Diversity and species composition of two different moth families (Lepidoptera: Arctiidae vs. Geometridae) along a successional gradient in the Ecuadorian Andes Dissertation zur Erlangung des Grades eines Doktors der Naturwissenschaften - Dr. rer. nat. - der Fakultät Biologie / Chemie / Geowissenschaften der Universität Bayreuth Vorgelegt von Nadine Hilt Bayreuth, Oktober 2005 Die vorliegende Arbeit wurde von März 2002 bis Oktober 2005 am Lehrstuhl Tierökologie I der Universität Bayreuth in der Arbeitsgruppe von Prof. Dr. Konrad Fiedler erstellt und von der Deutschen Forschungsgemeinschaft gefördert (Projekt FOR 402/1-1 TP 15, FOR 402/2-1 TP A3). Vollständiger Abdruck der von der Fakultät Biologie/Chemie/Geowissenschaften der Universität Bayreuth genehmigten Dissertation zur Erlangung des Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.). Tag der Einreichung: 17.10.2005 Tag des wissenschaftlichen Kolloquiums: 23.01.2006 1. Gutachter: Prof. Dr. K. Fiedler 2. Gutachter: Prof. Dr. S. Liede-Schumann Prüfungsausschuss: Prof. Dr. K. H. Hoffmann (Vorsitz) Prof. Dr. E. Beck Prof. Dr. K. Fiedler Prof. Dr. S. Liede-Schumann Prof. Dr. G. Rambold Contents 1 General introduction................................................................................. 1 2 Diversity and composition of Arctiidae moth ensembles along a successional gradient in the Ecuadorian Andes* ....................................... 19 3 Arctiid moth ensembles along a successional gradient in the Ecuadorian montane rainforest zone: how different are subfamilies and tribes?*.................................................................................................... 54 4 Morphological traits of arctiid moths along a succession gradient in southern montane Ecuador: clades differ more than habitats .............. 88 5 Diversity and ensemble composition of geometrid moths along a successional gradient in the Ecuadorian Andes* ........................ 119 6 Temporal dynamics of rich moth ensembles in the montane forest zone in southern Ecuador ................................................................ 150 7 Synopsis ................................................................................................ 179 8 Summary................................................................................................ 207 9 Zusammenfassung................................................................................ 213 10 Resumen ................................................................................................ 219 Darstellung des Eigenanteils ...................................................................... 225 Acknowledgements...................................................................................... 227 Appendix 1.................................................................................................... 229 Appendix 2.................................................................................................... 235 Erklärung....................................................................................................... 252 * published 1 General introduction Tropical rainforests are well known as centres of biodiversity (Mittermeier et al., 1999). Much interest has focused on the ecological processes responsible for generating and maintaining this diversity (e.g. Richards, 1996). The Andean rainforests are acknowledged as global diversity hotspots for vascular plants and vertebrates (Myers et al., 2000; Brooks et al., 2002; Myers, 2003). However, Andean montane forests have already been reduced to less than 10% of their original extent through human activities (Henderson et al., 1991; Hamilton et al., 1995; Armenteras et al., 2003). The annual deforestation rate is estimated at 0.8–2% (Doumenge et al., 1995; Food and Agriculture Association, 1995; Purvis & Hector, 2000), and the remaining remnants are often threatened by fire and logging (Paulsch, 2002), as well as by transformation into pastures or plantations of exotic tree species such as Eucalyptus and Pinus species (Cuarón, 2000). A particular problem in southern Ecuador is the encroachment of more than 25% of burned areas with bracken (Pteridium arachnoideum: Hartig & Beck, 2003). Thus, Andean montane rainforests must be considered as highly endangered ecosystems. Since 1997, the German Research Foundation (DFG) has been funding an interdisciplinary research project in the montane zone of southern Ecuador to examine the edaphic, climatic, zoological and botanical components of the forest ecosystem (see http://www.bergregenwald.de). Altitudinal as well as disturbance gradients were chosen as paradigms to gain a better understanding of patterns of biodiversity (e.g. through inventories of the forest and adjacent habitats), determinants of biodiversity, ecosystem processes, and the conservation needs of this ecosystem (Beck & Müller-Hohenstein, 2001). Even though preservation of large areas of continuous, pristine habitat is regarded as crucial for the conservation of many species (e.g. Brooks et al., 2002), human-dominated landscapes globally cover major proportions of land area (Ricketts et al., 2001). Fragmented landscapes and secondary habitats thus are becoming increasingly important for consideration with regard to the 1 General introduction conservation of biodiversity. A considerable number of studies have dealt with effects of logging on arthropod communities (e.g. Willott, 1999; Schulze 2000; Schulze et al., 2004; Meijaard et al., 2005). Clear-cut logging and subsequent conversion of forest into plantations and agricultural areas generally results in decreased insect diversity (Holloway et al., 1992; Bawa & Seidler, 1998), but effects of moderate disturbance are less clear (Hill, 1999) and frequently less dramatic (e.g. Willott, 1999; Schulze, 2000; Beck et al., 2002; Hamer et al., 2003). So far, little is known about the impact of different land use practices on species-rich neotropical insect assemblages native to the rainforest (Ricketts et al., 1999; Brehm & Fiedler, 2005) and the change of these assemblages along habitat gradients or man-made disturbance gradients, representing different successional stages of forest recovery after logging (e.g. Floren & Linsenmair, 2001 for an example from Borneo). Moreover, effects of forest disturbance on species diversity and composition are heavily scale dependent (Hamer & Hill, 2000; see also Rahbek, 2005). While relatively many studies were carried out for insects in Southeast Asian forests (e.g. Chey et al., 1997; Holloway, 1998; Willott, 1999; Schulze, 2000; Beck et al., 2002; Fiedler & Schulze, 2004), the neotropical region – even though much richer in species – has received surprisingly little attention (Ricketts et al., 2001; Brehm & Fiedler, 2005). Particularly few studies on insect diversity were performed in the montane regions of the Andes (Janzen et al., 1976; Braun, 2002; Brehm, 2002; Süßenbach, 2003). Total global species richness is currently estimated to be about 10 million species (summarized in May, 1990; Stork, 1993). A re-evaluation of Erwin’s estimate of tropical arthropod species richness arrived at a total of 4.8 million species (Gaston, 1991; Ødegaard, 2000; Novotny et al., 2002). Insects, including the species-rich Lepidoptera, play a central role in all terrestrial ecosystems. Lepidoptera are important herbivores, pollinators, and serve as food and hosts for multiple other organisms at higher trophic levels (Summerville & Crist, 2004; Summerville et al., 2004). The better-known groups of Lepidoptera (butterflies in particular) have often been advocated as useful indicators of environmental 2 1 General introduction change (Daily & Ehrlich, 1995; Hill et al., 1995; Hill & Hamer, 1998). However, more than 90% of the known lepidopteran species are moths, and the majority of them are nocturnal (Scoble, 1992; Young, 1997). Adult moths are accessible to standardized sampling through light traps, and the Macrolepidoptera (derived large moths; a monophyletic group: Solis & Pogue, 1999) are taxonomically relatively well known (Scoble, 1992, 1999; Holloway, 1993, 1996, 1997; Holloway et al., 2001). Moreover, moths appear to be at least as well suitable as butterflies for assessing environmental impacts (e.g. Holloway, 1985; Kitching et al., 2000; Beck et al., 2002; Fiedler & Schulze, 2004). Therefore, I chose two different moths families, namely the Arctiidae (Plate 1) and the Geometridae (Plate 2), as model organisms for a biodiversity study in disturbed and succession habitats at the edge of a natural montane rainforest in the Ecuadorian Andes. Species of both families tend to differ in their life histories, their habitat fidelity and habitat preferences (e.g. Holloway, 1984; Schulze, 2000; K. Fiedler, pers. comm.). I thus expected that arctiid and geometrid moths should respond differentially to environmental change. The Arctiidae include about 11,000 species worldwide, with more than 50% occurring in the Neotropics (Watson & Goodger, 1986). In this region, arctiid moths fall into two subfamilies, viz. the Lithosiinae and Arctiinae. The latter can be subdivided into four major tribes (Arctiini, Phaegopterini, Pericopini, and Ctenuchini with the latter including Euchromiini: Jacobson & Weller, 2002). Five characters support the monophyly of Arctiidae (Jacobson & Weller, 2002). Of the synapomorphies of adult Arctiidae, the most unambiguous is the presence of a pair of dorsal, eversible pheromone glands associated with the anal papillae of females. In addition,