<p>Although we often study mutualisms (interactions in which both species benefit) at the level of the individual partners, mutualistic interactions take place in the context of populations and communities. Sharing mutualists with others in a population could result in indirect interactions in the form of mutualist-mediated competition or facilitation. In my dissertation work I asked whether intraspecific competition or facilitation for ants might occur in an extrafloral nectary-bearing (EFN) plant, and what the consequences would be for long-term population dynamics of the plant. My focal species was <italic>Colubrina spinosa</italic> (Rhamnaceae), a neotropical treelet on which I observed 69 ant species at La Selva Biological Station, Costa Rica. </p><p> Demonstrating intraspecific competition for mutualists requires that 1) neighbor densities affect mutualist visits to an individual, and 2) change in mutualist visits results in reduced benefit. To determine how mutualist density affects plant benefit, I experimentally manipulated ant abundances on plants over two years and measured growth and survival. To assess competition for mutualists, I excluded ants from conspecific neighbors and followed ant abundance on focal plants. To consider long-term facilitation, in which greater local nectar resources increase local ant abundance, I manipulated nectar resources in a two-year field experiment and estimated ant abundance on <italic>C. spinosa</italic> plants and on baits. </p><p> Considering local neighbor density both within a 1m radius and in 5x5 m plots, ant densities on <italic>C. spinosa</italic> plants showed evidence for a small-scale competition effect and a contrasting plot-level facilitation effect. The small-scale competition was sized-based; smaller plants lost ants to larger plants. Ant benefit to plants also depended on plant size. For larger plants, those with greater size-adjusted ant density had higher growth and survival than those with fewer ants than expected for their size. </p><p> To determine whether these contrasting competition and facilitation effects could impact population growth or densities, I modeled population dynamics with an integral projection model (IPM). Growth and survival were functions of ant density, which in turn depended on conspecific neighbors, plant size, and mean background ants. Results suggest that larger-scale facilitation of mutualists impacts long-term population growth more than small-scale competition. Population growth rate increased with increasing background ant density, which depended on facilitation at the 5x5m plot scale. In contrast, small-scale competition caused a redistribution of mutualist ants among plants of different sizes, but had very little effect on long-term population growth. </p><p> I thus conclude that on the scale of individuals there is evidence of intraspecific competition for ants as well as facilitation in the EFN plant <italic>C. spinosa</italic>, but only facilitation effects lead to appreciable changes in population dynamics. If mutualist-mediated facilitation effects tend to occur over long time scales in other systems as well, facilitation might prove to be more important than competition in other mutualisms.</p> / Dissertation

Despite a long history of investigations on protective ant-plant interactions, since the late 19century (Thomas Belt 1874), a comprehensive quantitative understanding of the adaptations that facilitate these associations between plants and ants and the differential importance of these adaptations in predicting the benefits and costs to each partner and in shedding light on the evolutionary trajectories of this ecologically widespread interaction is yet to be realized. In the present study we have experimentally shown that the identity of the ant species (T. albipes), the abundance of this ant species and the composition of EFN produced by floral buds and young leaves of H. brunonis plant populations contribute to facilitating protection of floral bud inflorescences and young leaves of H. brunonis populations in a particular site. Of all the 16 ant species that inhabit the domatia of H. brunonis at several sites, the plant has found its protective ant partner only in the dolichoderine ant T. albipes and that too, only in the southernmost site In our experimental set up of three populations of H. brunonis at three different latitudes, where each site is separated from another site by at least 200 km, we have shown that only the southernmost population (Solaikolli) in protected by the ant species T. albipes. Although T. albipes is found in two of the three populations used in the present study (Sampaji and Solaikolli), it is present in a relatively larger abundance both within the domatia of H. brunonis and in the terrestrial stratum only in the southernmost site Solaikolli and is hence able to protect the EFN producing structures of H. brunonis at this site. In comparison, T. albipes is found in low abundance both in the domatia and in the terrestrial stratum in the site Sampaji and does not protect EFN producing plant parts of H. brunonis at this site. We have also experimentally shown that the young leaves of H. brunonis at this southernmost site Solaikolli are protected to a greater extent than the floral bud inflorescences, since T. albipes preferentially utilized EFN produced by young leaves compared to EFN produced by floral buds at this site. Moreover, we have also shown that the floral buds of H. brunonis in the northernmost site are unprotected for two reasons: i. the absence of the protective ant species T.albipes at this site. ii. The EPN produced by some of floral buds at this site is differentially utilized to a much lower extent than floral buds from other sites due to EFN from Agumbe being > 400-fold more viscous than the EFN produced by floral buds at the other two sites (sampaji and Solaikolli) (at 30º C). Although previous studies have independently shown that the identity of the ant partner (Janzen 1966, Schemske 1980, Horvitz and Schemske 1984, Heads 1986, Oliveira et al. 1987a, b; Jaffe et al. 1989, RicoGray and Thien 1989, Davidson et al. 1991) and the abundance of the protective ant species (Koptur 1984, Rocha and Bergallo 1992, Di Giusto et al. 2001) are important in predicting the protective outcome of the interaction between a plant species and its interacting ants, few studies have examined these two factors along with an examination of the EFN volume and composition (Inouye and Inouye 1980, Rudgers and Gardener 2004) and differential utilization of these qualitatively different EFNs produced by different populations of the same ant plant. Humboldtia brunonis and its interacting ant species provides a unique system to address questions about the evolution of ecological specialization and the evolution of preadapted plant traits that facilitate interactions between plants and ants (especially domatia), due to its unique polymorphism for the presence of caulinary domatia and its widespread abundance in the low-elevation wet-evergreen forests of the Western Ghats.

Extrafloral nectar (EFN) mediates food-for-protection mutualisms between plants and defensive insects. Senna mexicana var. chapmanii is a perennial legume native to the pine rockland habitats of south Florida. My dissertation focuses on how anthropogenic changes to the pine rocklands might affect EFN production by S. chapmanii, and the outcome of EFN mediated interactions. First, I investigated the influence of time of day, leaf damage, and leaf age on EFN production in S. chapmanii. Plants produced more nectar at night than during the day, and leaf damage resulted in increased EFN production. Furthermore, the response to leaf damage was greater when plants were damaged in the morning than when plants were damaged at night. Damage to young leaves elicited a stronger defensive response than damage to older leaves, in line with optimal defense theory. Second, I conducted a field experiment to determine the effects of ant activity, and light intensity, on herbivory rates, growth, and reproductive fitness in S. chapmanii. In shaded habitats, the presence of ants had no effect on herbivory rates, seed set, or plant size. In sunny habitats, however, plants with ants suffered less herbivore damage, produced more seeds, and grew larger over the duration of the one year study. Third, through a controlled greenhouse experiment I examined the effects of light intensity, and red/far-red light ratios, on EFN production in S. chapmanii. Plants in light-limited conditions produced less EFN, and leaf damage elicited increased EFN production regardless of light conditions. Ratios of red/far-red light, however, did not affect EFN production in either damaged or undamaged plants. Finally, I conducted a field study to determine how ants affect reproductive fitness in S. chapmanii. Over a period of eight months I observed the effects of ants on the activity of herbivores, predators, pollinators, and pre-dispersal seed predators. Relative pollinator efficiency, and rates of pre-dispersal seed predation, were unaffected by ants. Plants with ants, however, were quicker to establish, grew larger, and produced floral displays that attracted more pollinators. In S. chapmanii ants affected plant reproductive fitness simply by facilitating growth and establishment, with coincidental effects on reproductive investment.