Search

Several South American species of Iridaceae, especially those of Tigridieae, produce floral oils as rewards to oil-bee pollinators. The present study aimed to contribute to a deeper understanding of the reproductive biology, pollination ecology and level of specialization of the interactions of species encompassed in Tigridieae. Data on breeding and pollination systems were acquired from six species native to Southern Brazil. The visitation frequency and pollen load of pollen- and oil-collecting bees were also investigated. The results strongly suggest that the studied species are distributed along a specialization–generalization continuum. Three oil-producing taxa, Cypella herbertii, Cypella pusilla and Cypella amplimaculata, were pollinated effectively by oil-bees, whereas in the other two studied species, Kelissa brasiliensis and Herbertia pulchella, the oil-bees appeared to function as oil thieves, owing to failure to contact the plant reproductive parts during oil-foraging behaviour. New insights into aspects of the specialization–generalization continuum of pollination systems, differences in pollinator behaviour during oil and pollen foraging, and reproductive outputs of the studied species are provided. Taken together, our results provide a significant contribution towards a better understanding of reproductive biology and plant–pollinator interactions between Iridaceae and oil-collecting bees.

Understanding how disturbances influence interaction networks is a central but still poorly explored issue in ecology and management. The goal of this study was to test how the structure of plant–pollinator networks and the structuring processes are influenced by grazing in a subtropical grassland community on the southern hemisphere. Twelve sampling plots were allocated in order to cover a grazing gradient ranging from overgrazed to ungrazed sites. For each plot, we created a quantitative matrix containing all observed pairwise insect–plant interactions and described morphology, phenology and abundances of each species. We fitted a series of models to test the influence of grazing intensity on metrics describing networks structure. We finally used probabilistic matrices, maximum likelihood and model selection to investigate the processes influencing frequencies of interactions across the gradient of disturbance. Grazing intensity influenced connectance, specialization and interaction evenness, while the number of species and links, nestedness and modularity were less variable. Species abundance was the most important determinant of interaction frequencies regardless of grazing intensity. In contrast to northern hemisphere pollination networks studied so far, these subtropical plant–pollinator networks and their structuring processes were remarkably consistent along the grazing gradient. We argue that this results from the dominance of generalist Asteraceae species, which are selectively avoided by cattle and play a core role in attracting a wide range of pollinators and thereby structuring plant–pollinator interactions, providing therefore stability.

Synchrony of phenological events at the community level is a critical aspect for plants. Distinct phenological patterns, such as aggregated and segregated, act in the establishment of plant-pollinator interactions and can play a fundamental role in plant community structure. Working in grazed subtropical grasslands in southern Brazil, our goals were to describe the plant-pollinator interactions in the grassland community and investigate the phenological patterns. The study was performed in a grassland community of Southern Brazil. In 12 sample plots (100 m2 each) all flowering plant species with visually attractive flowers were recorded monthly during the period of six months. We used the Pianka index to estimate the flowering overlap, by comparison with 1000 pseudo-communities and a synchrony index to evaluate the phenological community pattern. All 82 flowering species present in the 12 study plots were observed monthly between six months. We observed that the community presented aggregated and non-synchronized flowering periods, which should reduce pollinator competition. We found no significant relationship between flowering synchrony and similarity of pollinators, that is, plant species that flower for a long time or together with many species did not share more pollinators than plants that flower for short time and alone. Some guilds of flowers visitors preferred flowers with a particular color, which indicated the important role of visual traits in plant-pollinator interactions. This study reports that flowering phenology could be structured by biotic components more than by random processes in these species-rich South Brazilian grasslands.

Diverse ecological factors (top-down and bottom-up) act on the richness and abundance patterns of insect galls. Host plant density is a kind of bottom-up factor and is related to “resource concentration hypothesis.” In this study, we analyzed forest patches that differ in vegetational structure and host plant abundance: homogeneous forest patches (Schinus-dominated) and more complex forest patches, to determine whether there are differences in the intensity of gall infestation among individual trees. Insect galls were sampled from 15 Schinus polygamus trees along five transects in each forest (150 trees in total). Two galler species, a psyllid leaf galler and cecidomyiid shoot galler, were recorded. We found that Psyllidae gall infestation in S. polygamus presents higher values of galls per individual tree in Schinus-dominated forest patches than forest patches that are more diverse in plant species. Cecidomyiidae galls did not differ between areas and were not related to host plant abundance. We did not observe any variation in gall infestation within individual. Our study demonstrated that density-dependence effect could be more important for some galls species than others. The results suggest that biological attributes of galling insects’ infestation could be associated with susceptibility to host plant density.

Offering floral rewards to pollinators such as nectar and pollen may not mean reproductive success. Some plant species present strategies to minimize this cost and dupe their pollinators. Begonia cucullata presents a mimicry system characterized by female rewardless flowers (mimic) that mimic male flowers (model) with a large amount of pollen. For the perfect functioning of mimetism, the proportion of mimics should not be higher than models. We tested different arrays of mimic-model flower proportions to test the efficacy of this system. We observed that bees avoided rewardless female flowers, although fruit set did not indicate pollen limitation., The proportion of mimics and models is a key parameter in mimetic systems. In monoecious plants with self-mimicry pollination systems, the mimic-model ratio is determined by the floral sex ratio. While an equal sex ratio (1:1) could provide the perfect balance between pollen donors and stigma surfaces able to receive the pollen, an unequal ratio could increase pollination by production of a greater number of rewarding, model flowers. The aim of the present study is to test the differences in visitation frequency and reproductive rates of different mimic and model flower arrays in order to assess the efficacy of the mimetic system in a Begonia cucullata population. The frequencies of visitors to groups of flowers with three distinctive sex ratio arrays (male-biased, female-biased and equal ratio) were compared using a Bayesian approach. The reproductive outcomes were compared in order to detect advantages of particular sex ratios. Low visitation frequency was recorded in all arrays. Pollinators showed similar behaviour regardless of sex ratio; they tended to avoid female, rewardless flowers. Pollination quality was highest in the equal sex ratio array. The current study shows that sex ratio plays a critical role in the pollination of B. cucullata and that the efficacy of the self-mimicry system appears to be doubtful. Visitation frequency may be associated with visual or chemical cues that allow pollinators to recognize models and mimics, regardless of their frequency in the population.

Floral traits are useful for understanding how natural selection has acted upon the evolution of plants over short time spans and in restrictive ecological scenarios. Furthermore, the spatio-temporal dynamic of inter and intraspecific interactions acts in the individual’s fitness and contributes to floral traits’ evolutionary history. Nicotiana alata presents large white flowers typical of hawkmoth pollination. A population of N. alata was monitored between 2011 and 2012, and we observed differential reproductive rates throughout the flowering period, with substantially higher values of fruit and seed set at the beginning than at the end of the flowering season. There was an indication of pollen limitation in the middle of the reproductive season, with difference in seed set between natural and hand pollination. Decreasing longevity over time was observed, as were different reproductive rates between distinctive floral life times, with higher values of fruit set in 5-day flowers (88.8 %). These results may indicate density-dependent effects in reproductive outputs and intraspecific interactions as one of the potential ecological factors impacting flowering time. Furthermore, although the hawkmoths are recognized for their unpredictability, the visits’ efficiency may contribute to adequate pollen loads in flowers with shorter longevity.