From March 2001 to December 2002, we studied the reproductive phenology, pollination ecology, and growth rates of Espeletia grandiflora Humb. and Bonpl. (Asteraceae), a giant caulescent rosette from the Páramos of the Eastern Andes of Colombia. Espeletia grandiflora was found to be predominantly allogamous and strongly self-incompatible. Bumblebees (Bombus rubicundus and B. funebris) were the major pollinators of E. grandiflora, although moths, hummingbirds, flies, and beetles also visited flowers. Inflorescence development began in March and continued through August to September. Plants flowered for 30 - 96 days with a peak from the beginning of October through November. The percentage of flowering plants strongly differed among size classes and between both years. Seed dispersal occurred as early as September through May of the following year. The average absolute growth rate for juveniles and adults rate was 7.6 cm/year. Given the scarcity of floral visitors at high altitudes due to climatic conditions, we suggest that even small contributions from a wide range of pollinators might be advantageous for pollination of E. grandiflora. Long-term studies on different populations of E. grandiflora are required to determine if the high growth rates are representative, to quantify the variation in the flowering behavior within and among populations, and to establish if nocturnal pollination is a trait that is exclusive to our population of E. grandiflora.

The tropical rain forests of northwest South America fall within the Chocó-Darien Global Ecoregion (CGE). The CGE is one of 25 global biodiversity hotspots prioritized for conservation due to its high biodiversity and endemism as well as threats due to deforestation. The analysis of land-use and land-cover (LULC) change within the CGE using remotely sensed imagery is challenging because this area is considered to be one of the rainiest places on the planet (hence high frequency of cloud cover). Furthermore, the availability of high-resolution remotely sensed data is low for developing countries before 2015. Using the Random Forest ensemble learning classification tree system, we developed annual LULC maps in the CGE from 2002 to 2015 using a time series of cloud-free MODIS vegetation index products. The MODIS imagery was processed through a Gaussian weighted filter to further correct for cloud pollution and matched to visual interpretations of land cover and land use from available high spatial resolution imagery (WorldView-2, Quick Bird, Ikonos and GeoEye-1). Validation of LULC maps resulted in a Kappa of 0.87 (Sd = 0.008). We detected a gradual replacement of forested areas with agriculture (mainly grassland planted to support livestock grazing), and secondary vegetation (agriculture reverting to forest) across the CGE. Forest loss was higher between 2010–2015 when compared to 2002–2010. LULC change trends, deforestation drivers, and reforestation transitions varied according to administrative organization (countries: Panamanian CGE, Colombian CGE, and Ecuadorian CGE).

Developing accurate methods to map vegetation structure in tropical forests is essential to protect their biodiversity and improve their carbon stock estimation. We integrated LIDAR (Light Detection and Ranging), multispectral and SAR (Synthetic Aperture Radar) data to improve the prediction and mapping of canopy height (CH) at high spatial resolution (30 m) in tropical forests in South America. We modeled and mapped CH estimated from aircraft LiDAR surveys as a ground reference, using annual metrics derived from multispectral and SAR satellite imagery in a dry forest, a moist forest, and a rainforest of tropical South America. We examined the effect of the three forest types, five regression algorithms, and three predictor groups on the modelling and mapping of CH. Our CH models reached errors ranging from 1.2–3.4 m in the dry forest and 5.1–7.4 m in the rainforest and explained variances from 94–60% in the dry forest and 58–12% in the rainforest. Our best models show higher accuracies than previous works in tropical forests. The average accuracy of the five regression algorithms decreased from dry forests (2.6 m +/− 0.7) to moist (5.7 m +/− 0.4) and rainforests (6.6 m +/− 0.7). Random Forest regressions produced the most accurate models in the three forest types (1.2 m +/− 0.05 in the dry, 4.9 m +/− 0.14 in the moist, and 5.5 m +/− 0.3 the rainforest). Model performance varied considerably across the three predictor groups. Our results are useful for CH spatial prediction when GEDI (Global Ecosystem Dynamics Investigation lidar) data become available.

The pink, tubular, nectariferous flowers of Melocactus intortus (Cactaceae) in Puerto Rico are visited by native hummingbirds (Anthracothorax dominicus), but also by invasive honeybees (Apis mellifera) and ants (Solenopsis sp.). We sought to determine if the bees and ants significantly alter the pollination of M. intortus by measuring pollinator effectiveness. Using traditional estimates of effectiveness (visitation rate and seed set), our results show that hummingbirds were the most effective pollinators as expected. Bees and ants were less effective, and their contributions were one-fourth to one-tenth of that observed for hummingbirds. We then modified this measure of effectiveness by adding two components, fitness of progeny and temporal availability of visitors, both of which refine estimates of flower visitor effectiveness. With these new estimations, we found that the effectiveness values of all three animal visitors decreased; however, the role of hummingbirds as the principal pollinator was maintained, whereas the effectiveness values of bees and ants approached zero. By these new measures of overall pollinator effectiveness, the invasive honeybees and ants have little effect on the reproductive success of M. intortus.

Tropical rain forests are suffering the highest deforestation and reforestation ever recorded. Interactions between direct (proximate or direct causes) and indirect (underling or indirect causes) drivers could cluster these forest cover changes forming hotspots (areas that exhibit significant spatial correlation of deforestation or reforestation transitions). Using land use–land cover maps and global (I) and local (Ii) Moran's tests, we identified these hotspots in the Chocó‐Darien Global Ecoregion (CGE) of South America, a natural region that was declared one of the top 25 hotspots for conservation priorities in the world. Subsequently, we tested and studied the effects and interactions between deforestation and reforestation hotspots and their direct and indirect drivers using Bayesian Structural Equation Modeling (Bayesian SEM). We found that deforestation and reforestation were spatially auto‐correlated forming hotspots (I = 0.49, P = 0.001 for deforestation transitions and I = 0.48, P = 0.001 for reforestation transitions). Also, hotspots of deforestation and reforestation were auto‐correlated within municipality borders (I = 0.5, P = 0.001 for deforestation transitions; I = 0.49, P = 0.001 for reforestation transitions). Eighteen municipalities located on the border between Colombia and Ecuador showed significant aggregations of deforestation hotspots, while thirty‐four municipalities in three areas of Colombia and the area between the Colombian and Ecuadorian border showed significant clustering of reforestation hotspots. Eleven of these municipalities presented significant clustering of both reforestation and deforestation hotspots. The Bayesian SEM for deforestation showed that population growth and road density were indirect drivers of deforestation hotspots (0.191 and 0.127 standard deviation units). The Bayesian SEM for reforestation found that armed conflicts, Gross Domestic Product, and average annual rain were indirect drivers related to reforestation hotspots (0.228, 0.076, and 0.081 standard deviation units, respectively). Our assessment shows a novel methodology to study interactions among direct and indirect drivers of forest change and their potential dissimilar effects on forest transitions.

Understanding spatial patterns of diversity in tropical forests is indispensable for their sustainable use and conservation. Recent studies have reported relationships between forest structure and α-diversity. While tree α-diversity is difficult to map via remote sensing, large-scale forest structure models are becoming more common, which would facilitate mapping the relationship between tree α-diversity and forest structure, contributing to our understanding of biogeographic patterns in the tropics. We developed a methodology to map tree α-diversity in tropical forest regions at 50 m spatial resolution using α-diversity estimates from forest inventories as response variables and forest structural metrics and environmental variables as predictors. To include forest structural metrics in our modelling, we first developed a method to map seven of these metrics integrating discrete light detection and ranging (LiDAR), multispectral, and synthetic aperture radar imagery (SAR). We evaluated this methodology in the Chocó region of Colombia, a tropical forest with high tree diversity and complex forest structure. The relative errors (REs) of the random forest models used to map the seven forest structural variables ranged from low (6%) to moderate (35%). The α-diversity maps had moderate RE; the maps of Simpson and Shannon diversity indices had the lowest RE (9% and 13%), followed by richness (17%), while Shannon and Simpson effective number of species indices had the highest RE, 27% and 47%, respectively. The highest concentrations of tree α-diversity are located along the Pacific Coast from the centre to the northwest of the Chocó Region and in non-flooded forest along the boundary between the Chocó region and the Andes. Our results reveal strong relationships between canopy structure and tree α-diversity, providing support for ecological theories that link structure to diversity via niche partitioning and environmental conditions. With modification, our methods could be applied to assess tree α-diversity of any tropical forest where tree α-diversity field observations coincident with LiDAR data.

Topography is a factor that can significantly affect the diversity and the distribution of trees species in tropical forests. Aniba perutilis, a timber species listed as vulnerable to extinction, is widely distributed in Andean forest fragments, especially in those with highly variable topography. Based on field surveys and logistic regression analyses, we studied the population structure and the effect of highly variable topography on the spatial distribution of this tree in three protected forest fragments in the central Andes of Colombia. Individuals of A. perutilis were mainly found on mountain ridges and hills with gentle slopes; no individuals were found in valleys. Using a species distribution model with presence/absence data, we showed that the available habitat for A. perutilis is significantly smaller than the extension of the fragments and much smaller than the extension of the currently protected areas. Our results have important implications for the conservation of A. perutilis and likely for other threatened Andean tree species, which can also have locally restricted distributions due to highly variable local topography.

No abstract