Plant-pollinator/seed-disperser, ant-plant, plant-herbivore, and model-mimic interactions give rise to many evolutionary (ultimate) and mechanistic (proximate) perspectives. The empirical studies on these questions (both in the laboratory and in the field) and theoretical modeling of the proximate mechanisms of herbivore foraging strategies that we are studying in my lab are summarized in this issue.

Philosophy and Approaches

Hypotheses regarding expected strategies of each player in the interaction are tested in our lab using the framework of the mechanisms involved in the interactions. The proximate mechanisms are investigated using a multi - disciplinary approach, using, for example, the tools of chemical analysis and mathematical and computer modeling, apart from behavioral studies.

The plurality of approach is also mirrored in the diversity of the systems. A brief summary of the major research projects is given in the next section.

Introduction

The Malabar Giant Squirrel, Ratufa indica, is an endangered, almost completely arboreal herbivore endemic to the closed canopy forests of India. Behavioural ecology of this animal is being studied by conducting extensive field studies, partly motivated by the conviction that a better understanding of the biology of this animal will serve to guide conservation policies.

Empirical Studies

Previous work in the field has led to extensive information on the preferences of the animal with respect to the biochemical compositions of its food resources. Samples were collected, by means of continuous focal animal sampling, high - resolution movement data of R. indica, along with behavioural observations of resource utilization. This is complemented by spatio-temporal data on availability of resources of differing quality. Empirically quantitating the spatial movement constraints that the animal has to contend with while foraging is being done. These constraints arise as a consequence of the animal being obligately arboreal, which means that almost all the movements of the animal are within the canopy of the forest, using the network of interconnected branches.

Theoretical Modelling

One of the primary motivations of the study is to elucidate the mechanisms by which R.indica  exercises its preferences for spatially distributed resources. As in most studies involving endangered animals, spatial movement data for R. indica  is more readily available than information on the behavioural mechanisms that guide the animal's movements. A methodology for inferring behavioural mechanisms from movement data, therefore, would be highly desirable. To this end, a model of animal movement was developed, which takes into account behavioural mechanisms that guide an animal's movements while predicting spatial movement patterns over time. Formerly, models used extensive computer simulations to predict the expected patterns of movement. Compared to this, now simple mathematical models are developed for greater reliability and transparency. Thus, for example, the movement pattern of an animal using an olfaction - guided navigation can be distinguished from that of an animal using compass direction - based navigation. Using these models, one can, in principle, infer some of the characteristics of the behavioural mechanisms from spatio-temporal movement data. Conversely, given the known behavioural mechanisms of movement, the probability of finding the animal at a point can be predicted. Refinement of the models is in progress so as to incorporate more realistic situations.

While these models are being developed with the objective of understanding R. indica ' s foraging movements, they can be used to interpret spatial movement data of any animal.

Theoretical representation of the territory of an individual squirrel as a movement space graph (Figure 1), with the circles representing the trees (nodes) and the lines (edges) representing the possibility of the animal moving from one tree to another in a single step. Presence on each node is considered as a 'state' of a Markov chain, with the transition matrix being determined by the behavioural mechanisms and the spatial constraints. Scale is in metres. Inset: The Malabar Giant Squirrel, Ratufa indica.

The Myrmecophyte Humboldtia brunonis (Fabaceae) and its Invertebrate Associates

Myrmecophytes, which are plants that house ants, have an extraordinarily diverse set of interactions with their ant partners. An interesting feature of these interactions is the high frequency with which these interactions are mutualistic, i.e., of benefit to both parties. Current focus is on the adaptive strategies that are responsible for the evolution and maintenance of the mutualisms, and their proximate manifestations. These questions in a system are being addressed, which shows tremendous variation at different spatial scales both in the degree of ant interaction as well as ant rewards provided by the plant, making it an ideal system for asking evolutionary questions.

The system under investigation: Humboldtia brunonis and its various associates

The association of H. brunonis  with a wide range of invertebrates is being studied and documented. Potential co-operations and conflicts are hidden within these interactions churning out a system ideal for investigation, rich in connections, exchanges and their respective consequences. The central hypothesis currently being tested / investigated is that H. brunonis  invests a part of its resources in recruiting ant species that form a line of defense against herbivory. From the proximate viewpoint, the resources are in (at least) two forms:

Domatia, which are stem modifications that provide nesting space to invertebrates. H. brunonis  is polymorphic for the presence of domatia.

Extrafloral nectar, which is produced by nectaries on the sepals of buds and young leaves.

Analyses of the chemical composition of the extrafloral nectar is done with the objective of quantifying the reward provided to the ants. This is complimented by behavioural studies on the ants to estimate the requirements and preferences towards the individual components of the nectar.

The system, apart from showing considerable within site variation, also displays between-site variation, with a gradation of increasing ant activity seen with decreasing latitude. The causes of this variation form another subject of investigation. Wherever it is found, H. brunonis  also happens to be a dominant understory tree. The hypothesis that this is largely due to vegetative propagation is being tested using molecular studies.

Domatia of Humboldtia brunonis	Domatia of Humboldtia brunonis
Extrafloral nectaries in Humboldtia brunonis	Ants on buds of H. brunonis

Distribution of Humboldtia brunonis in the Western Ghats

Wet Evergreen forests of the Western Ghats (low-elevation types, abundant in H. brunonis)       Dipterocarpus indicus Humboldtia brunonis Poeciloneuron indicum Dipterocarpus indicus Kingiodendron pinnatum Humboldtia brunonis

<>

Behavioral Ecology of Ant-Mimicking Spiders

Introduction

Myrmecomorphs are arthropods that have evolved a morphological resemblance to ants. Myrmecomorphy in spiders is considered to be a type of Batesian mimicry in which spiders gain protection from the predators through morphological and behavioural similarity to aggressive or unpalatable ants. Nine families of spiders have been recorded to exhibit mimicry, which include Clubionidae, Salticidae, Thomisidae and Therididae. Myrmecomorphs are found associated with a variety of terrestrial and arboreal ants of diverse habits in terms of nesting preferences, trail forming habits and feeding ecology.

The System(s) under Investigation:

The mechanisms and the adaptive significance of mimicry in several species of ant-mimicking spiders are being investigated. The current focus is on the Myrmecomorphs of the family Salticidae, occurring in and around Bangalore. A number of diverse ant species have mimics that are closely related, though morphologically and behaviourally dissimilar. For example, the Weaver Ant, Oecophylla smaragdina  has a mimic in the spider Myrmarachne plataleoides. The work also focuses on the intricate interactions between other ant species e.g. Camponotus and their spider mimics. The mimics are found in close proximity to their respective ant models.

Another interesting feature of myrmecophilic salticid spiders is their sexual dimorphism. The males have enlarged chelicerae which are also positioned parallel to the plane of the body unlike that of the female.

Questions Being Addressed

Given that the various morphs of spiders have evolved, from a common ancestor, under selection to mimic their respective models, the degree of morphological and behavioural similarity between each model - mimic pair is being investigated, as an estimate of adaptive change that a species can undergo under selection. Apart from behavioural studies, we are using the tools of digital image analysis towards this end. Along with predators, the spiders also need to avoid detection by the ant models. We are testing the hypothesis that the spiders achieve this by chemical mimicry, wherein the surface hydrocarbon profile, or the 'smell' of spiders is similar to that of their ant models. Chemical analysis (using mass spectrometry) is therefore being performed on these surface hydrocarbons, along with behavioural studies on the ants to estimate their ability to detect changes in these chemical 'signatures'. Field studies are being carried out to investigate the spatial proximity of each spider morph with its ant model.

Causes of sexual dimorphism in the system is another area of investigation. The dimorphism implies that the females are better mimics than the males. One of the interesting hypotheses we are pursuing is that the chelicerae of the males are objects of sexual selection, either as 'honest' signals of male quality by virtue of their being an impediment (the Handicap Principle - the 'good genes' hypothesis) or other, less adaptive reasons (the Fisherian 'good taste' hypothesis).

The Weaver Ant, Oecophylla smaragdina (Model)

Male Spider Myrmarachne plataleoides (Mimic)

Female Spider Myrmarachne plataleoides (Mimic)

Introduction

Interaction between fig wasps and their host plants (Ficus, Moraceae) range from mutualisms to parasitisms, and have served to test theories of co-evolution, sex allocation, kin selection, speciation and conflicts of resource management. All fig species are pollinated by mutualistic fig wasps (Agaonidae, Agaoninae) that are characterized by extreme host specificity possessing morphological adaptations and life cycles that depend entirely on host reproduction. However this mutualism is also exploited by a large number of non-pollinating wasps, which use figs as a breeding ground without offering pollination services. Elucidating the ultimate and proximate mechanisms involved in the maintenance of interactions between fig, fig -pollinator and parasite is one of the primary research goals.

Fig -Pollinator- Parasite Interactions:

Winged Female Wasp

Wingless Male Wasp

Non-Pollinating Wasp

Resembling a fruit, the fig is an enclosed inflorescence (syconium) containing many unisexual male and female flowers accessible to pollinators through an opening called the ostiole. Pollen carrying females are attracted by host specific fragrances released from receptive figs and lay eggs in a fraction of the flowers they pollinate. Their offspring feed on fig endosperm as larvae and mate in the fig cavity as adults. The inseminated female then leaves its natal fig with the pollen in search of other receptive figs.

A diverse assemblage of nonpollinating wasps and parasitoids also co-exist. Unlike the pollinators, fig-parasites oviposit within the flowers by piercing from the outside of the syconium with long ovipositors, laying their eggs in the ovaries but without the services of pollination, in turn competing with the pollinators for space and food.

As part of the current research interests, now common Indian Ficus species (F. racemosa, F. hispida, F. exasperata ) are being monitored with the objective of characterizing the identity and behavior of their respective pollinator and parasites.

Dioecy and Chemical Mimicry

In monoecious figs, all syconia in the tree are similar and produce both wasps and seeds. However, in dioecious figs two separate trees occur - male and female trees.

Male syconia contain male and short styled female flowers and produce many wasps, but few if any seeds. By contrast, female syconia contain only long styled female flowers and produce only seeds. While the entry of wasps into the female tree is beneficial to the tree it acts as a burial ground for the wasps. This generates a conflict of interest between the pollinator and the fig. Figs need agaonids to enter both types of syconium, but agaonids can only produce offspring in male syconia. From the perspective of the pollinator, therefore, the best strategy would involve discriminating between male and female figs, and avoiding the female figs altogether. From the fig's perspective, deceiving the pollinator into believing that the female fig is actually male would be the best strategy. The question, therefore, is whether there has been a selection on agaonids to discriminate between male and female syconia and on figs to prevent this, and what adaptations has each partner evolved so as to maximize its fitness.

The approach adopted to the above question is to investigate the mechanism by which the pollinators recognize and locate figs. Previous studies have shown figs release distinct chemical cues at different stages of their development, which are recognized by its pollinating wasps. These chemical cues act as attractants not only for the pollinators but also for frugivore animals, which help in dispersal of seeds. If the volatiles released by the male and female figs differ, then the wasps must be able to discriminate between them, while the figs should strive to minimize these differences.

Volatiles collected from receptive male and female trees were subjected to GC-MS analysis, so as to quantify differences in "chemical signatures". Behavioral studies are being performed on the wasps using olfactometers to test their ability to respond to differences in these signatures.