Southeast Asian Ant-Gardens - Diversity, ecology, ecosystematic significance, and evolution of mutualistic ant-epiphyte associations - Dissertation zur Erlangung des Doktorgrades der Naturwissenschaften vorgelegt beim Fachbereich Biologie und Informatik der Johann Wolfgang Goethe-Universität in Frankfurt am Main von Eva Kaufmann aus Bad Nauheim Bad Homburg 2002 Vom Fachbereich………………………………………………………………………… der Johann Wolfgang Goethe-Universität, Frankfurt am Main, als Dissertation angenommen. Dekan:………………………………………………………………………………………… Gutachter:……………………………………………………………………………………… Datum der Disputation:………………………………………………………………………….. Perhaps the most complex mutualism between plants and ants is the ant garden… (B. Hölldobler and E.O. Wilson, The Ants, 1990) Table of contents Contents 1. Introduction.........................................................................................1 1.1. Epiphytes...............................................................................................................1 1.2. Arboricolous ants...................................................................................................2 1.3. Ant-epiphyte interactions......................................................................................3 1.4. Topics of the present study....................................................................................4 2. Research Areas....................................................................................5 2.1. Ulu Gombak, Central Peninsular Malaysia...........................................................7 2.2. Pedawan Valley, Sabah, Borneo............................................................................7 3. Materials and Methods.....................................................................11 3.1. Inventory methods...............................................................................................11 3.2. Taxonomic classification.....................................................................................12 3.3. Microclimate of ant-gardens................................................................................12 3.4. Ecological significance of Southeast Asian ant-epiphyte associations...............12 3.5. Experimental investigations................................................................................13 3.5.1. Seed-carrying experiments..............................................................................13 3.5.2. Characterization of ant attractants in epiphyte seeds.......................................14 3.5.2.1. Seed-carrying experiments with non-AG ants.............................................14 3.5.2.2. Ageing, heating, washing, and extracts with different solvents..................15 3.5.2.3. Comparison of GC/MS profiles and IR spectra of several attractive seeds 15 3.5.2.4. HPLC and LC fractions in bioassays...........................................................16 3.5.3. Substrate properties.........................................................................................17 3.5.3.1. Analysis of nutrient content.........................................................................17 3.5.3.2. Analysis of water storing capacity...............................................................19 3.6. Statistical methods...............................................................................................19 4. Diversity of Southeast Asian ant-gardens........................................20 4.1. Results.................................................................................................................20 4.1.1. Inventory - True ant-garden mutualists and opportunists................................20 4.1.2. Detailed description of selected ant-garden systems.......................................25 4.1.2.1. Myrmicinae..................................................................................................27 Crematogaster sp 18......................................................................................27 KfmA Crematogaster sp 21......................................................................................39 KfmA Crematogaster sp 200....................................................................................41 KfmA Pheidole sp 33................................................................................................44 KfmA 4.1.2.2. Dolichoderinae.............................................................................................51 Philidris sp 37................................................................................................51 KfmA Philidris sp 160..............................................................................................56 KfmA 4.1.2.3. Formicinae...................................................................................................61 Camponotus sp 9............................................................................................61 KfmA Camponotus sp 240........................................................................................65 KfmA 4.1.2.4. Ponerinae.....................................................................................................73 Diacamma sp 111..........................................................................................73 KfmA 4.2. Discussion............................................................................................................78 4.2.1. Towards a classification of ant-garden systems..............................................78 4.2.2. Interspecific interactions..................................................................................84 4.2.2.1. Potential benefits of epiphytes to the ants...................................................84 4.2.2.2. Potential benefits of ants to the epiphytes...................................................86 4.2.2.3. Significance of additional ‘partners’ for ant-garden systems, and vice versa. .....................................................................................................................88 i Table of contents 5. Species sorting of ant-garden ants and ant-garden epiphytes........94 5.1. Results.................................................................................................................94 5.1.1. Interdependences of AG ant and epiphyte species..........................................94 5.1.2. Temperature.....................................................................................................96 5.1.3. Vertical position in host trees..........................................................................98 5.1.3.1. Ants.............................................................................................................98 5.1.3.2. Epiphytes...................................................................................................100 5.2. Discussion.........................................................................................................103 5.2.1. Crematogaster sp 18 (and Crematogaster sp 19)..............................103 KfmA KfmA 5.2.2. Philidris sp 160........................................................................................104 KfmA 5.2.3. Camponotus sp 9.....................................................................................104 KfmA 5.2.4. Camponotus sp 240.................................................................................105 KfmA 5.2.5. Diacamma sp 111....................................................................................105 KfmA 6. Assessment of the ecological significance of ant-gardens..............107 6.1. Results...............................................................................................................107 6.1.1. Importance of ants for the establishment of epiphytes in Peninsular Malaysia.. .......................................................................................................................107 6.2. Discussion.........................................................................................................108 6.2.1. Ecological importance of ant-epiphyte associations.....................................108 6.2.2. Radiation in response to ant-epiphyte interactions?......................................109 6.2.2.1. Philidris spp..............................................................................................110 6.2.2.2. The Asclepiadaceae...................................................................................111 6.2.2.3. The Rubiaceae...........................................................................................113 7. Characterization of ant attractants in epiphyte seeds...................114 7.1. Results...............................................................................................................114 7.1.1. Seed-carrying experiments with non-AG ants..............................................114 7.1.2. Ageing, heating, washing, and extracting.....................................................114 7.1.3. Comparison of GC/MS profiles and IR spectra of different epiphyte seeds.116 7.1.4. HPLC and LC fractions in bioassays............................................................120 7.2. Discussion.........................................................................................................123 7.2.1. Characteristics of ant attractants in wind-dispersed epiphyte seeds..............123 7.2.2. Myrmecochory – ‘Systems’..........................................................................125 7.2.2.1. Terricolous myrmecochores......................................................................125 7.2.2.2. Neotropical ant-gardens............................................................................127 7.2.2.3. General considerations on possible evolutionary pathways for myrmecochory..............................................................................................................129 7.2.2.4. Evolutionary ecology of myrmecochory in Southeast Asian ant-gardens 133 8. Concluding discussion.....................................................................135 8.1. The ‘ant epiphyte puzzle’ of Southeast Asia.....................................................135 8.2. Comparison of Southeast Asian and tropical American ant-gardens................138 8.2.1. Species composition......................................................................................138 8.2.2. Ant-garden establishment..............................................................................139 8.2.3. Ant nutrition and the significance of phorophytes........................................140 8.3. Evolution of palaeotropical ant-gardens............................................................140 8.3.1. Ants...............................................................................................................142 8.3.2. Epiphytes.......................................................................................................146 8.4. Perspectives for future studies...........................................................................150 9. Summary .........................................................................................152 10. Ausführliche Zusammenfassung....................................................154 11. References........................................................................................161 ii Table of contents 12. Appendix..........................................................................................173 12.1. Author index..................................................................................................173 12.2. Table Index....................................................................................................175 12.3. Figure Index...................................................................................................176 12.4. HPLC-Fractions.............................................................................................177 12.5. Parameters recorded for an inventory of ant-epiphyte interactions in Southeast Asia .......................................................................................................................177 12.6. Wind-dispersed seeds used in seed-retrieving bioassays..............................179 12.7. Item list for bioassays....................................................................................180 12.8. Instructions for the use of reflectometric tests...............................................185 12.8.1. Ammonium....................................................................................................185 12.8.2. Nitrate............................................................................................................186 12.8.3. Phosphate.......................................................................................................187 12.9. Distribution matrices for ants and epiphytes in Ulu Gombak.......................188 12.9.1. AG-ants..........................................................................................................188 12.9.2. Opportunists...................................................................................................189 12.10. List of epiphytes occurring in Peninsular Malaysia......................................190 12.11. Previous reports on palaeotropical ant-gardens.............................................196 12.12. Neotropical ant-garden epiphytes..................................................................198 12.13. Publications...................................................................................................200 13. Acknowledgements..........................................................................201 iii Introduction 1. Introduction Tropical rain forest canopies are among the most complex and at the same time the least studied of all terrestrial habitats. This is mainly because access to this stratum is particularly difficult. Technical progress over the last 40 years has helped to solve this problem to some extent, and many studies have been published since. However, these revealed that a large part of total arthropod biodiversity still remains to be discovered in tropical forest canopies (Erwin, 1983; Stork, 1988). Two groups of organisms are particularly important for canopy ecosystems (e.g. Benzing, 1990; Stuntz, 2001): Ants and epiphytes. In contrast to forests in temperate zones, epiphytes form a significant part of plant diversity in tropical rain forests (up to 30-50 % of the total vascular flora (Benzing, 1990)). They contribute to the complex structure of rain forest canopies and provide important resources for many canopy arthropods (Stuntz, 2001). On the other hand, ants make an important contribution in sheer numbers and biomass, though not in species richness (Adis et. al., 1984; Stork, 1988). Ants as eusocial insects form long-living colonies and have a major impact on many other organisms. It is not surprising, therefore, that ants and epiphytes interact in many ways. Such ant-epiphyte associations are the topic of the present thesis. Before various special types of ant-epiphyte associations are presented, an account will be given of common traits of epiphytes and some of the major problems arising from an epiphytic life style. The second section will do the same for arboricolous ants. Both chapters focus on topics relevant to ant-epiphyte interactions. 1.1. Epiphytes About 10 % of all vascular plant species are epiphytes (Kress, 1986), most of them occurring in tropical rain forests. Growing on a host tree, these plants use the space available in the third dimension. This life style involves a number of special problems touching on more or less all aspects of plant life: First of all, the propagule has somehow to come across a suitable growing site. Depending on species-specific requirements, ‘suitability’ might be confined by habitat, tree species or bark characteristics, vertical position in a host tree, and growth substrate (e.g. Went, 1940; Johansson, 1974; Steege and Cornelissen, 1989; Benzing, 1995; Callaway et al., 2002). According to Benzing (1990), the growing conditions of canopy flora often resemble those on the ground with regard to rooting substrate. In view of the inhomogeneity of tree crown structure and especially the fragmentation and scarcity of suitable substrate for epiphytes, the way in which epiphyte seeds or spores reach a new appropriate growing site becomes a crucial question. Several problems have to be solved in connection with the growth substrate. Water supply is difficult and temporary drought is likely to be lethal, especially for seedlings (Benzing, 1991). All sorts of physiological (e.g. Winter et al., 1983; Sinclair, 1983a,b, 1984; Goh and Kluge, 1989; Griffiths, 1989; Benzing, 1995), morphological (Johansson, 1974; Benzing, 1990, 1995) and ecological (Benzing, 1990) tricks to cope with temporary aridity have been ‘invented’ by epiphytes. Supply with mineral nutrients is 1 Introduction similarly difficult, a problem often solved by extremely slow growth (Benzing, 1991). Even pollination seems to require special strategies since epiphytes are usually hyperdispersed (Madison, 1977; Ackerman, 1986). Associations with arboricolous ants are one possible strategy to overcome some of these problems: Ants can provide reliable seed dispersal to suitable growing sites, accumulate substrate throughout the life of an epiphyte and thus fertilize the plants. These substrate accumulations can also serve as short-term water-reservoirs, thus reducing water stress (Davidson and Epstein, 1989). 1.2. Arboricolous ants Tropical arboricolous ants have been subject of many studies especially in the past 40 years. Fogging techniques revealed a high abundance and diversity of canopy ants (Erwin, 1983; Adis et al., 1984; Wilson, 1987; Stork, 1988; Tobin, 1991; Floren, 2000). According to Wilson (1959) and Tobin (1995), there is little overlap between ground nesting and arboricolous ant species, i.e. arboricolous ants are usually restricted to this habitat for nesting and often also for foraging. There are, however, some typical ground nesting genera also occurring in humus accumulations in the canopy (Longino and Nadkarni, 1990). Special adaptive characteristics of dominant ant species for canopy life were summarized by Tobin (1995) and Orivel and Dejean (1999a). These comprise morphological features protecting ants from desiccation (Hölldobler and Wilson, 1990; Tobin, 1995), special cognitive abilities for three-dimensional orientation (Hölldobler, 1980), and the use of a wider range of food sources in general and especially the use of plant sap as major food source (Tobin, 1994, 1995). The latter fact has consequences for the nitrogen economy of these ants1 because nitrogen is a highly limited resource in the canopy (Orivel and Dejean, 1999a). Though dominant species account for the major part of biomass and abundance in canopy habitats, many subdominant species, not sharing all the features proposed as special adaptations for arboricolous life style, account for the major part of ant diversity in this habitat (Orivel and Dejean, 1999a). The abundance of ants in tree crowns is surprising because ants, with their movable brood, were originally typical ground arthropods, specialized in building their nests in soil or using natural cavities in dead wood (Wilson, 1959; Sudd, 1967). Unlike bees with their comb-restricted brood, ants cannot produce waxy substances to build their own nests, and unlike wasps they usually do not have sticky gland products that enable them to construct nests from glued plant fibers. Therefore, at first glance, tree crowns seem not to be an ideal habitat for ants because sturdy and permanent nesting space – vital for social insects - is rather scarce. Dead twigs, hollow branches and stems as well as epiphytes provide the only natural – often short-lasting - cavities, suitable especially for 1 Modifications discussed in this context are: morphological modifications reducing the need for nitrogen; modification of the digestive system in order to process large amounts of liquid food; venom lacking nitrogen; maybe microbial endosymbionts and reduction of metapleural glands (Hölldobler and Wilson, 1990; Davidson and Patrell-Kim, 1996; Davidson, 1997; Orivel and Dejean, 1999a). 2