«Von der Mathematisch-Naturwissenschaftlichen Fakultät der Rheinisch-Westfälischen Technischen Hochschule Aachen zur Erlangung des akademischen ...»
Seed dispersal by ants
and its consequences for the phenology of plants
A study system for mutualistic animal-plant-interactions
Von der Mathematisch-Naturwissenschaftlichen Fakultät
der Rheinisch-Westfälischen Technischen Hochschule Aachen
zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften
Diplom–Biologe Reik Oberrath
Referent: PD Dr. rer. nat. K. Böhning-Gaese
Korreferent: Universitätsprofessor Dr. rer. nat. H. Wagner Tag der mündlichen Prüfung: 30.06.2000 Wenn weder Zahlen noch Figuren Sind Schlüssel aller Kreaturen, Wenn die, so singen oder küssen, Mehr als die Tiefgelehrten wissen, Wenn sich die Welt ins freie Leben, Und in die Welt wird zurück begeben, Wenn dann sich wieder Licht und Schatten Zu echter Klarheit werden gatten, Und man in Märchen und Gedichten Erkennt die wahren Weltgeschichten, Dann fliegt vor einem geheimen Wort Das ganze verkehrte Wesen fort.
(FRIEDRICH VON HARDENBERG,1772-1801) Nach Jahren quantitativer Forschung und zahllosen statistischen Berechnungen eine Erinnerung, daß Empirie nicht alles ist, das zählt.
1. General Introduction ____________________________________________________ 1
1.1 Community ecology and the meaning of phenology __________________________ 1
1.2 The comparative approach in community ecology ___________________________ 3
1.3 Study system and aim of thesis __________________________________________ 4
2. The Signed Mantel test to cope with autocorrelation in comparative analyses _________________________________________________ 6
2.1 Introduction _________________________________________________________ 6
2.2 The principle of the Mantel test __________________________________________ 8
2.3 Signed distances _____________________________________________________ 9 2.3.1 Limitation of absolute distances _______________________________ 10 2.3.2 Advantage of signed distances ________________________________ 10 2.3.3 The sign transformation _____________________________________ 12
2.4 Types of distance measures in trait variables ______________________________ 16 2.4.1 The use of categorical variables _______________________________ 17 2.4.2. The difference metric _______________________________________ 18 2.4.3 The quotient metric _________________________________________ 18
2.5 Validation Procedure for Mantel analyses _________________________________ 19 2.5.1 The validation criterion ______________________________________ 19 2.5.2 Numerical examples: Hypothetical data _________________________ 20 2.5.3 Numerical examples: Ecological field data _______________________ 21 2.5.4 Validation of autocorrelation effects ____________________________ 22
2.6 The computer program _______________________________________________ 23 2.6.1 Implementation ____________________________________________ 23 2.6.2 Variants of the Signed Mantel test _____________________________ 24 2.6.3 Graphical data exploration ___________________________________ 25
2.7 Advantages and limitations ____________________________________________ 25
2.8 Conclusions ________________________________________________________ 26
2.9 Summary __________________________________________________________ 27
3. Effects of pollination and seed dispersal mode on the reproductive phenology of a temperate plant community _______________28
3.1 Introduction _________________________________________________________28
3.2 Methods____________________________________________________________29 3.2.1 Study area and plant species __________________________________29 3.2.2 Phenological variables _______________________________________30 3.2.3 Biotic influences ____________________________________________30 3.2.4 Abiotic influences ___________________________________________31 3.2.5 Statistics __________________________________________________32
3.3 Results_____________________________________________________________34 3.3.1 Biotic influences ____________________________________________34 3.3.2 Abiotic influences ___________________________________________39
3.4 Discussion __________________________________________________________40
3.5 Summary ___________________________________________________________44
4. Phenological adaptation of ant dispersed plants to the seasonal variation in ant activity ____________________________________46
4.1 Introduction _________________________________________________________46
4.2 Methods____________________________________________________________48 4.2.1 Study area and plant community _______________________________48 4.2.2 Plant phenology ____________________________________________48 4.2.3 Plant attributes _____________________________________________48 4.2.4 Plant phylogeny ____________________________________________49 4.2.5 Seasonal variation of ant activity _______________________________50
4.3 Results_____________________________________________________________52 4.3.1 Plant phenology ____________________________________________52 4.3.2 Seasonal variation in ant activity________________________________56 4.3.3 Test on phenological adaptation________________________________58
4.4 Discussion __________________________________________________________58
4.5 Summary ___________________________________________________________62
5. General Conclusions___________________________________________________ 64
6. References ___________________________________________________________ 65
7. Appendix_____________________________________________________________ 76
8. Acknowledgment ______________________________________________________ 81
1.1 Community ecology and the meaning of phenology Ecology is the study of interactions between organisms and their biotic as well as abiotic environment. Ecological studies can be performed on the individual, population and community level. In the last case, which is my area of research, the aim is to find and understand ecological patterns and mechanisms that concern interactions between organisms of different species (RICKLEFS 1990, BEGON et al. 1996). Basically, the interactions can be positive, neutral or negative for any of the interaction partners. Interactions that are positive for both partners are called mutualistic. Mutualistic interactions play an important role in nature mainly for two reasons. First, they are very common in natural communities, second, the reproduction and hence the survival of species depend - sometimes critically - on the benefit of mutualists (FEINSINGER 1983, HOWE & WESTLEY 1988, PIANKA 1994, BEGON et al. 1996). In my opinion, the study of mutualistic animal-plant-interactions is one of the most fascinating fields in community ecology.
In temperate regions two types of mutualistic animal-plant-interactions are both widespread and eye-catching: pollination and seed dispersal by animals. While animals gain food from these interactions, plants benefit from gene flow (BEATTIE 1978). With pollination, animals transfer male gametes to other plant individuals of the same species, therefore being necessary for the sexual reproduction of many plant species. Cross fertilization between plants of the same and, more rarely, of different subpopulations is assumed to be of major importance for the reproduction and fitness of plants (FAEGRI & VAN DER PIJL 1976, HOWE & WESTLEY 1988, BOND 1995). With seed dispersal, animals disperse plant embryos. Successful seed dispersal can contribute to the avoidance of predators, pathogens and competition, to increased genetic heterogeneity and successful cross-fertilization among plants of the following generations, to the spatial expansion of the plant population, and to the establishment of new plant populations (HOWE & SMALLWOOD 1982, HOWE 1986, HOWE & WESTLEY 1988, BOND 1995, BONN & POSCHLOD 1998). Besides seedling establishment, pollination and seed dispersal appear to be the dominant factors influencing reproduction in plants (FENNER 1985).
Given the importance of pollination and seed dispersal for plant reproduction, one can assume that plants have evolved special attributes and adaptations that increase the probability
1. GENERAL INTRODUCTIONfor successful interactions (FEINSINGER 1983, FENNER 1985, HOWE & WESTLEY 1988). In fact, many plant attributes are discussed as adaptations to animal pollinators and dispersers. With respect to pollination, adaptive traits often appear to be tuned to particular animal species and correspondingly reflect the pollinator spectrum of a plant species. A group of different plant attributes, e.g. form, color and odor of flowers, represents the so-called pollination syndrome (HOWE & WESTLEY 1988). An example for the pollination syndrome of plants pollinated by nocturnal moths is the concurrence of nectar producing, deep flowers with bright colors and intensive odors. Corresponding syndromes are known for the seed dispersal interactions. For example, fleshy intensely colored fruits in exposed positions are typical for bird dispersed plant species (HOWE & WESTLEY 1988). In comparison to pollination, seed dispersal by animals appears to be more diffuse (JANZEN 1983, HERRERA 1985, HOWE & WESTLEY 1988, HANDEL & BEATTIE 1990), indicated by a lack of highly specialized, tight animal-plant-interactions and by species attributes which can be less obviously explained by adaptations.
Morphological adaptations, e.g. the form of flowers and fruits, as well as physiological attributes, e.g. the production of nutrients as reward for pollinators and dispersers, have been intensely studied for many decades. For seed dispersal by ants, for example, it has been known since 1873 by observations of J. F. Moggridge (see SERNANDER 1906) that some plant species have seeds with food bodies attached which represent nutrients for ants. Phenological attributes of plants, however, have been considered only for the last few decades. The phenology, the development of organisms during one season, appears to be important because selection is assumed to favor species that avoid unfavorable times (FENNER 1998). Plant phenology includes vegetative growth, flowering, fruit ripening and fruiting. As reproductively essential periods, flowering and fruiting phenology appear to be most important. Low availability of pollinators and dispersers, high competition for resources between plant species, and intensive predation on flowers, seeds or fruits are assumed as selection forces which influence and shape flowering and fruiting times (RATHCKE & LACEY 1985, RATHCKE 1988, FENNER 1998). Thus, some plant species may flower or fruit earlier or later in the season in order to reduce interspecific competition and predation or to take advantage of higher pollinator or disperses activities. In pollination ecology some evidence exists for the adaptive nature of plant phenology, in seed dispersal ecology evidence for the adaptive significance is rare and mostly appears to be speculative (see FENNER 1998).
1.2 The comparative approach in community ecology Most of the earlier ecological studies investigated animal-plant-interactions between a single plant species and their pollinators or dispersers. These studies revealed detailed aspects for the plant species studied, but their results could at best be partly generalized. Recently, more studies investigate a group of plant species defined by their phylogenetic relationship (SMITH-RAMIREZ et al. 1998), their habitat (MURALI & SUKUMAR 1994), or their life history strategy (RATHCKE 1988). Such comparative studies which consider many different species are able to identify more universal patterns. In general, the more species are comparatively investigated in the same study the broader the view one gets from the results on possible overall patterns and relationships (BROWN 1995).
A fundamental problem arises with comparative data, e.g. the fruiting times of a list of plant species, and statistical standard procedures to test possible relationships on significance.