topics:

1. The role of management intervention in enhancing breeding-habitat networks in the European tree frog (Le Lay et al. 2012).

2. The potential genetic consequences of seed mixtures in restoration (Aavik et al. 2012).

3. Landscape genetics of seven insect species in intensive agriculture: new ecological insights (Keller et al. 2012).

4. Connectivity of river habitats: population genetic survey on the effect of river fragmentation on Swiss brown trout (Araki et al. 2012).

5. We initiated a one week Master and PhD winter school on landscape genetics for ETH, conducted at WSL (Bolliger et al. 2010; Bolliger et al. 2012).

(14)

References

Aavik, T., P. Edwards, R. Holderegger, and R. Billeter. 2012. The potential genetic consequences of seed mixtures. EN- HANCE: Enhancing ecosystem connectivity through intervention – benefits for nature and society? Final Report.

Araki, H., A. Peter, J. Junker, and L. Langeloh. 2012. Connectivity of river habitats: population genetic survey on the effect of river fragmentation on Swiss brown trout. ENHANCE: enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 17–26.

Balkenhol, N., F. Gugerli, S. A. Cushman, L. P. Waits, R. Holderegger, H. H. Wagner, and Participants of the Landscape Genetics Research Agenda Workshop 2007. 2009. Identifiying future research needs in landscape genetics: where to go from here? Landscape Ecology 24: 455–463.

Bolliger, J., R. Holderegger, and F. Gugerli. 2012. Winter school on landscape genetics. ProClim Flash, April 2012.

Bolliger, J., D. Keller, and R. Holderegger. 2011. When landscape variables do not explain migration rates: an example from an endangered dragonfly (Leucorrhinia caudalis). European Journal of Entomology 108: 327–330.

Bolliger, J., G. Le Lay, and R. Holderegger. 2010. Landscape genetics – how landscapes affect ecological processes. Gaia 19: 238–240.

Cushman, S. A. 2006. Effects of habitat loss and fragmentation on amphibians: a review and prospects. Biological Conser- vation 128: 231–240.

Epperson, B. K. 2003. Geographical Genetics. Princeton University Press, Princeton.

Fahrig, L. 2003. Effects of habitat fragmentation on biodiversity. Annual Review of Ecology and Evolution 34: 487–515.

Holderegger, R., and H. H. Wagner. 2008. Landscape genetics. BioScience 58: 199–207.

Keller, D., M. J. Van Strien, J. Ghazoul, and R. Holderegger. 2012. Landscape genetics of insects in intensive agriculture:

new ecological insights. ENHANCE: Enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 27–36.

Le Lay, G., S. Angelone, C. Flory, R. Holderegger, and J. Bolliger. 2012. The role of management intervention in enhancing breeding-habitat networks in the European tree frog. ENHANCE: Enhancing ecosystem connectivity through intervention – benefit for nature and society? Final Report. 13–16.

Manel, S., M. Schwartz, G. Luikart, and P. Taberlet. 2003. Landscape genetics: combining landscape ecology and population genetics. Trends in Ecology and Evolution 18: 189–197.

Sander, A.-C., T. Purtauf, I. J. Holzhauer, and V. Wolters. 2006. Landscape effects on the genetic structure of the ground beetle Poecilus versicolor STURM 1824. Landscape Ecology 15: 245–259.

Segelbacher, G., S. A. Cushman, B. K. Epperson, M. J. Fortin, O. Francois, O. J. Hardy, R. Holderegger, P. Taberlet, L. P.

Waits, and S. Manel. 2010. Applications of landscape genetics in conservation biology: concepts and challenges. Con- servation Genetics 11: 375–385.

Storfer, A., M. A. Murphy, J. S. Evans, C. S. Goldberg, S. Robinson, S. F. Spear, R. Dezzani, E. Delmelle, L. Vierling, and L.

P. Waits. 2007. Putting the “landscape” in landscape genetics. Heredity 98: 128–142.

Wagner, H. H., M. A. Murphy, R. Holderegger, and L. Waits. 2012. Developing an interdisciplinary distributed graduate course for tweny-first century scientists. BioScience 62: 182–188.

(15)

The role of management intervention in enhancing breeding- habitat networks in the European tree frog

Gwenaëlle Le Lay¹, Sonia Angelone^1,3Rolf Holderegger¹, Christoph Flory², Felix Kienast¹, Janine Bolliger¹

1 Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland

2 CreaNatira, Stroppelstr. 9, 5417 Untersiggenthal

3 Grün Stadt Zürich, Beatenplatz 2, 8001 Zürich

Summary

Functional connectivity in fragmented landscapes is a critical factor to facilitate movement between habitats and populations. Management actions which aim at fostering habitat networks are often promoted but rarely functionally evaluated. Here, we explored the number and origin of colonising individuals at newly created stepping-stone ponds for the threatened European tree frog (Hyla arborea) in the Reuss valley, Switzerland. Collecting buccal material of 85 frogs sampled at newly built stepping-stone ponds, we identified source populations of migrating frogs by means of 11 microsatellite markers. Results show that new ponds were colonized within one year, and that the number of colonising frogs increased with the new ponds’ age. Some migrating individuals originated from small or distant populations (up to 5 km), and even crossed expected barriers such as a river. The landscape in the study area appeared thus quite permeable to tree frogs. Our measures revealed that building new stepping-stone breeding ponds is an efficient and successful conservation action.

State of the art and motivation: how effective is the provisioning of stepping- stone breeding habitats for tree-frogs?

Amphibians are highly affected by human-impacted ecosystems (Stuart et al. 2004) and one third of amphibian species are threatened by extinction (Baillie et al. 2004). Main factors are the loss of habitat area and its fragmentation (Cushman 2006b; Funk et al. 2005). A few recent studies looked at de- fragmentation such as the provisioning of stepping-stone ponds (Lesbarrères et al. 2010, Le Lay et al.

submitted), however, the effectiveness of mitigation actions are difficult to measure as it remains unclear whether the individuals colonising new stepping stones stem from nearby populations or whether they form a much broader sample of individuals originating from more distant populations. Thus, detailed evaluation of the functional permeability of a landscape is crucial, in particular, as the decision on the extent and location of measures in conservation management rarely rely on organism-based functional connectivity (Lesbarrères et al. 2010). One promising tool to perform such analyses includes landscape genetics. Landscape genetics seeks to assess how ecological processes such as migration, dispersal and gene flow are affected by landscape structure and composition. The field amalgamates population genetics and landscape ecology by combining theory, concepts and methods of population genetics with the spatially dynamic framework of landscape ecology, spatial statistics and modelling (Manel et al.

2003). Landscape ecology and population genetics are both based on well-established disciplinary methods. Landscape ecology assesses the relationship between spatially dynamic patterns and processes in landscapes by using statistical models including the characterization of landscape structure (e.g., suitable versus unsuitable habitats, barriers and corridors), while population genetics uses a variety of genetic techniques (e.g., AFLPs, microsatellites or SNPs) to describe the genetic structure of populations or individuals. Landscape genetics thus provides a powerful tool for explaining genetic structure and gene flow based on spatially dynamic patterns and processes (Fahrig 2003; Sander et al. 2006; Cushman 2006a) and allows to identify the effects of explicit landscape properties on ecological processes.

(16)

Genetic methods for conservation management

We applied genetic assignment tests (Manel et al. 2005) to address the role of management intervention in enhancing landscape permeability by building new stepping stone breeding ponds for the threatened European tree frog Hyla arborea L., Figs. 1–3). Genetic methods permit reliable assessments of contem- porary or recent movement directions, which allow drawing conclusions on the functional landscape connectivity. We benefit from a rare well documented dataset, i.e. genetic data encompassing all populations before and after the construction of new ponds. First, we identified the number of colonising frogs present in the newly built ponds which were implemented to improve the breeding-habitat network for the tree frog. Second, we identified the origins of colonising tree frogs to assess the spatial range of movement.

The Reuss valley offered a unique situation to study the geographic origin of tree frogs colonising new stepping-stone ponds in real time (Fig. 2). The study of Angelone and Holderegger (2009) provided a data set of all established tree-frog populations in the Reuss valley in 2006, i.e. before new ponds were built.

Our sampling campaign in 2009 used the 2006 data as reference to evaluate the effectiveness of newly built stepping-stone breeding ponds. In spring 2009 we sampled buccal swabs from 85 tree frogs at four new ponds for genetic analysis. We applied the same allelic scoring scheme as presented by Angelone and Holderegger (2009). In brief, DNA extraction was conducted with the DNeasy Tissue Kit (QUIAGEN), and microsatellites were amplified in four multiplexed polymerase chain reactions and analysed on a 3130 automated sequencer (Applied Biosystems). Peaks were scored using GENEMAPPER 3.7 (Applied Biosystems).

Key insight: stepping-stone ponds are a successful conservation-management strategy to foster tree-frog habitat networks

Within one to three years after construction, all newly built ponds were occupied by tree frogs. This confirms the species’ assumed strong colonization ability (Fog 1993, Lesbarrères et al. 2010) but also demonstrates that the tree-frog habitat network in the study area was already well functioning (Angelone et al. 2011). We not only found fast and abundant colonisation of new ponds, results also showed that the age of the new ponds mattered as the oldest newly created pond exhibited the largest amount of new colonizers. Colonizing individuals originated from populations at distances of up to 5 km, with 50 % of the first generation migrants originating from source populations at distances of less than 2 km. Thus, the landscape of the Reuss valley appears to be generally well permeable for tree frogs. As the number of colonising frogs increased during the first years of the pond creation, and as these frogs come from a large variety of population sources, we conclude that the provisioning of new stepping-stone breeding ponds is a successful conservation strategy to increase the number of breeding sites and to enhance the gene flow between tree-frog populations. The so supplemented habitat network de-centralizes individual habitat importance which leads to an increased resilience of the entire habitat network (Bolliger et al., in prep).

Fig. 1. The European tree-frog (Hyla arborea); © Gwenaëlle Le Lay.

(17)

Fig. 2. Tree-frog sampling sites in the Reuss valley, Switzerland; © Gwenaëlle Le Lay.

Fig. 3. Newly created stepping-stone breeding habitats for the tree frog in the Reuss valley, Switzerland;

(18)

References

Angelone, S. & Holderegger, R. (2009) Population genetics suggests effectiveness of habitat connectivity measures for the European tree frog in Switzerland. Journal of Applied Ecology, 46: 879–887.

Angelone, S., Kienast, F. & Holderegger, R. (2011) Where movement happens: scale-dependent landscape effects on genetic differentiation in the European tree frog. Ecography, 34: 714–722.

Baillie, J. E. M., Hilton-Taylor, C. & Stuart, S. N. (2004) Red List of threatened species. A global species assessment. IUCN, Gland, Switzerland and Cambridge, UK.

Bolliger, J., Angelone, S., Le Lay, G., Flory, C. & Holderegger, R. (in prep) Importance of habitat and connectivity in tree-frog breeding networks. Molecular Ecology.

Cushman, S. A. (2006a) Effects of habitat loss and fragmentation on amphibians: a review and prospects. Biological Con- servation, 128: 231–240.

Cushman, S. A. (2006b) Effects of habitat loss and fragmentation on amphibians: A review and prospectus. Biological Conservation, 128: 231–240.

Fahrig, L. (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology and Evolution, 34: 487–515.

Fog, K. (1993) Migration in the tree frog Hyla arborea. Ecology and Conservation of the European Tree Frog (eds A. H. P.

Stumpel & U. Tester), pp. 55–64. Institute for Forestry and Nature Research, Wageningen.

Funk, W. C., Blouin, M. S., Corn, P. S., Maxell, B. A., Pilliod, D. S., Amish, S. & Allendorf, F. W. (2005) Population structure of Columbia spotted frogs (Rana luteiventris) is strongly affected by the landscape. Molecular Ecology Resources, 14:

483–496.

Le Lay, G., Angelone, S., Holderegger, R., Flory, C. & Bolliger, J. (submitted) The role of management intervention in enhancing breeding-habitat networks in the European tree frog. Conservation Biology.

Lesbarrères, D., Fowler, M. S., Pagano, A. & Lode, T. (2010) Recovery of anuran community diversity following habitat replacement. Journal of Applied Ecology, 47: 148–156.

Manel, S., Gaggiotti, O. E. & Waples, R. S. (2005) Assignment methods: matching biological questions techniques with appropriate. Trends In Ecology & Evolution, 20: 136–142.

Manel, S., Schwartz, M., Luikart, G. & Taberlet, P. (2003) Landscape genetics: combining landscape ecology and population genetics. Trends in Ecology and Evolution, 18: 189–197.

Sander, A.-C., Purtauf, T., Holzhauer, I. J. & Wolters, V. (2006) Landscape effects on the genetic structure of the ground beetle Poecilus versicolor STURM 1824. Landscape Ecology, 15: 245–259.

Stuart, S. N., Chanson, J. S., Cox, N. A., Young, B. E., Rodrigues, A. S. L., Fischman, D. L. & Waller, R. W. (2004) Status and trends of amphibian declines and extinctions worldwide. Science, 306: 1783–1786.

(19)

The potential genetic consequences of seed mixtures

Tsipe Aavik¹, Peter Edwards², Rolf Holderegger¹, Regula Billeter²

1 Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf

2 Institute for Integrative Biology, ETH Zürich

Scientific summary

Sowing with commercial wildflower seed mixtures is a common restoration practice in areas with impoverished species pools. The potential genetic consequences of using seed mixtures, however, are poorly understood and often not considered in practical restoration. One of the key objectives of such restoration measures is to enhance connectivity among isolated natural and semi-natural habitats, e. g. by establish- ing new habitat patches and (sown) populations. Nevertheless, improvement to functional habitat connectivity is rarely assessed when evaluating restoration success. Furthermore, the impact of landscape structure on habitat connectivity is seldom examined.

Our study area was located in an intensively managed agricultural landscape in the Oberaargau area, Switzerland. A few years before our study took place, several extensively managed grasslands were recreated in this landscape. Restoration activities included sowing with commercially produced wildflower seed mixtures. We studied the genetic diversity and structure of nine sown and 17 naturally occurring populations of Lychnis flos-cuculi, which is an insect-pollinated grassland species. We found that sown and natural populations had similar gene diversity and allelic richness. Inbreeding, by contrast, was significantly higher in sown populations. Source populations, where the seeds for propagation in seed companies were collected, may have been small and/or inbred resulting in higher inbreeding in sown populations. Inbreeding can also be caused by repeated regeneration of the same seed stock over several cycles at the seed company. We also found that sown populations were genetically very distinct from natural ones despite the fact that source populations of sown plants originated from the same seed zone as the restored sites. These results suggest that seeds for propagation should be collected from numer- ous individuals in large and non-isolated populations nearby restored sites. Stocks for the production of seed mixtures should only be propagated for a small number of generations to avoid negative effects such as inbreeding and loss of local adaptation.

To study the effect of provenance, genetic diversity and composition on plant fitness, we measured the fitness of the study populations of L. flos-cuculi. In addition, we established an experiment in the study area as well as in an experimental garden in Zürich by sowing seeds originating from natural populations, sown populations and seed companies. We recorded the establishment, survival and fitness of the experimental plants. We detected no significant effect of genetic diversity on the fitness of study plants, which suggests that neutral genetic diversity examined in the present study may not have a direct relationship with the adaptive genetic variation.

In order to evaluate functional connectivity among restored and remnant grasslands, we examined contemporary gene flow patterns of L. flos-cuculi using assignment tests and first-generation migrant tests. Assignment tests, which reflect gene flow during several generations, revealed high gene flow among the natural populations of L. flos-cuculi. By contrast, little gene flow occurred between sown and natural populations. Furthermore, we detected only a few first-generation migrants among sown and natural populations as well as among natural populations, which indicates insufficient spatial connectivity of extensively managed grasslands in this landscape. Alternatively, gene flow occurred more often than detected, but lower adaptation of sown genotypes to local environmental conditions or higher inbreeding observed in these populations could have impeded establishment.

Additionally, we examined the effects of Euclidian distance, cumulative elevation change and the propor- tion of various landscape elements (forest, settlements, ditch verges, agricultural land) in corridors (cor-

(20)

ridor width of 50, 100, 200, 300, 500, 1000 m) between the natural populations of L. flos-cuculi on gene flow among those populations. Only forest had a significant positive effect on the genetic differentiation FST. Forest inhibits gene flow among the study populations most probably through hampering the movement of pollinators.

The findings of the current study are of high significance for nature conservation and ecological restoration. In particular, we suggest that more attention should be paid to the genetic quality of seed mixtures used in ecological habitat restoration. In addition, the evaluation of restoration measures should include assessing the improvement to connectivity of habitats. Our findings suggest that enhancing spatial connectivity through restoration measures does not necessarily increase functional connectivity in short term.

State of the art (pre-ENHANCE)

Sowing with commercial seed mixtures has become a common practice for restoring species diversity in areas with impoverished species pools (Kiehl et al. 2010). In spite of the increasingly broad application of this measure all over Europe, the genetic composition and diversity of plant populations originating from commercial seed mixtures has received practically no attention. Furthermore, the potential influence of gene exchange between sown and local natural populations on the genetic properties and subsequent fitness of species is largely unknown. If seeds from seed mixtures are characterised by lower genetic diversity, which is often related to lower fitness in case of self-incompatible species (Leimu et al. 2006), then sowing commercial seed mixtures in large quantities may become detrimental to natural populations by “polluting” the natural gene pool with genotypes exhibiting lower fitness. Gene exchange among plants from seed mixtures and local populations may result in outbreeding depression, which can also have negative consequences for plant fitness (Montalvo and Ellstrand 2001). Additionally, seeds of non-local origin may be poorly adapted to the environmental conditions at the restoration site, and therefore exhibit significantly lower fitness compared to local natural populations (Bischoff et al. 2006). Furthermore, the ex-situ propagation of seeds in gardens favours genotypes that are well suited to garden conditions but not necessarily to those in the restored habitat (Ensslin et al. 2011). To decrease the negative effects of using unsuitable seeds, conservation biologists have suggested sowing seeds, which originate from the same seed zone as the restored site (Vander Mijnsbrugge et al. 2010). However, seed zones are mostly defined on the basis of climatic and bio-geographical data, while there is no information on how these zones correspond to natural patterns of genetic variation (Kramer and Havens 2009). There is also insufficient data about the effects of repeated propagation on the genetic variation of seed stocks.

The success of ecological restoration measures is often judged by the increase in the number and abun- dance of plant species. In practical conservation, the area and spatial connectivity of habitats are fre- quently considered as indicators of restoration success. However, there is limited knowledge, how restoration measures influence functional connectivity among plant populations. Direct estimation of functional connectivity of plants (e.g. by tracking the movement of seeds or pollen) is complicated, and can only be done for certain plant species (Sork 1984; Van Geert et al. 2010). Furthermore, ecological methods may strongly misjudge the extent and amount of seed and pollen flow (Kamm et al. 2009). Genetic methods offer an alternative means of estimating functional connectivity by assessing historical as well as contemporary gene flow (Lowe and Allendorf 2010). Regrettably, genetic methods have rarely been applied for studying the functional connectivity of plants in response to restoration measures.

Genetic connectivity among plant populations is mostly predicted as a function of Euclidian distance between populations (Honnay et al. 2007; Jacquemyn et al. 2007; Mix et al. 2006). However, genetic structuring may also be influenced by landscape characteristics between populations and not by geographic distance alone (Holderegger and Wagner 2008). Landscape genetic approaches, which combine the methods of population genetics and landscape ecology, have increasingly been applied to study the effect of landscape properties on the dispersal of individuals. However, most landscape genetic studies are focused on vertebrata, while only about 15 % of studies deal with plants (Holderegger et al. 2010;

Storfer et al. 2010).

(21)

Motivation and research questions for the project

Little is known about the genetic diversity and composition of commercially produced seed mixtures used in ecological restoration. Therefore, our first aim was to examine whether natural and sown populations of the study species differ in regard to their genetic characteristics. In particular, we wanted to know if there are any differences in the genetic diversity (measured as gene diversity, allelic richness, observed heterozygosity, inbreeding) and structure of sown and natural populations. Because genetic diversity may have an effect on the fitness of plants, we also examined the relationship between the genetic properties and plant fitness of sown and naturally occurring populations.

The assessment of restoration effectiveness does usually not encompass evaluating the improvements to functional connectivity. Our study system of sown and natural plant populations enabled us to examine whether and how much gene flow occurred among restored and natural populations since restoration. In one of the two study regions, half of the populations had been restored by sowing wildflower seed mixtures eight years before the study; in the other region, two populations had been sown three years prior to our sampling, while most populations were of natural origin. Thus we could also compare the effect of age and amount of restored populations on gene flow.

Few studies have examined the effect of landscape structure on the gene flow among plant populations, whilst most studies concentrate on animals. In the present study, we firstly examined how much the spatial connectivity of habitats (measured by geographic distance) does mirror the functional connectivity among populations (measured as genetic differentiation FST). Secondly, we studied the effect of various landscape elements on gene flow among the study populations.

Technical issues: material, methods and sampling

Study species

Lychnis flos-cuculi L. (Silene flos-cuculi; Caryophyllaceae) is a diploid polycarpic perennial herb (Fig. 1).

It grows in moist, open habitats such as floodplains and calcareous fen meadows, and is distributed throughout Europe (Jalas 1986). L. flos-cuculi is an insect-pollinated species visited by a wide range of pollinators: Lepidoptera, Diptera and Hymenoptera (Van Rossum and Triest 2010). The species is self- compatible, but the protandrous flowers of L. flos-cuculi are predominantly outcrossed (Biere 1991). In addition to sexual reproduction, L. flos-cuculi forms vegetative rosettes from axillary stem buds. Plants overwinter as rosettes. In the second year, they produce stems which are 30–90 cm tall and flower between April and June. Ripe capsules contain an average of about 150 seeds. Seeds are released by vibrations of the stiffened stalks (Biere 1991).

Study populations and fieldwork

The study was carried out in an intensively managed agricultural landscape located in the Cantons of Bern and Aargau in Switzerland (Fig. 2). As part of a restoration programme, several new streams were created in the area between 2001 and 2003. The verges of these watercourses were sown with standard wildflower seed mixtures developed for extensively managed meadows or wet meadows, containing the study species L. flos-cuculi (UFA seed company, Winterthur). In 2006–2007, a few wet and mesotrophic grasslands were restored using the same type of seed mixtures. Most of these sown areas were managed as ecological compensation areas: no fertilizers were used and mowing took place once per year after June 15. In addition to nine sown populations of L. flos-cuculi, we found 17 naturally occurring popu- lations of L. flos-cuculi in the study area (Tab 1).

In large populations, we sampled leaves from approximately 30 individuals for genetic analysis (Table 1).

However, in small populations, fewer individuals could be sampled. Most of the populations at ditch verges and field margins were long and narrow. To take into account the spatial structure of these linear populations, we divided them into sectors of approximately 70–100 m. Within each sector, we collected the leaves of about 30 individuals, with a distance of approximately 2–3 meters between sampled plants.

(22)

(a) (b)

Fig. 1. The study species Lychnis flos-cuculi (a) and one of the restored habitats in the study area (b) hosting a sown popu- lation of L. flos-cuculi.

The largest linear population comprised 13 sectors (238 individuals were sampled). In total, 1413 indi- viduals of L. flos-cuculi were sampled (813 in natural populations and 600 in sown populations). In each study population, soil samples were collected to estimate soil nitrogen and phosphor content. In addition, soil moisture was measured using the HydroSense Soil Water Measurement Systems CD620 (Campbell Scientific).

In order to study the effect of origin (sown, natural) and genetic characteristics on the fitness of study populations, we measured various fitness traits of approximately 30 randomly selected reproductive individuals in 20 of the study populations of L. flos-cuculi in spring 2011. We counted the number of flow- ers and stalks per plant and measured plant height. Ripe seeds were collected from 30 randomly selected individuals in each population. A hundred seeds per individual were weighed. Fifty seeds of 15 individuals per population were placed on filter paper in Petri dishes (in total 300 Petri dishes). Petri dishes were placed in a greenhouse at 20 °C with 16 hours of light and were regularly watered with tap water. The number of germinated seeds was counted after 30 days.

We established an experiment in the study area and in an experimental garden to examine the effect of provenance and genetic characteristics on plant fitness. The seeds from two natural populations (Natural 13, Natural 23) and two sown populations (Sown 1, Sown 3) were collected in summer of 2010 for the experiment. In addition, seeds of L. flos-cuculi were ordered from two seed companies in Switzerland (UFA seed company, CH-Wildblumen). In September 2010, the experiment was set up at two sites in the study area and in the experimental garden at ETH Hönggerberg in Zürich. In all sites, four blocks with the size of 2 x 3 m were established. Within each block, six plots with the size of 0.75 x 0.75 m were created by removing the above- and below-ground vegetation. In the experimental garden, 24 pots with the size of 0.74 x 0.56 x 0.37 m were filled with a mixture of soil and sand and were covered with a 10-cm layer of humus on top. At the beginning of October 2010, we sowed 200 seeds in each plot/pot. Every block con- tained a plot with seeds of different origin: two sown populations, two natural populations and two plots with seeds originating from two seed companies. At the end of May 2011, seedlings were counted and harvested so that in every plot ten seedlings remained (when possible). At the beginning of July, August and September, we recorded survival and measured the fitness of experimental plants (plant diameter, length of the longest leaf, height, number of stalks, number of flowers). At the beginning of September we collected the above-ground parts of plants for biomass measurements. Biomass samples were weighed after drying in an oven at 70 °C for 48–72 hours.

(23)

Fig. 2. Black continuous lines denote the locations of natural populations and black dashed lines the locations of sown populations of Lychnis flos-cuculi. Grey lines indicate ditches and other water bodies with flowing water; grey polygons show forested areas.

Genetic analysis

Plant material collected for genetic analysis was dried and kept in silica gel until used. DNA was extracted from 10 mg of dry leaf material using the Dneasy 96 Plant Kit (QIAGEN). We used three microsatellite markers developed for L. flos-cuculi: Cuculi 4, Cuculi 17, Cuculi 19 (Galeuchet et al. 2002), and, in ad- dition, a selection of primers developed for Silene latifolia (Caryophyllacea; Moccia et al. 2009), a close relative of L. flos-cuculi: SL_eSSR13, SL_eSSR17 and SL_eSSR49. Polymerase chain reactions (PCR) were carried out as described in Aavik et al. (2012). PCR products were analyzed on an ABI 3730 auto- mated sequencer (Applied Biosystems) using 400 ROXä size standard. Allele lengths were visualized and scored using GENEMAPPER 3.7 (Applied Biosystems).

Table 1. Origin (sown, natural), age, coordinates (E, N), population size, sample size, allelic richness (AR), gene diversity (HE), observed heterozygosity (HO) and inbreeding coefficient (FIS) of the study populations of Lychnis flos-cuculi in two study regions.

Location Genetic

population Population

age (years) E N Pop.

size Sample

size AR HE HO FIS

Region 1

Sown 1 Sown I 8 621167 227643 1360 176 5.12 0.66 0.55 0.110

Sown 2 Sown I 8 622172 227874 2050 60 5.23 0.68 0.58 0.142

Sown 3 Sown I 8 623210 227882 1040 123 6 0.7 0.62 0.121

Sown 4 Sown I 8 624749 229019 600 75 5.46 0.68 0.58 0.119

Sown 5 Sown I 8 624628 229915 170 45 5.19 0.67 0.56 0.175

Sown 8 Sown II 8 623342 228322 2000 30 5.75 π0.7 0.65 0.068

Sown 9 Sown II 8 623570 228875 500 30 5.25 0.65 0.53 0.173

Natural 11 Natural 11 Natural 621919 230642 100 30 5.23 0.66 0.62 0.056

Natural 12 Natural 12 Natural 622202 230752 100 30 5.34 0.7 0.717 -0.026

Natural 16 Natural 16 Natural 624406 229357 20 20 6 0.68 0.64 0.062

(24)

Region 2

Sown 6 Sown I 3 627183 230936 12 12 5.33 0.64 0.53 0.176

Sown 7 Sown I 3 627331 232740 300 49 6.10 0.71 0.65 0.067

Data analyses

We calculated allelic richness AR, gene diversity HE, inbreeding coefficient FIS and observed heterozygosity HO of each study population using FSTAT 2.9.3.2 (Goudet 1995). The differences in HE, AR, HO and FIS

between sown and natural populations using log-transformed population size as a covariate were tested with non-parametric distance-based (Euclidian) permutation tests implemented in R (vegan package;

Oksanen, et al. 2008). The distribution of molecular variation among sown and natural populations, within groups and within populations was evaluated with the analysis of molecular variance (AMOVA; Excoffier et al. 1992) implemented in ARLEQUIN 3.11. Individuals were clustered by applying the Bayesian Monte Carlo Markov Chain (MCMC) method implemented in STRUCTURE 2.3.3 (Pritchard et al. 2000).

The effects of origin (sown, natural), genetic diversity and environmental variables on the fitness of study populations were analysed using linear mixed-effects models in R (packages nlme (Pinhero et al. 2012) and lme4 (Bates et al. 2011)). The same models were used for studying the effect of seed origin (sown, natural, seed company) on the fitness of experimental plants in the study area and in an experimental garden.

Recent gene flow among sown and natural populations was estimated using assignment tests (Rannala and Mountain 1997) and first-generation migrant tests (Paetkau et al. 2004) implemented in GENE- CLASS 2.0 (Piry et al. 2004). In this analysis, we divided the study area into two regions being spatially separated by the town of Langenthal. Genetic differentiation FST among sown populations was very low.

Thus, there was a high probability that assignment tests would place a migrant originating from any of the sown populations to a wrong source population due to high genetic similarity. We therefore pooled the sown populations within the same genetic cluster (according to STRUCTURE 2.3.3; Pritchard et al. 2000) together in assignment and first-generation migrant tests (“genetic population” in Table 1).

We carried out a corridor analysis to examine the effect of landscape variables on gene flow among the natural populations of L. flos-cuculi. Using ARCGIS 9.3.1 (ESRI), we calculated the amount of various landscape elements (agricultural land, settlements, forests, ditch verges) within corridors between populations (corridor widths of 50, 100, 200, 300, 500 and 1000 m). The effect of rank-transformed percentages of landscape elements within corridors on pairwise genetic differentiation FST among populations was estimated using multiple regression on distance matrices provided in R (package ecodist; Goslee and Urban 2007).

Location Genetic

population Population

age (years) E N Pop.

size Sample

size AR HE HO FIS

(25)

Innovation, gains, new insights and main results thanks to ENHANCE

Our first aim was to examine the genetic characteristics of sown and natural populations of L. flos-cuculi.

Genetic analysis revealed no significant differences between gene diversity HE and allelic richness AR of sown and natural populations (Fig. 3a). However, sown populations were characterized by significantly lower observed heterozygosity HO and correspondingly higher inbreeding coefficients FIS in comparison with natural populations (Fig. 3b, Table 1). High inbreeding may have negative consequences for various aspects of fitness such as germination, survival and reproductive output (Hauser and Loeschcke 1995;

Galeuchet et al. 2005), and can thus seriously jeopardize restoration success. Several reasons may have caused higher inbreeding in sown populations. First, the source populations, where the seeds were collected for propagation at the corresponding seed company, may have been small and inbred. Secondly, inbreeding in sown populations may have arisen when only a few source individuals were sampled causing genetic bottleneck effects and increasing the influence of genetic drift (Williams 2001). Thirdly, an increase in inbreeding can occur due to repeated regeneration of the same seed stock over several cycles (Schoen and Brown 2001), which is a common practice of seed companies. Consequently, to avoid inbreeding in seed mixtures, seeds for propagation should be collected from a substantial number of individuals in large and well connected populations. Additionally, seed stocks should be renewed after a few regeneration cycles.

(a) (b)

Fig. 3. Mean (bars) and standard errors (whiskers) of gene diversity HE (a) and inbreeding coefficient FIS (b) in natural (white bars) and sown (grey bars) populations of Lychnis flos-cuculi. Asterisks denote a significant difference (**P < 0.01) between the two groups. (From Aavik et al. 2012).

Sown populations of L. flos-cuculi were genetically very distinct from natural populations (Aavik et al.

2012), although they originated from the same seed zone as the restored site. Furthermore, amongst the nine sown populations, we could distinguish two very distinct gene pools, which most likely represent two different source populations used for seed propagation at the seed company. However, despite the genetic differences between sown and natural populations, measurements of population fitness as well as an experiment in the study area revealed no significant influence of gene diversity or inbreeding on plant fitness. Neutral genetic diversity examined in the present study may not have a direct relationship with the adaptive genetic variation (Reed and Frankham 2001), which could be one of the reasons for the lack of correlation between fitness and genetic diversity. It is, nevertheless, also possible that the studied range of inbreeding (FIS = 0–0.15) and gene diversity (HE = 0.57–0.71) was too narrow to detect a response of fitness.