Plant distributions are changing at unprecedented rates, primarily due to habitat clearance and the spread of alien invasive species. Landscape pattern and local density can affect plant sexual processes, particularly those mediated by biotic vectors, by acting on the composition and behavior of pollinators and seed dispersers. Ecologists are now grappling with the likely effects of these altered processes on future forest composition as existing plant reproductive mutualisms break down or adjust to new spatial circumstances. Here, we introduce five papers that address pollinator responses and pollination outcomes in a variety of human-dominated landscapes and emphasize the need to better understand the dynamic nature of plant–pollinator interactions.

Forest fragmentation and the resulting spatial isolation of tree species can modify the activity of pollinators and may have important implications for the reproductive success and mating systems of the plants they pollinate. The objectives of this study were to (1) evaluate the effect of forest fragmentation on pollinator activity in bat-pollinated bombacaceous trees and (2) determine the effects of forest fragmentation on reproductive success and mating systems of bombacaceous trees. We studied these parameters in three bombacaceous tree species in tropical seasonal forest of Chamela, Jalisco, Mexico, and Osa and Guanacaste, Costa Rica. For Ceiba aesculifolia, more visits were observed in fragments by both Glossophaga soricina and Leptonycteris curasoae. For Ceiba grandiflora, Musonycteris harrisoni visited flowers exclusively in forest and G. soricina visited more flowers in forest than in fragments; no difference was shown by L. curasoae. For Ceiba pentandra in Chamela, no differences were found in visitation by G. soricina between forest and fragments; L. curasoae visited significantly more flowers in forest. Ceiba pentandra received more visits by Phyllostomus discolor than G. soricina in Guanacaste, whereas no bat visitors were observed in Osa. Total mean flower production was greater in fragments than forest for C. aesculifolia, whereas no difference was observed for C. grandiflora. Fruit set was greater in forest than in fragments for C. grandiflora, whereas no difference was observed for C. aesculifolia. Outcrossing rates were high for C. aesculifolia and C. grandiflora in Chamela, and for C. pentandra in Guanacaste, independent of tree habitat, while C. pentandra in Osa showed a mixed-mating system. The effects of forest fragmentation on bat pollinators, plant reproductive success, and mating patterns varied depending upon the bombacaceous species. This variability was associated with the effects that forest fragmentation may have on differences in flowering patterns, bat foraging behavior, and plant self-incompatibility systems.

Fruit set is pollen-limited in the self-incompatible tree Heterophragma quadriloculare (Bignoniaceae), pollinated by long-distance flying carpenter bees, and in the self-compatible shrub Lasiosiphon eriocephalus (Thymeleaceae), pollinated by weak-flying, sedentary beetles. We studied a single H. quadriloculare population over high and low flowering years and found no difference in bee visitation rates between these years. For H. quadriloculare, neighborhood floral display did not make a significant contribution to reproductive success. We investigated dense and sparse L. eriocephalus populations in the same year. In the low density L. eriocephalus population, individual floral displays were higher than in the dense population, yet reproductive success was low, indicating that plant isolation was a major factor influencing fruit set. This result was due to the extremely low number of beetles per plant and per flower in this population. In the dense L. eriocephalus population, although the displays of individual neighbors were smaller and plants were closer, neighborhood floral display did not contribute significantly to reproductive success, whereas the effect of individual floral display was ambiguous. Species with self-incompatible rather than self-compatible breeding systems are expected to experience neighborhood effects on reproductive success; however, at the spatial scale and floral display levels of plants in this study, only individual floral display affected fruit set in H. quadriloculare, whereas neither individual nor neighborhood display influenced fruit set in L. eriocephalus. Therefore, pollinator type, pollinator behavior, and plant and population isolation, rather than breeding system alone, will determine if neighborhood floral display affects fruit set.

The species-rich mist-belt grasslands of southern Africa have been severely reduced in extent as a result of commercial afforestation, thus confining many endemic plants and animals to small populations in habitat fragments. We investigated the influence of various environmental factors on seed production and seedling recruitment in remnant populations of the endangered grassland herb Gerbera aurantiaca (Asteraceae). Experiments with color traps showed that Eriesthis beetles, which appear to be the primary pollinators of G. aurantiaca, were most abundant in the two largest extant populations. Seeds are produced in a very small proportion (typically <10%) of the ca 80 female florets in a capitulum. The mean number of seeds produced in undamaged capitula was found to be significantly lower in small than in large populations. Pollen limitation was evident from a significant overall increase in seed set after supplemental pollination in three populations over two seasons. The proportion of capitula containing seed predators did not differ markedly among populations or years, but lepidopteran larvae, which destroy all of the seeds in a capitulum, were most abundant in the two largest populations in 2003. The presence of juvenile plants varied markedly among populations, but this could not be linked firmly to estimates of seed production. Clonal growth is likely to contribute to the persistence of small isolated populations of G. aurantiaca, even when seed production is severely compromised by pollen limitation and predation.

The indirect impacts of Shorea siamensis-logging on the reproductive ecology of Dipterocarpus obtusifolius, a self- incompatible butterfly-, moth-, and bird-pollinated tree, were studied in tropical dry forest in Thailand. Pollinator activity at D. obtusifolius trees and subsequent seed production were recorded in three forest areas subject to differing intensities of S. siamensis extraction. The pollinator and plant understory communities in these areas were also noted. Forest areas subject to high S. siamensis extraction intensities had very high understory flowering plant cover, dominated by the exotic invader Chromolaena odorata. Activity of butterfly pollinators at D. obtusifolius trees decreased in these disturbed areas, although their abundance remained comparable to other forest areas subject to only moderate or no extraction. For sphingid moth pollinators, there was no difference across differentially disturbed forest areas in either abundance or in the proportion bearing pollen. Pollinator activity by birds increased at highly disturbed locations but was not sufficient to offset a decline in overall pollinator activity at D. obtusifolius canopies in areas of heavy S. siamensis extraction. Thus, extraction of S. siamensis indirectly affected the pollination of D. obtusifolius, primarily by causing changes in the foraging behavior of butterfly pollinators rather than their abundance. A shift in the relative abundance of floral nectar resources from the canopy to the understory, a consequence of S. siamensis extraction and invasion by C. odorata, led to a parallel shift in foraging location of the principal diurnal pollinators, the butterflies, toward the understory. Despite reduced pollination at disturbed sites, behavioral changes did not translate into a D. obtusifolius seed set effect, possibly because pollination by birds (or moths) at the disturbed site compensated for reduced butterfly pollination.

Animals, especially insects, are principal pollen vectors of tropical trees and have behavior patterns that affect gene dispersal. Here, we explore complex pollination systems using a new simulation model Eco-Gene and considering, among other factors, flowering synchrony, spatial distribution of trees, degree of selfing, population densities, pollinator flight distances, pollen deposition, and pollinator response to floral display size. Sensitivity analyses using two contrasting tree data sets (Jacaranda copaia and Dipteryx odorata) determined the importance of each parameter on three response variables: the proportion of seeds from self-pollination, effective population size, and pollen dispersal. Spatial considerations and attractiveness of floral displays were prominent features determining the population genetic result of pollinators, and some biological implications of the results are discussed.

The five papers in this Special Section illustrate that pollination processes are often disrupted by changing plant distributions. They further demonstrate that plant pollination and seed production outcomes are highly complex due to multidimensional spatial effects and numerous interactive biological responses of both plants and pollinators at individual and population scales. Nevertheless, these studies also lend support to some emerging generalities that (through modeling) begin to provide a coherent framework by which further research, and ultimately management planning, could proceed.

The soil microbial biomass is largely responsible for the transformation of organic matter and the associated mineralization of important nutrients that regulate plant productivity. While the controls on soil microbial biomass and activity in temperate ecosystems have been well investigated, there is a paucity of such information from tropical rain forest ecosystems. Hence, we used a natural soil gradient (nutrient-poor oxisols and nutrient-rich mollisols) to investigate seasonal changes in the size and activity of the soil microbial community in two primary tropical rain forest sites in southwestern Costa Rica. In a nutrient-poor oxisol, microbial biomass ranged from 952 μg/g in the dry season (February) to 1967 μg/g in the wet season (June). The greater microbial biomass estimates corresponded to greater rates of microbial activity in the oxisol forest; however, the active fraction of the soil microbial biomass remained consistently low, regardless of soil C availability. In addition, while there was very little seasonal variability in microbial C:N ratios, we found extreme variations in seasonal microbial C:P ratios, with the highest microbial C:P ratio occurring when microbial biomass C was at a maximum. Our results suggest the importance of P availability in regulating soil microbial processes and that decomposition during the wet season (of litter accumulated in the previous dry season) may play an important role in driving seasonal changes in microbial biomass and activity.

We studied nectar characteristics in 12 species of Neotropical hummingbird-pollinated plants: Aphelandra golfodulcensis, A. sinclairiana, A. storkii, Justicia aurea (Acanthaceae); Heliconia irrasa, H. latispatha, H. mathiasii (Heliconiaceae); Pavonia dasypetala (Malvaceae); Isertia haenkeana, Palicourea guianense, Pentagonia macrophylla (Rubiaceae); and Renealmia cernua (Zingiberaceae). The diurnal patterns in sugar content and cumulative volume of nectar were documented in flowers that were protected from flower visitors. Sugar content in these species (mean of all flowers sampled per species ranged from 16 to 28%) was comparable to those reported for other hummingbird-pollinated species. Mean nectar volume per flower spanned an order of magnitude (i.e., 8.8 [Palicourea]–72.7 μl [H. latispatha]), indicating that hummingbirds visited flowers with a wide range of reward. We observed low intraspecific variation in sugar content but high variation in nectar volume. Among species, 10–50 percent of the variation in nectar volume was explained by time of day. For a subset of four species (A. sinclairiana, Isertia, Palicourea, and Pentagonia), we studied variation among individuals in nectar volume. With time of day controlled, individual explained as much as 70 percent of the variation in nectar volume (A. sinclairiana) and as little as 20 percent (Pentagonia). We suggest that nectar volume likely has a genetic component in these species, but also that variation may itself be a trait, especially in Palicourea and Pentagonia. In three of the four species for which we had data from individuals, mean nectar volume and standard deviation were not correlated, suggesting that these traits may vary independently (Pentagonia was the exception with mean and standard deviation positively correlated). For two species (Isertia and Palicourea), we found evidence of seasonal variation in nectar production, with more nectar per flower produced early in the flowering season than later.

We documented the diurnal pattern of nectar volumes in flowers of seven Neotropical hummingbird-pollinated plant species that were open to flower visitors and compared these standing crop data to nectar volumes in flowers protected from visitors. Standing crop of nectar bore little relationship to nectar in bagged flowers either in terms of volume or temporal pattern of availability. There was almost no nectar in open flowers except in those collected at dawn and those of Aphelandra sinclairiana; these were apparently too abundant to be depleted by visitors. Nectar volumes in open flowers were even more variable than in bagged flowers. We argue that understanding the evolution of plant– pollinator interactions requires understanding the relationship between nectar production and standing crop, as well as the impact of high variability on decisions by foraging animals. We conducted experiments to determine the effect of nectar removal on nectar production. Nectar removal via simulated legitimate visits had no impact on total production in flowers of Pavonia, Isertia, and Palicourea. For A. sinclairiana and Pentagonia, total nectar production in visited flowers was reduced compared to unvisited flowers. Data from individuals of these last two species indicated that the population-level pattern was assembled from individuals that responded differently to nectar removal; clearly, understanding the evolution of nectar traits demands data from individuals. We argue that detecting patterns of plant responses to nectar removal requires the context of floral characteristics (e.g., longevity, ovule number, or pollinators). Nectar removal via simulated nectar-robbing visits had no impact on total nectar production in A. sinclairiana but reduced total production in Isertia. Nectar robbing did not cause plants to invest more energy in nectar than they would have the absence of robbing. Finally, we found no difference in seed set by robbed and unrobbed flowers of A. sinclairiana. Our results add to a growing body of literature suggesting that nectar robbers are not always detrimental to plant fitness.

This study compared elephant use of woody vegetation on termite mounds with surrounding woodlands in western Zimbabwe. Twelve sites consisting of paired plots on termite mounds and in woodlands were selected. At each site, soil and vegetation samples (leaf and stem) were collected for chemical analysis. Both soil and plant samples were analyzed for calcium, magnesium, potassium, sodium, and phosphorus, and plant samples were also analyzed for crude protein concentration. Two indices of elephant feeding damage were computed: the median number of stems and branches removed per plant, and the mass of stems and branches removed by elephants per unit area. Termite mound soils had higher concentrations of all elements tested than soils from woodlands, and termite mounds differed from woodland plots in terms of plant species composition. Trees growing on termite mounds had higher concentrations of all nutrients except sodium and crude protein, and were subjected to more intense feeding by elephants than trees from the surrounding vegetation matrix. Termite mounds may play an important role in determining food availability and spatial feeding patterns by elephants and other herbivores.

We investigated the distribution patterns of leaf mining insects along an elevational gradient in cerrado vegetation of southeastern Brazil. We tested four hypotheses related to the distribution of mining insects: (1) the “altitudinal gradient hypothesis,” which predicts that mining insect species richness will decrease with altitude or elevation; (2) the “habitat- mediated richness hypothesis,” which predicts that mining insect species richness will be higher in mesic habitats than in xeric habitats; (3) the “plant species richness hypothesis,” which predicts that mining insect species richness will be positively correlated with plant species richness; and (4) the “plant architecture hypothesis,” which predicts a positive correlation between mining insect species richness and plant structural complexity. A total of 33,000 herbs, 3520 shrubs, and 1760 trees were sampled at 44 sites across an elevational gradient of 700 m. Mining insect species richness and plant species richness showed a negative correlation with elevation in xeric habitats, while in mesic habitats mining insect species and plant species richness did not show any statistically significant relationship with elevation. The differential distribution of mining insect species between xeric and mesic habitats supported the habitat-mediated richness hypothesis, which states that miners would be more speciose in mesic, more favorable habitats. Mining species richness also increased with increasing plant structural complexity. The results suggest that the mining habit may not represent a strong adaptive strategy in protecting mining insects against desiccation.

The Wood Stork (Mycteria americana) is a wading bird inhabiting subtropical and tropical regions of the American continent. This species is considered endangered in the United States. We compared variability and genetic structuring of nine Brazilian Pantanal subpopulations with an North American population using eight allozyme loci data (MPI, ICD, NSP, EST-D, LDH, PGM, 6PGD, and PEP-A) and four microsatellite loci data (WS1, WS2, WS4, and WS6). Average expected heterozygosity of Pantanal population was similar (0.198 ± 0.065) to that expected for the North American population (0.231 ± 0.066). No significant genetic differentiation was found among Pantanal subpopulations (F_st = 0.012) and low differentiation was detected between Pantanal and North American populations (F_st = 0.023). Lack of differentiation among Pantanal subpopulations may have been due to high gene flow level among birds of neighbor breeding colonies and low natal philopatry. We propose that low differentiation between North and South American populations has arisen either because these populations occupied neighboring regions during late glaciation or because there is a continuous gene flow between them, via Central American or northern South American populations.

We examined the effects of shrimp on insects in Puerto Rico by assessing insect assemblages in streams relative to varying shrimp densities and also by manipulating shrimp presence. Insect assemblages were typical of tropical island streams and insects were found in all stream habitats. Shrimp impacted insects directly via predation and competition and indirectly through removing particles from the water column. Variation in shrimp distribution is likely to be an important factor determining insect-assemblage structure in tropical streams.

Workers of Paraponera clavata, a common Neotropical ant, collect both nectar and insect prey. Previous reports show that nectar accounts for up to 90 percent of the ants' food loads, while calculations suggest that nectar contributes only 10 percent of colonies' energy supply. We assessed the trophic source of carbon and nitrogen in adult workers using stable isotope analysis. Carbon in adult workers was largely derived from plant sources. Worker nitrogen isotopic ratios varied significantly among colonies and were enriched compared to prey. Prey nitrogen isotope ratios suggest considerable intercolonial variation in diet, with some colonies collecting prey from lower trophic levels than other colonies. The importance of nectar as a source of metabolic carbon in adult worker biomass, coupled with the high frequency of nectar collection, supports the conclusion that omnivory is a key to supporting this species' biomass in Neotropical wet forests.

The co-extinction of interdependent species in relation to massive, long-term habitat disturbance has not been thoroughly investigated. Using logistic regression analyses, we examined the relationship between the loss of butterflies and their specific hostplants from the tropical island of Singapore and report the probable occurrence of their co- extinctions at the community level. Our simulation shows that the number of extinct butterfly species is expected to increase exponentially with that of extinct hostplants. The close association between butterflies and their hostplants suggests that the preservation of whole habitats is urgently needed if we are to avoid the possible cascading effects of species (co-)extinctions.