Patterns of arthropod distribution and determinants of arthropod assemblage composition in a natural West African savannah Dissertation zur Erlangung des naturwissenschaftlichen Doktorgrades der Bayerischen Julius-Maximilians-Universität Würzburg vorgelegt von Karsten Mody aus Mannheim Würzburg 2003 Eingereicht am: 25.04.2003................................................................. Mitglieder der Promotionskommission: Vorsitzender: Prof. Dr. Rainer Hedrich............................................ Gutachter: Prof. Dr. K. Eduard Linsenmair.................................. Gutachter: Prof. Dr. Bert Hölldobler............................................. Tag des Promotionskolloquiums: ..................................................................................... Doktorurkunde ausgehändigt am: ..................................................................................... From large to small scales - and vice versa? or How does it affect an African caterpillar when I go to work by car? Frontside (top): Planet Earth showing Antarctica at bottom, and Africa and Madagascar at centre (arrow indicates Côte d’Ivoire). From: Drury 1998, modified. Frontside (bottom): Chrysopsyche imparilis, an African caterpillar, meeting a Camponotus ant on Combretum fragrans. Table of Contents I. General Introduction......................................................................................................1 II. Ökologische Gemeinschaften und Verständnis von Biodiversität: Welche Faktoren strukturieren artenreiche Arthropodengemeinschaften?...................................9 III. Study Sites, Species and Methods................................................................................15 IV. Persistent Composition of Arthropod Communities on Combretum fragrans ...............25 V. Organisation of Arthropod Assemblages in Individual African Savannah Trees ...........................................................................................................................31 VI. Neighbouring Plants Host Different Ants: Analysis of the Ant Mosaic on a Myrmecophilic Tree....................................................................................................49 VII. Plant-attracted Ants Affect Arthropod Community Structure but not Necessarily Herbivory.................................................................................................65 VIII. Finding its Place in a Competitive Ant Community: Leaf Fidelity of Camponotus sericeus...................................................................................................77 IX. Persistent Aggregation of Herbivorous Beetles on Conspecific Host Plant Individuals: What are the Causes? ...............................................................................89 X. Intraspecific Host-Plant Change Affects Growth and Fecundity in a Specialist Caterpillar: Roaming for Fitness? ..............................................................107 XI. General Discussion....................................................................................................117 XII. Summary...................................................................................................................123 XIII. Zusammenfassung.....................................................................................................127 XIV. References.................................................................................................................131 XV. Danksagung...............................................................................................................145 I. General Introduction and Thesis Outline 1 I. General Introduction Human activities alter global, regional and local processes (Vitousek et al. 1997, Chapin et al. 2000). They affect biogeochemical cycles, habitats, ecosystems, biotic communities and the distribution of species. Although they may locally increase species numbers by introduction of nonindigenous species, human activities result in a world-wide decline in species diversity, a homogenisation of species communities and a loss of genetic information (Ehrlich and Ehrlich 1981, Meffe and Carroll 1997, Lodge and Shrader-Frechette 2003). Human-induced environmental changes thereby reduce opportunities for future use and evolution of species (Western 2001), and they can negatively affect the productivity and vital functions of ecosystems (Schulze and Mooney 1993, Tilman 1999). This would not only deteriorate the quality of human life but can eventually threaten the whole functionality of Planet Earth’s biosphere. Thus, the development of measures counteracting these negative impacts should be in everybody’s interest whose thinking is not restricted to a present time welfare. The success of such measures will depend on the fundamental understanding of processes connected with anthropogenic alterations and with conservation and restoration efforts (Dobson et al. 1997, Symstad et al. 2003). In order to understand biological processes and the effects of anthropogenic activities on them, an adequate knowledge of unaltered processes is essential as they have been the basis for the current and (still) working systems (Linsenmair 1990, Ewel 1999). However, this source of information is vanishing since many anthropogenic influences (e.g. climate change, rising atmospheric carbon dioxide concentrations and nitrogen deposition) are already affecting even the most remote and best protected areas (Gallagher and Carpenter 1997). If natural systems can provide unique insights into the functioning of environmental processes, it thus seems necessary to understand them as fast as possible. Biological systems are always influenced by processes acting on very different spatial and temporal scales, regardless of whether they are strongly determined by human activities (urban landscapes, agricultural fields, plantations, etc.) or more or less natural (remote, protected areas). Consequently, they cannot be fully understood without reference to local, regional and global processes (Ricklefs 1987, Morton and Law 1997, Lawton 2000). The studies presented here will focus on local, small-scale processes acting on coexisting and potentially interacting populations of organisms, using arthropod communities on different species of savannah trees as study system. The results of this work, however, should be considered in a broader context. They will serve to establish larger-scale studies investigating the influence of regional factors, for example comparisons of arthropod communities of the same habitat-entities (trees of the same species) in different West African countries, possibly embedded in differing habitat matrices. They will allow to test for the effects of anthropogenic land conversion and to monitor influences coming along with a continuously changing environment due to ‘global change’ (both in tropical and temperate regions)(Singh 2002). The study of arthropod communities on trees and shrubs (arboricolous arthropod communities) of an unsettled West African savannah (Comoé National Park, Republic of Côte d’Ivoire) is especially informative and promising for several reasons. 1) The study area can be considered to be almost natural. Although changes in composition of savannah vegetation occurred in historical times as a consequence of an anthropogenically increased bushfire-frequency, it still represents the largest area of protected and unconverted savannah habitat in West Africa. At least for the studied processes (arthropod species pool, plant 2 I. General Introduction and Thesis Outline population structure), it provides the basic data on natural systems mentioned above, which may be unavailable from areas more heavily altered by anthropogenic influences. 2) The type of habitat investigated has not been studied in detail hitherto with respect to insect-plant interrelationships. Therefore, the studies provide new information for an ecosystem which is typical for large areas of Africa between the Sahelian and the tropical forest zone and which is strongly threatened by climate change and land conversion. The studies thereby also contribute to the concerted rapid ‘action plan’ called for by Basset et al. (2003b) to investigate canopies in African savannahs in a concerted, immediate way. 3) Arboricolous arthropod communities are regarded as a model system to study the relationship between local environmental variables and the structure of speciose communities (Moran and Southwood 1982). These communities can easily be delimited (all arthropods living on one tree) and sampled in a representative, almost complete way (depending on the size of the tree and the sampling methods used). This allows to investigate the factors regulating community structure (abundance, identity and number of species) and to compare different theories that have been proposed to explain community structure (Morin 1999). For the understanding of community organisation, two explanation-approaches are especially important (Wiens 1984, Linsenmair 1990). The first states communities to be a product of species- specific habitat requirements and interspecific competition in an environment characterised by limited resources – so called equilibrium communities or ‘deterministic’ community organisation. In case the specific requirements are known for the species, the composition of individual communities would then be predictable. The second approach highlights the importance of random processes (colonisation events, disturbances), which allow different species