168 PARALLEL SESSION HALL H IV THURSDAY LINK FOR SURVIVALPARALLE SESSION THURSDAY H III: LINK FOR SURVIVAL169HABITAT THRESHOLD EFFECTS ON GENETIC DIVERSITY ANDDIFFERENTIATION: EMPIRICAL EVIDENCE FROM A SPECIALISTSPECIES IN THREE ATLANTIC FOREST LANDSCAPESNiko Balkenhol 1 , Renata Pardini 2 , Cintia Cornelius 2 , Fabiano Fernandes 1 ,Simone Sommer 11IZW Berlin, Berlin, DE, balkenhol@izw-berlin.de, 2 Departmento de Zoologia,Instituto de Biociencias, Universidade de Sao Paulo, Sao Paulo, BrazilTHURSDAY 17:15 Hall H IVSimulation studies and literature reviews suggest the existence of afragmentation threshold, around 10-30% of remaining habitat, below whichbiodiversity loss is accelerated. Habitat thresholds could also affect differentcomponents of genetic variation, namely the amount of genetic variation(i.e., genetic diversity) and the distribution of genetic variation (i.e., geneticdifferentiation). Here, we provide the first test of habitat threshold effects ongenetic variation. Specifically, we used 12 microsatellite markers to analyze 529samples of Marmosops incanus, a small marsupial that is restricted to areas ofnative <strong>for</strong>est at both local and range-wide scales. Samples were gathered fromthree 10,000-ha landscapes differing only in the proportion of remaining <strong>for</strong>est(30, 50, and >80%). Using landscape- and patch-level analyses, we used thisdata set to address two interrelated questions: (1) Do genetic diversity, geneticdifferentiation and effective (i.e., genetic) immigration differ among landscapeswith varying proportion of <strong>for</strong>est cover? (2) Does the importance of local patchcharacteristics in determining genetic diversity, differentiation and effectiveimmigration vary among these landscapes? Congruent with our expectations,genetic diversity was significantly lower in the most de<strong>for</strong>ested landscape (30%<strong>for</strong>est cover) compared to the two most <strong>for</strong>ested landscapes. In contrast, geneticdifferentiation and effective migration did not differ between the 30% and 50%<strong>for</strong>est cover landscapes, while being significantly higher in the most <strong>for</strong>estedlandscape (> 80% <strong>for</strong>est cover). Finally, local patch characteristics explainedgenetic variation only in the most de<strong>for</strong>ested landscape. Our study providesstrong support <strong>for</strong> the existence of habitat threshold effects on genetic variation.Results demonstrate that different components of genetic variation respondto habitat loss at varying rates, and that local patch characteristics becomeincreasingly important at higher levels of habitat loss. Furthermore, our studyillustrates the utility of genetic estimates of migration rates to evaluate habitatloss effects on effective dispersal. These findings have important consequences<strong>for</strong> understanding landscape- and patch-level effects on genetic variation, and<strong>for</strong> conserving the most basic level of biodiversity in natural and human-alteredtropical landscapes.THURSDAY 17:15 Hall H IVgtö<strong>Society</strong> <strong>for</strong> <strong>Tropical</strong> <strong>Ecology</strong> | Gesellschaft für Tropenökologie e.V. Status and future of tropical biodiversity | Frankfurt, 21 - 24 February 2011gtö
170PARALLEL SESSION HALL H IV THURSDAY LINK FOR SURVIVALECONOMIC INCENTIVES FOR MITIGATING FRAGMENTATION ANDEDGE-EFFECTS IN THE BRAZILIAN ATLANTIC FORESTFlorian Hartig 1 , Sandro Pütz 1 , Cristina Banks-Leite 2 , Andreas Huth 1 ,Martin Drechsler 11UFZ Leipzig, Leipzig, DE, florian.hartig@ufz.de, 2 University of São Paulo, SãoPaulo, BRHabitat loss and fragmentation are the main causes of biodiversity loss in theMata Atlântica region. Rain<strong>for</strong>est is still being lost at an alarming rate, but thereis hope that <strong>for</strong>est protection laws and carbon sequestration initiatives suchas REDD (Reducing Emissions from De<strong>for</strong>estation and Degradation) will createincentives <strong>for</strong> preservation and re<strong>for</strong>estation in the future.Practically all of these initiatives express their targets in terms of <strong>for</strong>est areawithin a larger region. However, area-based policies alone bear the danger ofcreating <strong>for</strong>est patches that are too small and too isolated to support importantecosystem functions.In our contribution, we discuss how international <strong>for</strong>est protection initiativessuch as REDD and national <strong>for</strong>est protection laws could be altered to considernot only total area, but also size and connectivity of <strong>for</strong>est fragments. Weshow evidence from theoretical and empirical studies that have examined theapplicability of such structural incentives <strong>for</strong> conservation policy. We argue thatthe implementation of incentives that consider the spatial distribution of <strong>for</strong>eston a regional scale, together with prioritization of biodiversity hotspots onlarger spatial scales, will maximize cobenefits between carbon sequestration,conservation, and other ecosystem services that arise from preserving theAtlantic Forest.THURSDAY 17:30 Hall H IVgtö<strong>Society</strong> <strong>for</strong> <strong>Tropical</strong> <strong>Ecology</strong> | Gesellschaft für Tropenökologie e.V.