Pollination biology and inter-island geographical variation in the mutualistic Heliconia (Heliconiaceae)- hummingbird (Trochilidae) interaction of the Eastern Caribbean Islands Open Access
Downloadable ContentDownload PDF
Plants from the family Heliconiaceae (predominantly Neotropical family; 1 genus, ~200 species) are known to be exclusively pollinated by hummingbirds (Trochilidae, Stiles 1975; Kress 1983a). The adaptive interactions between heliconias and hummingbirds probably represent one of the well-documented cases of plant-pollinator mutualisms from the tropics. Although heliconias are most diverse in South America, the Caribbean Heliconia-hummingbird interactions (only two species: H. bihai and H. caribaea and 3-8 hummingbird species) provide a unique opportunity to investigate adaptation in plant-pollinator interactions in a natural set-up. Ecological studies investigating the mutualistic interactions between the Eastern Caribbean heliconias and their hummingbird pollinators initiated by Temeles et al. (2000) and Temeles and Kress (2003) have shown that the Heliconia-hummingbird interactions on the two islands of St. Lucia and Dominica show sexual specialization where each sex of the Purple-throated Carib hummingbird (Eulampis jugularis; hereafter PTC) is specialized to one of the two native heliconias. Based on the morphological associations between flower size of heliconias and bill size of hummingbirds and pollinator visitations, they further showed that trap-lining females were the primary pollinators of H. bihai, whereas territorial males were the primary visitors of H. caribaea (Temeles 2000, Temeles and Kress 2003, Gowda and Kress in prep). While several studies have investigated the behavior of the native Purple-throated Carib hummingbirds (refer to studies by Wolf, Feinsinger and Temeles) very little is known about the biology and ecology of the plants. In a mutualistic plant-pollinator hinteraction, fitness of the interacting plants can be affected by their breeding system and reproductive strategies. In this study I first examined the heliconias in their natural habitats and investigated their breeding system and pollination biology. I used a fluorescent-dye based approach to study pollination success in the two heliconia species because fruit-set was frustrated by the loss of experimental treatments in the field. Although presence of pollen tubes serves only as a more proximate yet still indirect measure of pollination success, it was observed that although self-compatibility is present (53-59%) in both heliconia species, the mean number of pollen tubes observed per style was significantly smaller in self versus out-crossed treatments (self = 0.52-1; outcross = 2.98-3.98). Outcrossing treatments also showed the highest pollination success (100%) in both heliconia species and autonomous pollination was generally rare (30-33%). I next investigated the effectiveness and importance of the two sexes of Purple-throated Carib hummingbirds as pollen vectors on H. bihai and H. caribaea. Pollinator effectiveness measures the total number of pollen grains deposited by a pollinator in a single visit while pollinator importance represents a simplified estimate of total pollen delivering capacity of the pollinator. Pollinator importance is a simple product of visitation frequency and pollinator effectiveness. As female Caribs were the sole pollinators of H. bihai, their pollinator effectiveness and importance estimates were indisputable. However, in H. caribaea male and female Caribs showed similar estimates for pollinator effectiveness (PE Male = 11.33 compare with PE Male = 11.67). However, male Caribs had higher values of pollinator importance than female Caribs due to their significantly higher visitation rates (PI male = 2.97 compare with PI female= 1.4). The second suite of factors that can affect a mutualistic plant-pollinator interaction includes the foraging behavior of the pollinator. Given the similar breeding system of the heliconias and the similar pollinator effectiveness of the male and female Caribs, I next developed and characterized microsatellite markers to investigate the effect of contrasting foraging behavior of male and female Caribs on gene flow patterns within each species. Our results of the genetic population structure of the two heliconias show that H. bihai indeed has a lower inbreeding rate than H. caribaea (FIS, bihai Vs. caribaea = 0.07 Vs. 0.15). This study therefore provides one of the first examples where foraging patterns of pollinators indeed produce the predicted effects on the gene flow patterns of the plants. In summary, the results show that territorial male Caribs do promote inbreeding within H. caribaea, while traplining female Caribs promote outcrossing within H. bihai. The third factor that can influence a plant-pollinator mutualism is the flowering phenology of the plant. Flowering phenologies of plants are suggested to have coevolved as a response to their pollinators and have been shown to affect Heliconia-hummingbird interactions. Our results on flowering phenologies show that H. bihai has a circum-annual flowering season with a distinct peak from March to June. This coincides with the peak but strictly-seasonal flowering pattern of H. caribaea, which flowers only from March to June. I further investigated the relationship between flowering phenologies and visitation rates of the two heliconia species. Despite contrasting flowering phenologies it can be concluded that both Heliconia species show specialization with their primary pollinators, male Purple-throated Caribs for H. caribaea and female PTC for H. bihai. However, temporal specialization patterns differ in the two heliconia species: H. bihai shows a phenologically-correlated specialization where a positive increase in flower availability is reinforced with higher visitation rate by the traplining female Caribs; in H. caribaea, increase in flower availability does not show a positive increase in visitation rates of the territorial male Caribs. However, the male Caribs also do not abandon their territories in times of least flower output and thus maintain almost a steady state of visitation rate through the flowering season of the plants. Finally, it is important to compare and contrast plant-pollinator mutualisms across a wide geographical scale to understand whether specializations are ubiquitous, and if not, does it reflect local pollinator influence. I therefore compared the same plant groups (the two heliconia species) across three different islands (North - St. Kitts, Central - Dominica, and South - St. Vincent) that varied in the distribution of the two native heliconias. Inter-island comparisons of plant-pollinator interactions suggest that on islands where H. caribaea are present, sex-specific pollinator specialization in H. bihai is present. On islands where H. caribaea is absent (such as St. Vincent), however, sexual specialization of H. bihai is lost and adaptation in nectar traits and pollinator response (in terms of visitation rates) can be observed. Results from this study suggests that hummingbirds are important selection agents for the presence of high sugar and volume of nectar in St. Vincent H. bihai. Although sexual specialization was not found in H. caribaea, species-specific specialization is prevalent. However, on St. Kitts and Dominica, variable H. caribaea-pollinator interactions were observed. In St. Kitts, the most generalized plant-pollinator interaction was observed where two additional species of hummingbirds (Green-throated Carib, Antillean Crested) visited H. caribaea across its temporal flowering season. In contrast, Dominica had the most species-specific plant-pollinator interaction where male and female Caribs were the dominant pollinators and Green-Throated and Antillean Crested hummingbirds could be occasionally observed only on non-defended clumps of H. caribaea. Results from this study provide insight into the plant's perspective of geographic variations in the Heliconia-hummingbird mutualism in the Caribbean Islands. This study is a first step towards extending our current knowledge of the breeding system and inbreeding rates of the plants, and the pollinator effectiveness and importance of the hummingbirds, onto other heliconias from the Caribbean Islands and mainland South America. The microsatellites developed in this study will be a key factor in addressing not only questions on inbreeding rates but also parentage analysis, phylogeography, colonization history and ultimately the evolution of the two Caribbean heliconia species.