Climatic Gradient

Floral phenotype and pollination system of a plant may be influenced by the abiotic environment and the local pollinator assemblage. This was investigated in seven plant-hummingbird assemblages on the West Indian islands of Grenada, Dominica and Puerto Rico. We report all hummingbird and insect pollinators of 49 hummingbirdpollinated plant species, as well as six quantitative and semi-quantitative floral characters that determine visitor restriction, attraction and reward. Using nonmetric multidimensional scaling analysis, we show that hummingbirdpollinated plants in the West Indies separate in floral phenotypic space into two gradients-one associated with the abiotic environment and another with hummingbird size. Plants pollinated by large, long-billed hummingbirds had flowers with long corolla tube, large amounts of nectar and showy orange-red colouration. These attracted few or no insect species, whereas plants pollinated by small, shortbilled hummingbirds were frequently pollinated by insects, particularly lepidopterans. The separation of plants related to environmental factors showed that species in the wet and cold highlands produced large amounts of dilute nectar, possessed no or a weak odour, and were associated with few insects, particularly few hymenopterans, compared to plants in the dry and warm lowlands. The most specialised hummingbird-pollinated plants are found in the West Indian highlands where they are pollinated by mainly large, long-billed hummingbirds. At the other extreme, highly generalised plants growing in the dry and warm lowlands are pollinated by small, short-billed hummingbirds and numerous insect species. This illustrates that, even within the hummingbird-pollinated flora, pollination syndrome and the degree of specialisation may vary tremendously depending on pollinator morphology and environment.

I want to especially thank all of the members of the Holt Lab, whom I have had the joy of working with over the past five years. I'm especially grateful to Dr. Jodie Holt, my wonderful advisor, for her welcoming spirit, her understanding, her advice on many experiments, grant applications, presentations and manuscripts as well as professional and career matters. I would also like to thank Doug Holt for technical assistance and for brightening up many lab lunches and parties. Dr. Rana Tayyar is a wonderful person and was incredibly helpful for many aspects of my research, not to mention many wise words v and lessons for life in general. Kai Palenscar was an unbelievable lab-mate, without whom much of this research would not have been possible, and I want to thank him for his practical support, advice and unflappably positive attitude during many long hours in the field and in the lab. I would also like to thank those lab members who either mentored or supported me early in my career in the lab including Polly Johnson, Dr.

Climate change is expected to alter the geographic distribution of many plant species worldwide. However, there is still no clear evidence showing a generalized direction and magnitude of these distribution shifts. Here, we have surveyed, in nine mountainous regions in Spain, an array of tree species along entire elevational ranges, as surrogates of their global climatic ranges, to test for elevational shifts towards cooler locations. We analysed the distribution recruitment patterns of five dominant tree species, recording the abundance and measuring the primary growth of juveniles in 306 plots. Three of the species have a temperate-boreal distribution with populations at their southern edge in the Mediterranean mountain ranges: Pinus sylvestris, P. uncinata and Fagus sylvatica; and the other two species have a Mediterranean distribution: Quercus ilex and P. nigra. Despite the contrasting phylogenies and biogeographies, we identified a similar pattern in recruitment abundance across species, with an asymmetric distribution of juveniles (more recruits in the middle-upper elevation of their range), but higher annual growths at lower elevations. This survival-growth trade-off at the early recruitment stage may potentially counterbalance at population level the negative effect of global warming on recruit survival at the lower edge of species ranges. These findings suggest a demographic stabilisation process at the early recruitment stage of these tree species, and highlight the importance of considering the different demographic stages across the whole climatic range to understand the effects that climate change may exert on species distributions and population dynamics.

Various aspects of the reproductive biology of Atamisquea emarginata Miers ex Hook. et Arn. were studied in a population from Tuc-uma´n, northern Argentina. Flowering took place in late (austral) spring to early summer. The creamy white, entomophilous flowers were protandrous and lasted 3.5-4.5 days. A pleasant odor along with nectar and pollen attracted 36 mostly diurnal insect species of which 41.7% were Hymenoptera, 38.9% Lepidoptera and 19.4% Diptera. However, 75% from all recorded species were infrequent, and domestic bees made ca. 94% of all visits. One halictid bee, one sphecid wasp, one myzinid wasp, four small-to medium-sized butterflies and a hawkmoth completed the set of relatively frequent insect visitors. Pollen and stigma were presented in the same place within the flower, but interference was prevented by a movement-assisted, complete dichogamy mechanism in which the bending ability of the gynophore played a fundamental role. This form of dichogamy was previously unknown in the Capparaceae. In comparison with the known species of the closely allied genus Capparis L., A. emarginata has diurnal, smaller, longer-lasting flowers, that offer much smaller nectar amounts, and that attract a rather generalized array of nonhovering visitors, while Capparis behaves as a more crepuscular-biased sphingophile or ornithophile or chiropterophile. Low levels of seed set under bagging indicates dependence on insect visits and are partly congruent with an earlier report of selfincompatibility in A. emarginata. Pollen tubes arrived to ovules in 90.6% of the inspected flowers. Fruit set of open-pollinated flowers was 43.6%-53.7%. Ninety-five percent of all ovules were regularly aborted in each fruiting ovary; thus, each fruit contained 1.2 developed seeds. Full-sized individuals were expected to produce ca. 10,700 potential offspring per year.

Understanding how climate and soil hydrology control tree growth is critical to predict the response of Siberian ecosystems to climate change. The general aim of this study was to (i) characterize the soil water budget and identify the factors controlling aspen (Populus tremula L.) radial growth in southwestern Siberia, and (ii) assess its potential response to future climate change. Along a gradient of climate and soil hydrological conditions, soil water budgets were reconstructed by modeling at four sites, and dendrochronological analyses were performed. Aspen growth potential was simulated in response to different climate change scenarios represented by shifts in soil water budgets. Simulated soil water budgets varied with climate variables, specifically increased temperature and drier summer combined with varying winter precipitation occurring as snowfall. We show that plant-available soil water and drainage gradually increased while stress decreased from the warmest and driest (south, forest-steppe zone) site to the coldest and wettest (north, southern taiga zone) site. Aspen radial growth was mainly limited by summer temperature in the north and by summer water deficit in the south. Surprisingly, we did not find clear evidence of snow level impact on radial growth, either positively in the south (water supply and protection against soil freezing) or negatively in the north (water-logging and drainage). In the context of climate change, water stress intensity could increase dramatically in the south inhibiting aspen growth; in those places summer soil water content depends on the refilling that occurs at snow-melt and increasing winter precipitation could alleviate stress levels. Conversely, in the north, aspen growth may mostly benefit from rising temperature.

Understanding how climate and soil hydrology control tree growth is critical to predict the response of Siberian ecosystems to climate change. The general aim of this study was to (i) characterize the soil water budget and identify the factors controlling aspen (Populus tremula L.) radial growth in southwestern Siberia, and (ii) assess its potential response to future climate change. Along a gradient of climate and soil hydrological conditions, soil water budgets were reconstructed by modeling at four sites, and dendrochronological analyses were performed. Aspen growth potential was simulated in response to different climate change scenarios represented by shifts in soil water budgets. Simulated soil water budgets varied with climate variables, specifically increased temperature and drier summer combined with varying winter precipitation occurring as snowfall. We show that plant-available soil water and drainage gradually increased while stress decreased from the warmest and driest (south, forest-steppe zone) site to the coldest and wettest (north, southern taiga zone) site. Aspen radial growth was mainly limited by summer temperature in the north and by summer water deficit in the south. Surprisingly, we did not find clear evidence of snow level impact on radial growth, either positively in the south (water supply and protection against soil freezing) or negatively in the north (water-logging and drainage). In the context of climate change, water stress intensity could increase dramatically in the south inhibiting aspen growth; in those places summer soil water content depends on the refilling that occurs at snow-melt and increasing winter precipitation could alleviate stress levels. Conversely, in the north, aspen growth may mostly benefit from rising temperature.

Evidence of tree regeneration failure of some species in the Iberian Peninsula forests warns us about the impact that the global change may exert on the preservation of Mediterranean forests, such as we know them. Predictions agree about an exacerbation of the summer drought there, acknowledged as the main limiting factor for the recruits' survival. On the other hand, many studies have also proved the relevant role that local heterogeneity has over the spatial distribution of forest species recruitment by providing safe sites. Therefore, to unravel how climate interacts with local factors over juveniles' performance seems crucial for the design of successful management strategies that allow facing the global warming. Here, we surveyed the natural recruitment of four dominant tree species in seven mountainous regions in the Iberian Peninsula, along entire elevational ranges as surrogates of their climatic ranges. Two of them have alpine and temperate distributions with populations at their rear edge in the Spanish mountains: Fagus sylvatica and Pinus uncinata; and the other two have a genuine Mediterranean distribution: Quercus ilex and P. nigra. Our main goal was to analyze for each species the effect of climate, local factors (i. e. light availability, stand structure and ground cover) and the interactions among them to identify the main drivers leading the regeneration process, assessed in terms of presence, abundance and mean annual growth of juveniles. The results showed different environmental factors determining the recruitment patterns of each species. Nevertheless, they highlighted the pervasive role exerted by both climate and fine scale factors, particularly the co-occurring vegetation on recruits' abundance, and the light availability on their growth. Moreover, we found some interactions among annual mean temperature and local factors, suggesting that climate and local heterogeneity act hierarchically, i. e. the local conditions may mitigate or exacerbate the impact of climate on juveniles. These results advocate for further research to increase our knowledge on the complex net of interactions among factors involved in recruitment at different scales, which in turn should be taken into account and incorporated in forthcoming management strategies.

Climate change is expected to alter the geographic distribution of many plant species worldwide. However, there is still no clear evidence showing a generalized direction and magnitude of these distribution shifts. Here, we have surveyed, in nine mountainous regions in Spain, an array of tree species along entire elevational ranges, as surrogates of their global climatic ranges, to test for elevational shifts towards cooler locations. We analysed the distribution recruitment patterns of five dominant tree species, recording the abundance and measuring the primary growth of juveniles in 306 plots. Three of the species have a temperate-boreal distribution with populations at their southern edge in the Mediterranean mountain ranges: Pinus sylvestris, P. uncinata and Fagus sylvatica; and the other two species have a Mediterranean distribution: Quercus ilex and P. nigra. Despite the contrasting phylogenies and biogeographies, we identified a similar pattern in recruitment abundance across species, with an asymmetric distribution of juveniles (more recruits in the middle-upper elevation of their range), but higher annual growths at lower elevations. This survival-growth trade-off at the early recruitment stage may potentially counterbalance at population level the negative effect of global warming on recruit survival at the lower edge of species ranges. These findings suggest a demographic stabilisation process at the early recruitment stage of these tree species, and highlight the importance of considering the different demographic stages across the whole climatic range to understand the effects that climate change may exert on species distributions and population dynamics.

Evidence of tree regeneration failure of some species in the Iberian Peninsula forests warns us about the impact that the global change may exert on the preservation of Mediterranean forests, such as we know them. Predictions agree about an exacerbation of the summer drought there, acknowledged as the main limiting factor for the recruits' survival. On the other hand, many studies have also proved the relevant role that local heterogeneity has over the spatial distribution of forest species recruitment by providing safe sites. Therefore, to unravel how climate interacts with local factors over juveniles' performance seems crucial for the design of successful management strategies that allow facing the global warming. Here, we surveyed the natural recruitment of four dominant tree species in seven mountainous regions in the Iberian Peninsula, along entire elevational ranges as surrogates of their climatic ranges. Two of them have alpine and temperate distributions with populations at their rear edge in the Spanish mountains: Fagus sylvatica and Pinus uncinata; and the other two have a genuine Mediterranean distribution: Quercus ilex and P. nigra. Our main goal was to analyze for each species the effect of climate, local factors (i. e. light availability, stand structure and ground cover) and the interactions among them to identify the main drivers leading the regeneration process, assessed in terms of presence, abundance and mean annual growth of juveniles. The results showed different environmental factors determining the recruitment patterns of each species. Nevertheless, they highlighted the pervasive role exerted by both climate and fine scale factors, particularly the co-occurring vegetation on recruits' abundance, and the light availability on their growth. Moreover, we found some interactions among annual mean temperature and local factors, suggesting that climate and local heterogeneity act hierarchically, i. e. the local conditions may mitigate or exacerbate the impact of climate on juveniles. These results advocate for further research to increase our knowledge on the complex net of interactions among factors involved in recruitment at different scales, which in turn should be taken into account and incorporated in forthcoming management strategies.

The belowground microbiota are directly affected by vegetation via resource inputs and indirectly affected through the effects on physico-chemical soil properties. This study examines the relationship between vascular plant community diversity and the functional diversity of soil bacteria. The total vascular plant diversity of trees, shrubs and forest floor vegetation was investigated along an altitudinal climatic gradient in the Polish Carpathians. Soil (upper 10 cm) was collected in forest stands at three elevations (600, 900, and 1200 m a.s.l.) on five separate mountain slopes (treated as independent replicates). The overall microbial activity was determined by the soil respiration rate, whereas the bacterial catabolic activity and functional diversity (H bact) was determined using Biolog ® ECO plates. The vegetation diversity (H plant) decreased with elevation. The organic matter content, water holding capacity and dissolved organic nitrogen differed significantly between elevations and were highest for stands at 1200 m a.s.l. Soil physico-chemical characteristics varied widely between sites and only showed an effect on the bacterial catabolic activity. Within microbial indices, only H bact was correlated with elevation (p < 0.03). Both bacterial catabolic activity and functional diversity were positively correlated with H plant (p < 0.005 and p < 0.02, respectively). Among the six chemical substrate groups on the Biolog ® ECO plates, the most variable chemical substrate group utilised by soil bacteria were amines; their utilisation was highest for soil bacteria originated from 600 m a.s.l. and lowest for bacteria from 1200 m a.s.l. Our results underline the close associations between above-and belowground biological diversity in forest ecosystems, which can emerge as a result of carbon and nitrogen cycling.

Plant Á/animal mutualistic interactions have a pervasive influence on community dynamics and diversity. Since Darwin (1859), biologists have been increasingly aware of the importance and complexity of pollination interactions in communities (Kearns and Inouye 1997,

Climate change is expected to alter the geographic distribution of many plant species worldwide. However, there is still no clear evidence showing a generalized direction and magnitude of these distribution shifts. Here, we have surveyed, in nine mountainous regions in Spain, an array of tree species along entire elevational ranges, as surrogates of their global climatic ranges, to test for elevational shifts towards cooler locations. We analysed the distribution recruitment patterns of five dominant tree species, recording the abundance and measuring the primary growth of juveniles in 306 plots. Three of the species have a temperate-boreal distribution with populations at their southern edge in the Mediterranean mountain ranges: Pinus sylvestris, P. uncinata and Fagus sylvatica; and the other two species have a Mediterranean distribution: Quercus ilex and P. nigra. Despite the contrasting phylogenies and biogeographies, we identified a similar pattern in recruitment abundance across species, with an asymmetric distribution of juveniles (more recruits in the middle-upper elevation of their range), but higher annual growths at lower elevations. This survival-growth trade-off at the early recruitment stage may potentially counterbalance at population level the negative effect of global warming on recruit survival at the lower edge of species ranges. These findings suggest a demographic stabilisation process at the early recruitment stage of these tree species, and highlight the importance of considering the different demographic stages across the whole climatic range to understand the effects that climate change may exert on species distributions and population dynamics.

Aim An understanding of the relationship between forest biomass and climate is needed to predict the impacts of climate change on carbon stores. Biomass patterns have been characterized at geographically or climatically restricted scales, making it unclear if biomass is limited by climate in any general way at continental to global scales. Using a dataset spanning multiple climatic regions we evaluate the generality of published biomass-climate correlations. We also combine metabolic theory and hydraulic limits to plant growth to first derive and then test predictions for how forest biomass should vary with maximum individual tree biomass and the ecosystem water deficit. Location Temperate forests and dry, moist and wet tropical forests across North, Central and South America. Methods A forest biomass model was derived from allometric functions and power-law size distributions. Biomass and climate were correlated using extensive forest plot (276 0.1-ha plots), wood density and climate datasets. Climate variables included mean annual temperature, annual precipitation, their ratio, precipitation of the driest quarter, potential and actual evapotranspiration, and the ecosystem water deficit. The water deficit uniquely summarizes water balance by integrating water inputs from precipitation with water losses due to solar energy. Results Climate generally explained little variation in forest biomass, and mixed support was found for published biomass-climate relationships. Our theory indicated that maximum individual biomass governs forest biomass and is constrained by water deficit. Indeed, forest biomass was tightly coupled to maximum individual biomass and the upper bound of maximum individual biomass declined steeply with water deficit. Water deficit similarly constrained the upper bound of forest biomass, with most forests below the constraint. Main conclusions The results suggest that: (1) biomass-climate models developed at restricted geographic/climatic scales may not hold at broader scales; (2) maximum individual biomass is strongly related to forest biomass, suggesting that process-based models should focus on maximum individual biomass; (3) the ecosystem water deficit constrains biomass, but realized biomass often falls below the constraint; such that (4) biomass is not strongly limited by climate in most forests so that forest biomass may not predictably respond to changes in mean climate.

Primary Biogenic Organic Aerosols (PBOA) were recently shown to be produced by only a few types of microorganisms, emitted by the surrounding vegetation in the case of a regionally homogeneous field site. This study presents the first comprehensive description of the structure and main sources of airborne microbial communities associated with temporal trends in Sugar Compounds (SC) concentrations of PM10 in 3 sites under a climatic gradient in France. By combining sugar chemistry and DNA Metabarcoding approaches, we intended to identify PM10-associated microbial communities and their main sources at three sampling-sites in France, under different climates, during the summer of 2018. This study accounted also for the interannual variability in summer airborne microbial community structure (bacteria and fungi only) associated with PM10-SC concentrations during a 2 consecutive years’ survey at one site. Our results showed that temporal changes in PM10-SC in the three sites are associat...

The importance of intraspecific trait variation (ITV) is increasingly acknowledged among plant ecologists. However, our understanding of what drives ITV between individual plants (ITVBI) at the population level is still limited. Contrasting theoretical hypotheses state that ITVBI can be either suppressed (stress-reduced plasticity hypothesis) or enhanced (stress-induced variability hypothesis) under high abiotic stress. Similarly, other hypotheses predict either suppressed (niche packing hypothesis) or enhanced ITVBI (individual variation hypothesis) under high niche packing in species rich communities. In this study we assess the relative effects of both abiotic and biotic niche effects on ITVBI of four functional traits (leaf area, specific leaf area, plant height and seed mass), for three herbaceous plant species across a 2300 km long gradient in Europe. The study species were the slow colonizing Anemone nemorosa, a species with intermediate colonization rates, Milium effusum, an...

This study presents the analysis of 381 phytosociological relevés describing predominantly ombrotrophic South Patagonian lowland peatland vegetation along a gradient of increasing continentality. Numerical methods such as cluster analysis and detrended correspondence analysis (DCA) were carried out to explore the data set. Cluster analysis resulted in nine vegetation types that were also distinctly separated in DCA ordination. The major floristic coenocline along the first DCA axis reflected a gradient of continentality ranging from pacific blanket bogs dominated by cushion plants to Sphagnum-dominated continental raised bogs. Increasing continentality along the first axis was parallel with decreasing peat decomposition and increasing peat depth and acidity. In contrast, floristic variation along the second DCA axis represented a water level gradient. The typical sequence of vegetation types along the hollow-hummock moisture gradient that is well established for north hemispherical peatlands could also be observed in Sphagnum-dominated South Patagonian raised bogs with a surprising similarity in floristic and structural features. Concerning the gradient of continentality significant differences in comparison with the northern hemisphere could be established. Most obvious was the dominance of cushion building plants (e.g. Astelia pumila, Donatia fascicularis) in South Patagonian oceanic peatlands, whereas this life form is totally absent from the northern hemisphere. Similar to the continental Sphagnum bogs the cushion plant vegetation of hyperoceanic peatlands exhibited a clear separation along the moisture gradient.

Aim An understanding of the relationship between forest biomass and climate is needed to predict the impacts of climate change on carbon stores. Biomass patterns have been characterized at geographically or climatically restricted scales, making it unclear if biomass is limited by climate in any general way at continental to global scales. Using a dataset spanning multiple climatic regions we evaluate the generality of published biomass-climate correlations. We also combine metabolic theory and hydraulic limits to plant growth to first derive and then test predictions for how forest biomass should vary with maximum individual tree biomass and the ecosystem water deficit. Location Temperate forests and dry, moist and wet tropical forests across North, Central and South America. Methods A forest biomass model was derived from allometric functions and power-law size distributions. Biomass and climate were correlated using extensive forest plot (276 0.1-ha plots), wood density and climate datasets. Climate variables included mean annual temperature, annual precipitation, their ratio, precipitation of the driest quarter, potential and actual evapotranspiration, and the ecosystem water deficit. The water deficit uniquely summarizes water balance by integrating water inputs from precipitation with water losses due to solar energy. Results Climate generally explained little variation in forest biomass, and mixed support was found for published biomass-climate relationships. Our theory indicated that maximum individual biomass governs forest biomass and is constrained by water deficit. Indeed, forest biomass was tightly coupled to maximum individual biomass and the upper bound of maximum individual biomass declined steeply with water deficit. Water deficit similarly constrained the upper bound of forest biomass, with most forests below the constraint. Main conclusions The results suggest that: (1) biomass-climate models developed at restricted geographic/climatic scales may not hold at broader scales; (2) maximum individual biomass is strongly related to forest biomass, suggesting that process-based models should focus on maximum individual biomass; (3) the ecosystem water deficit constrains biomass, but realized biomass often falls below the constraint; such that (4) biomass is not strongly limited by climate in most forests so that forest biomass may not predictably respond to changes in mean climate.

Current challenges to global food security require sustainable intensification of agriculture through initiatives that include more efficient use of nitrogen (N), increased protein self-sufficiency through homegrown crops, and reduced N losses to the environment. Such challenges were addressed in a continental-scale field experiment conducted over 3 years, in which the amount of total nitrogen yield (Ntot ) and the gain of N yield in mixtures as compared to grass monocultures (Ngainmix ) was quantified from four-species grass-legume stands with greatly varying legume proportions. Stands consisted of monocultures and mixtures of two N2 -fixing legumes and two nonfixing grasses. The amount of Ntot of mixtures was significantly greater (P ≤ 0.05) than that of grass monocultures at the majority of evaluated sites in all 3 years. Ntot and thus Ngainmix increased with increasing legume proportion up to one-third of legumes. With higher legume percentages, Ntot and Ngainmix did not continu...

The importance of intraspecific trait variation (ITV) is increasingly acknowledged among plant ecologists. However, our understanding of what drives ITV between individual plants (ITVBI) at the population level is still limited. Contrasting theoretical hypotheses state that ITVBI can be either suppressed (stress-reduced plasticity hypothesis) or enhanced (stress-induced variability hypothesis) under high abiotic stress. Similarly, other hypotheses predict either suppressed (niche packing hypothesis) or enhanced ITVBI (individual variation hypothesis) under high niche packing in species rich communities. In this study we assess the relative effects of both abiotic and biotic niche effects on ITVBI of four functional traits (leaf area, specific leaf area, plant height and seed mass), for three herbaceous plant species across a 2300 km long gradient in Europe. The study species were the slow colonizing Anemone nemorosa, a species with intermediate colonization rates, Milium effusum, an...

Aim An understanding of the relationship between forest biomass and climate is needed to predict the impacts of climate change on carbon stores. Biomass patterns have been characterized at geographically or climatically restricted scales, making it unclear if biomass is limited by climate in any general way at continental to global scales. Using a dataset spanning multiple climatic regions we evaluate the generality of published biomass-climate correlations. We also combine metabolic theory and hydraulic limits to plant growth to first derive and then test predictions for how forest biomass should vary with maximum individual tree biomass and the ecosystem water deficit. Location Temperate forests and dry, moist and wet tropical forests across North, Central and South America. Methods A forest biomass model was derived from allometric functions and power-law size distributions. Biomass and climate were correlated using extensive forest plot (276 0.1-ha plots), wood density and climate datasets. Climate variables included mean annual temperature, annual precipitation, their ratio, precipitation of the driest quarter, potential and actual evapotranspiration, and the ecosystem water deficit. The water deficit uniquely summarizes water balance by integrating water inputs from precipitation with water losses due to solar energy. Results Climate generally explained little variation in forest biomass, and mixed support was found for published biomass-climate relationships. Our theory indicated that maximum individual biomass governs forest biomass and is constrained by water deficit. Indeed, forest biomass was tightly coupled to maximum individual biomass and the upper bound of maximum individual biomass declined steeply with water deficit. Water deficit similarly constrained the upper bound of forest biomass, with most forests below the constraint. Main conclusions The results suggest that: (1) biomass-climate models developed at restricted geographic/climatic scales may not hold at broader scales; (2) maximum individual biomass is strongly related to forest biomass, suggesting that process-based models should focus on maximum individual biomass; (3) the ecosystem water deficit constrains biomass, but realized biomass often falls below the constraint; such that (4) biomass is not strongly limited by climate in most forests so that forest biomass may not predictably respond to changes in mean climate.

Plant carbon (C) and nitrogen (N) stoichiometry play an important role in the maintenance of ecosystem structure and function. To decipher the influence of changing environment on plant C and N stoichiometry at the subcontinental scale, we studied the shoot and root C and N stoichiometry in two widely distributed and dominant genera along a 2,200-km climatic gradient in China&#39;s grasslands. Relationships between C and N concentrations and soil climatic variables factors were studied. In contrast to previous theory, plant C concentration and C:N ratios in both shoots and roots increased with increasing soil fertility and decreased with increasing aridity. Relative N allocation shifted from soils to plants and from roots to shoots with increasing aridity. Changes in the C:N ratio were associated with changes in N concentration. Dynamics of plant C concentration and C:N ratios were mainly caused by biomass reallocation and a nutrient dilution effect in the plant-soil system. Our res...

Current challenges to global food security require sustainable intensification of agriculture through initiatives that include more efficient use of nitrogen (N), increased protein self-sufficiency through homegrown crops, and reduced N losses to the environment. Such challenges were addressed in a continental-scale field experiment conducted over 3 years, in which the amount of total nitrogen yield (Ntot ) and the gain of N yield in mixtures as compared to grass monocultures (Ngainmix ) was quantified from four-species grass-legume stands with greatly varying legume proportions. Stands consisted of monocultures and mixtures of two N2 -fixing legumes and two nonfixing grasses. The amount of Ntot of mixtures was significantly greater (P ≤ 0.05) than that of grass monocultures at the majority of evaluated sites in all 3 years. Ntot and thus Ngainmix increased with increasing legume proportion up to one-third of legumes. With higher legume percentages, Ntot and Ngainmix did not continu...

We describe the geographical variation of corolla and nectar guide size in seven populations of Mimulus luteus (Phrymaceae) in central Chile, and examine whether flower phenotypes associate with taxonomic composition and flower visit patterns of pollinators across populations. Flowers showed higher variation in nectar guide size than corolla size. Mean corolla size increased with the proportion of bees and decreased with the proportion of lepidopterans in the pollinator assemblages. Nectar guide size increased with the proportion of hummingbirds in the pollinator assemblages. When the frequency of flower visits rather than taxonomic composition was considered, the results revealed similar patterns. Because these traits previously have been described as targets of bee-and hummingbird-mediated selection in M. luteus, our results have implications for understanding the processes that determine flower diversification in Chilean Mimulus. Although we cannot rule out ecological sorting as an explanation for the geographical association between pollinators and flower phenotypes, changes in the prevalence and importance of bees and hummingbirds across populations appear to account, at least in part, for the flower phenotypic variation across populations. The extent to which insect and hummingbird pollination in M. luteus produces pollinator-mediated divergence among populations needs to be examined in future studies.

This study presents the analysis of 381 phytosociological relevés describing predominantly ombrotrophic South Patagonian lowland peatland vegetation along a gradient of increasing continentality. Numerical methods such as cluster analysis and detrended correspondence analysis (DCA) were carried out to explore the data set. Cluster analysis resulted in nine vegetation types that were also distinctly separated in DCA ordination. The major floristic coenocline along the first DCA axis reflected a gradient of continentality ranging from pacific blanket bogs dominated by cushion plants to Sphagnum-dominated continental raised bogs. Increasing continentality along the first axis was parallel with decreasing peat decomposition and increasing peat depth and acidity. In contrast, floristic variation along the second DCA axis represented a water level gradient. The typical sequence of vegetation types along the hollow-hummock moisture gradient that is well established for north hemispherical peatlands could also be observed in Sphagnum-dominated South Patagonian raised bogs with a surprising similarity in floristic and structural features. Concerning the gradient of continentality significant differences in comparison with the northern hemisphere could be established. Most obvious was the dominance of cushion building plants (e.g. Astelia pumila, Donatia fascicularis) in South Patagonian oceanic peatlands, whereas this life form is totally absent from the northern hemisphere. Similar to the continental Sphagnum bogs the cushion plant vegetation of hyperoceanic peatlands exhibited a clear separation along the moisture gradient.

Floral phenotype and pollination system of a plant may be influenced by the abiotic environment and the local pollinator assemblage. This was investigated in seven plant-hummingbird assemblages on the West Indian islands of Grenada, Dominica and Puerto Rico. We report all hummingbird and insect pollinators of 49 hummingbirdpollinated plant species, as well as six quantitative and semi-quantitative floral characters that determine visitor restriction, attraction and reward. Using nonmetric multidimensional scaling analysis, we show that hummingbirdpollinated plants in the West Indies separate in floral phenotypic space into two gradients-one associated with the abiotic environment and another with hummingbird size. Plants pollinated by large, long-billed hummingbirds had flowers with long corolla tube, large amounts of nectar and showy orange-red colouration. These attracted few or no insect species, whereas plants pollinated by small, shortbilled hummingbirds were frequently pollinated by insects, particularly lepidopterans. The separation of plants related to environmental factors showed that species in the wet and cold highlands produced large amounts of dilute nectar, possessed no or a weak odour, and were associated with few insects, particularly few hymenopterans, compared to plants in the dry and warm lowlands. The most specialised hummingbird-pollinated plants are found in the West Indian highlands where they are pollinated by mainly large, long-billed hummingbirds. At the other extreme, highly generalised plants growing in the dry and warm lowlands are pollinated by small, short-billed hummingbirds and numerous insect species. This illustrates that, even within the hummingbird-pollinated flora, pollination syndrome and the degree of specialisation may vary tremendously depending on pollinator morphology and environment.

We studied the effect of climate on the plant-pollinator communities in the West Indies. We constructed plots of 200 m × 5 m in two distinct habitats on the islands of Dominica, Grenada and Puerto Rico (total of six plots) and recorded visitors to all plant species in flower. In total we recorded 447 interactions among 144 plants and 226 pollinator species. Specifically we describe how rainfall and temperature affect proportional richness and importance of the different pollinator functional groups. We used three measures of pollinator importance: number of interactions, number of plant species visited and betweenness centrality. Overall rainfall explained most of the variation in pollinator richness and relative importance. Bird pollination tended to increase with rainfall, although not significantly, whereas insects were significantly negatively affected by rainfall. However, the response among insect groups was more complex; bees were strongly negatively affected by rainfall, whereas dipterans showed similar trends to birds. Bird, bee and dipteran variation along the climate gradient can be largely explained by their physiological capabilities to respond to rainfall and temperature, but the effect of climate on other insect pollinator groups was more obscure. This study contributes to the understanding of how climate may affect neotropical plant-pollinator communities.

We studied the effect of climate on the plant-pollinator communities in the West Indies. We constructed plots of 200 m × 5 m in two distinct habitats on the islands of Dominica, Grenada and Puerto Rico (total of six plots) and recorded visitors to all plant species in flower. In total we recorded 447 interactions among 144 plants and 226 pollinator species. Specifically we describe how rainfall and temperature affect proportional richness and importance of the different pollinator functional groups. We used three measures of pollinator importance: number of interactions, number of plant species visited and betweenness centrality. Overall rainfall explained most of the variation in pollinator richness and relative importance. Bird pollination tended to increase with rainfall, although not significantly, whereas insects were significantly negatively affected by rainfall. However, the response among insect groups was more complex; bees were strongly negatively affected by rainfall, whereas dipterans showed similar trends to birds. Bird, bee and dipteran variation along the climate gradient can be largely explained by their physiological capabilities to respond to rainfall and temperature, but the effect of climate on other insect pollinator groups was more obscure. This study contributes to the understanding of how climate may affect neotropical plant-pollinator communities.

Floral phenotype and pollination system of a plant may be influenced by the abiotic environment and the local pollinator assemblage. This was investigated in seven plant-hummingbird assemblages on the West Indian islands of Grenada, Dominica and Puerto Rico. We report all hummingbird and insect pollinators of 49 hummingbirdpollinated plant species, as well as six quantitative and semi-quantitative floral characters that determine visitor restriction, attraction and reward. Using nonmetric multidimensional scaling analysis, we show that hummingbirdpollinated plants in the West Indies separate in floral phenotypic space into two gradients-one associated with the abiotic environment and another with hummingbird size. Plants pollinated by large, long-billed hummingbirds had flowers with long corolla tube, large amounts of nectar and showy orange-red colouration. These attracted few or no insect species, whereas plants pollinated by small, shortbilled hummingbirds were frequently pollinated by insects, particularly lepidopterans. The separation of plants related to environmental factors showed that species in the wet and cold highlands produced large amounts of dilute nectar, possessed no or a weak odour, and were associated with few insects, particularly few hymenopterans, compared to plants in the dry and warm lowlands. The most specialised hummingbird-pollinated plants are found in the West Indian highlands where they are pollinated by mainly large, long-billed hummingbirds. At the other extreme, highly generalised plants growing in the dry and warm lowlands are pollinated by small, short-billed hummingbirds and numerous insect species. This illustrates that, even within the hummingbird-pollinated flora, pollination syndrome and the degree of specialisation may vary tremendously depending on pollinator morphology and environment.

system permits detailed comparison with the organizing parameters of other wellstudied groups and with predictions from general ecological theory. We concentrate on neotropical hum? mingbirds because they are the most intensively studied nectar-feeding birds and ap? parently the most dependent upon nectar. We summarize published information and previously unpublished data to (a) list the types of food resources distinguished by hummingbirds; <b) discuss the relation? ships among m)rphological features of hummingbirds .hat determine broadly which resource types they exploit; (c) link morphology with behavior and ecology to distinguish the community roles available to different hummingbirds; and (d) assemble food resources and hummingbird roles into natural communities. Finally, we compare the organization of hummingbird assemblages to that of other consumer groups.

Aim: We examined the effects of space, climate, phylogeny and species traits on module composition in a cross-biomes plant-hummingbird network. Location: Brazil, except Amazonian region. Methods: We compiled 31 local binary plant-hummingbird networks, combining them into one cross-biomes metanetwork. We conducted a modularity analysis and tested the relationship between species' module membership with traits, geographical location, climatic conditions and range sizes, employing random forest models. We fitted reduced models containing groups of related variables (climatic, spatial, phylogenetic, traits) and combinations of groups to partition the variance explained by these sets into unique and shared components.

The biodiversity of selected plant and animal groups was assessed along an altitudinal transect on the island of Dominica. Certain species of trees, herbaceous plants, vertebrates, and insects were surveyed at several sites from the Check Hall River to the top of Fifi and Mt. Joy trails at the Archbold Tropical Research Center, Springfield. In general, biodiversity decreased with the increase in altitude.

Geographic pattern of phenotypic variation can appear in a clinal or a mosaic fashion and can evidence adaptive or non-adaptive variation. To shed light on the mechanisms underlying this variation, we studied the relationships between geographic variation of floral traits and both biotic and abiotic factors of the hummingbird-pollinated plant, Nicotiana glauca, across its natural range. We obtained floral measures of 38 populations from an area about 1600 km long and 1050 km wide and an altitude range from 7 to over 3400 m. We used a MANOVA to detect between-population differentiations in flower traits and a DFA to determine the traits that best discriminate between populations. To test for associations between floral traits and climatic variables we used correlation analysis. We explored any possible distance-based pattern of variation (either geographic or altitudinal) in floral traits or bill length of pollinators using Mantel tests. Finally, we used a multiple regression to analyze simultaneously the effects and relative importance of abiotic predictor variables and bill length on corolla length. We found a high variation in flower traits among populations. Morphometric traits were the ones that best discriminated across populations. There was a clinal pattern of floral phenotypic variation explained by climatic factors. Differences in floral phenotypic distances were structured by altitudinal distances but not by geographic distances. Bill length of the hummingbird pollinators was structured both by altitudinal and geographic distances. Differences in bill length of hummingbird pollinators explained differences in corolla length across populations. Our findings support the assumption of flower evolution at a broad geographic scale. Floral traits seem to be structured not only by altitude but also by climatic factors.

Background: Despite the increasing knowledge of plant-pollinator interaction networks, the effects of human-induced disturbances on them have barely been studied. We analyzed whether land-use changes modified the structure and topology of plant-hummingbird interaction networks or promoted the integration of exotic plant species. Methods: Fieldwork was carried out in two vegetation areas in Mexico: a protected tropical dry forest and nearby disturbed sites. For two years we registered hummingbird-plant interactions monthly in each area. Then, we constructed interaction matrices from these data and compared their assemblage structure. Results: The conversion of original dry forest to disturbed habitats impacted some assemblage attributes of the plant-hummingbird network. In the disturbed sites, there were more plant species, mainly exotics, and one additional humming-bird species. Most network attributes remained the same except niche width and nestedness (pattern of interactions where generalists and specialists tend to interact with generalists whereas specialist-to-specialist interactions are infrequent), which were higher in the disturbed network. The generalist core in the disturbed network contained half of the core species in the conserved network. Implications for conservation Exotic plants that strongly integrated into the disturbed network may exert a large influence on network dynamics in these areas. Identifying the interacting species and their role provides valuable insights for their conservation and protection. Hummingbirds attracting native plant species have a potential for practical or ornamental use, and hummingbirds presence in human-modified landscapes not only provides positive aesthetic value to people but can additionally contribute to conserving native plants and the biodiversity associated with them.

As global surface temperatures rise, global precipitation rates are predicted to increase. These localised increases in rainfall patterns may significantly affect plant-pollinator interactions in multiple ways. Detrimental effects to plant-pollinator interactions could have significant ecological and economic consequences, and so it is important to understand the effects that rain has on these mutualisms. Increased rainfall has the potential for population-level effects but there also wide scope for individual-level effects, which have received surprisingly little attention. Changes in rainfall patterns could alter the timings of phenological phases while also increasing the likelihood of pollen degradation and nectar dilution, each having detrimental effects to the fitness of the plant, the pollinator or both parties. Pollinators could also be affected through mechanical and energetic constraints, along with disruption of foraging patterns and disruption to sensory signals. In this review, we demonstrate that there are clear gaps in our knowledge of these events, the exploration of which should open new areas of debate surrounding the effects of climate change on biological systems.

Observamos la respuesta de alimentación del colibrí (Lampornis amethystinus) en presencia de dos especies de avispas (Polistes sp. y Eumenes sp.) sobre un bebedero artificial colocado en condiciones naturales. El objeto fue examinar la influencia de las avispas sobre el consumo de "néctar" e investigar si la interacción bebedero-colibrí-avispa es un buen modelo para analizar los efectos indirectos y modificadores de interacciones mutualistas de un visitante floral no polinizador en la interacción mutualista planta-colibrí. La avispa Eumenes no influyó el comportamiento de alimentación de los colibríes, mientras que la avispa Polistes provocó que L. amethystinus no se alimentara en el 72% de las visitas registradas al bebedero. En ausencia de Polistes, los colibríes se alimentaron en el 91% de los casos. La posición y movimientos de Polistes actúan como un estímulo que desencadena un determinado patrón de comportamiento en los colibríes (intentar comer, alimentarse, huir, etc.). Sugerimos que una alta densidad de Polistes puede modificar indirectamente el resultado de la interacción directa planta-colibrí, ya que es posible que la presencia de Polistes represente un efecto negativo en el éxito reproductivo de las plantas que dependen de los colibríes para su polinización.

Pollinating insects are vital to the survival of many primary producers in terrestrial ecosystems, as up to 80–85 % of the world's flowering plants require pollinators for reproduction. Over the last few decades however, numerous pollinating insect populations have declined substantially. The causes of these declines are multifaceted and synergistic, and include pesticides, herbicides, mono-culture, urbanization, disease, parasites, and climate change. Here, we present evidence for a generally understudied yet potentially significant source of pollinator mortality, collisions with vehicles. Negative impacts from roads have been observed for the majority of vertebrate groups but studies of the effects on invertebrates have remained largely absent from the scientific literature. We documented road mortality of pollinating insects along a 2 km stretch of highway in Ontario, Canada and used our findings to extrapolate expected levels of road mortality across a number of landscape scales. Our extrapolations demonstrate the potential for loss of hundreds of thousands (on our studied highway) to hundreds of billions (generalised across North America) of Lepidopterans, Hymenopterans and pollinating Dipterans each summer. Our projections of such high levels of annual road mortality highlight the need for research to assess whether the mortality levels observed are contributing to the substantial declines of pollinating insects occurring on a global scale, thus putting the ecological functioning of natural areas and agricultural productivity in jeopardy.

Generalization prevails in flower-animal interactions, and although animal visitors are not equally effective pollinators, most interactions likely represent an important energy intake for the animal visitor. Hummingbirds are nectar-feeding specialists, and many tropical plants are specialized toward hummingbird-pollination. In spite of this, especially in dry and seasonal tropical habitats, hummingbirds may often rely on non-ornithophilous plants to meet their energy requirements. However, quantitative studies evaluating the relative importance of ornithophilous vs. nonornithophilous plants for hummingbirds in these areas are scarce. We here studied the availability and use of floral resources by hummingbirds in two different areas of the Cerrado, the seasonal savannas in Central Brazil. Roughly half the hummingbird visited plant species were nonornithophilous, and these contributed greatly to increase the overall nectar availability. We showed that mean nectar offer, at the transect scale, was the only parameter related to hummingbird visitation frequency, more so than nectar offer at single flowers and at the plant scale, or pollination syndrome. Centrality indices, calculated using hummingbird-plant networks, showed that ornithophilous and non-ornithophilous plants have similar importance for network cohesion. How this foraging behaviour affects reproduction of nonornithophilous plants remains largely unexplored and is probably case specific, however, we suggest that the additional energy provided by non-ornithophilous plants may facilitate reproduction of truly ornithophilous flowers by attracting and maintaining hummingbirds in the area. This may promote asymmetric hummingbird-plant associations, i.e., pollination depends on floral traits adapted to hummingbird morphology, but hummingbird visitation is determined more by the energetic "reward" than by pollination syndromes.

Aim To investigate the role of alien plants in mutualistic plant–hummingbird networks, assessing the importance of species traits, floral abundance and insu-larity on alien plant integration. Location Mainland and insular Americas. Methods We used species-level network indices to assess the role of alien plants in 21 quantitative plant–hummingbird networks where alien plants occur. We then evaluated whether plant traits, including previous adaptations to bird pollination, and insularity predict these network roles. Additionally, for a subset of networks for which floral abundance data were available, we tested whether this relates to network roles. Finally, we tested the association between hummingbird traits and the probability of interaction with alien plants across the networks. Results Within the 21 networks, we identified 32 alien plant species and 352 native plant species. On average, alien plant species attracted more humming-bird species (i.e. aliens had a higher degree) and had a higher proportion of interactions across their hummingbird visitors than native plants (i.e. aliens had a higher species strength). At the same time, an average alien plant was visited more exclusively by certain hummingbird species (i.e. had a higher level of complementary specialization). Large alien plants and those occurring on islands had more evenly distributed interactions, thereby acting as connectors. Other evaluated plant traits and floral abundance were unimportant predictors of network roles. Short-billed hummingbirds had higher probability of including alien plants in their interactions than long-billed species. Main conclusions Once incorporated into plant–hummingbird networks, alien plants appear strongly integrated once incorporated into plant–hummingbird networks and, thus, may have a large influence on network dynamics. Plant traits and floral abundance were generally poor predictors of how well alien species are integrated. Short-billed hummingbirds, often characterized as functionally generalized pollinators, facilitate the integration of alien plants. Our results show that plant–hummingbird networks are open for invasion.

Ecological communities that experience stable climate conditions have been speculated to preserve more specialized interspecific associations and have higher proportions of smaller ranged species (SRS). Thus, areas with dispropor-tionally large numbers of SRS are expected to coincide geographically with a high degree of community-level ecological specialization, but this suggestion remains poorly supported with empirical evidence. Here, we analysed data for hummingbird resource specialization, range size, contemporary climate, and Late Quaternary climate stability for 46 hummingbird –plant mutualistic networks distributed across the Americas, representing 130 hum-mingbird species (ca 40% of all hummingbird species). We demonstrate a positive relationship between the proportion of SRS of hummingbirds and community-level specialization, i.e. the division of the floral niche among coexisting hummingbird species. This relationship remained strong even when accounting for climate, furthermore , the effect of SRS on specialization was far stronger than the effect of specialization on SRS, suggesting that climate largely influences specialization through species' range-size dynamics. Irrespective of the exact mechanism involved, our results indicate that communities consisting of higher proportions of SRS may be vulnerable to disturbance not only because of their small geographical ranges, but also because of their high degree of specialization.

Background: In the past decade, accumulating evidence of pollinator decline has raised concerns regarding the functioning of terrestrial ecosystems and the sustainability of crop production. Although land-use changes have been advanced as the major causes, the affinities of most wild pollinators with the main land-use types remain unknown. Filling this gap in our knowledge is a prerequisite to improving conservation and management programmes.

The knowledge of relative performance of plants across environmental gradients is critical for their effective management and for understanding future range expansion. Pennisetum setaceum is an invasive perennial grass found along roadsides and other disturbed sites in South Africa. The performance of this grass in response to competition, habitat characteristics and resources was experimentally tested in three biomes (Karoo, Fynbos and Savanna) of South Africa. A total of 846 young P. setaceum seedlings were translocated to study sites in May 2007. The seedlings were grown in 94 plots along random transects, of which alternate halves were cleared of vegetation. Despite a variety of environmental hazards at these sites, over 30% of the transplanted seedlings survived over 15 months. Competition from resident vegetation was a major factor limiting the establishment of seedlings. However, under adequate rainfall and historical disturbance (mine dump), competition effects were overridden. Survival of seedlings was greatest in the Karoo National Park, possibly because of summer rainfall that occurred shortly after translocation. Despite differences in the survival and growth rates, seedlings remained alive at all sites, especially if they survived the first six months after translocation. P. setaceum is capable of persisting across a broad range of environmental conditions. Management efforts should aim to reduce seed production and establishment along roadsides that act as conduits into protected sites. This could be best achieved by maintaining as much indigenous cover along road verges as possible, as seeds survive best where competition is low.

† Background and Aims Convergent floral traits hypothesized as attracting particular pollinators are known as pollination syndromes. Floral diversity suggests that the Australian epacrid flora may be adapted to pollinator type. Currently there are empirical data on the pollination systems for 87 species (approx. 15 % of Australian epacrids). This provides an opportunity to test for pollination syndromes and their important morphological traits in an iconic element of the Australian flora. † Methods Data on epacrid-pollinator relationships were obtained from published literature and field observation. A multivariate approach was used to test whether epacrid floral attributes related to pollinator profiles. Statistical classification was then used to rank floral attributes according to their predictive value. Data sets excluding mixed pollination systems were used to test the predictive power of statistical classification to identify pollination models. † Key Results Floral attributes are correlated with bird, fly and bee pollination. Using floral attributes identified as correlating with pollinator type, bird pollination is classified with 86 % accuracy, red flowers being the most important predictor. Fly and bee pollination are classified with 78 and 69 % accuracy, but have a lack of individually important floral predictors. Excluding mixed pollination systems improved the accuracy of the prediction of both bee and fly pollination systems. † Conclusions Although most epacrids have generalized pollination systems, a correlation between bird pollination and red, long-tubed epacrids is found. Statistical classification highlights the relative importance of each floral attribute in relation to pollinator type and proves useful in classifying epacrids to bird, fly and bee pollination systems.

We describe the geographical variation of corolla and nectar guide size in seven populations of Mimulus luteus (Phrymaceae) in central Chile, and examine whether flower phenotypes associate with taxonomic composition and flower visit patterns of pollinators across populations. Flowers showed higher variation in nectar guide size than corolla size. Mean corolla size increased with the proportion of bees and decreased with the proportion of lepidopterans in the pollinator assemblages. Nectar guide size increased with the proportion of hummingbirds in the pollinator assemblages. When the frequency of flower visits rather than taxonomic composition was considered, the results revealed similar patterns. Because these traits previously have been described as targets of bee- and hummingbird-mediated selection in M. luteus, our results have implications for understanding the processes that determine flower diversification in Chilean Mimulus. Although we cannot rule out ecological sorting as an explanation for the geographical association between pollinators and flower phenotypes, changes in the prevalence and importance of bees and hummingbirds across populations appear to account, at least in part, for the flower phenotypic variation across populations. The extent to which insect and hummingbird pollination in M. luteus produces pollinator-mediated divergence among populations needs to be examined in future studies.

We describe the geographical variation of corolla and nectar guide size in seven populations of Mimulus luteus (Phrymaceae) in central Chile, and examine whether flower phenotypes associate with taxonomic composition and flower visit patterns of pollinators across populations. Flowers showed higher variation in nectar guide size than corolla size. Mean corolla size increased with the proportion of bees and decreased with the proportion of lepidopterans in the pollinator assemblages. Nectar guide size increased with the proportion of hummingbirds in the pollinator assemblages. When the frequency of flower visits rather than taxonomic composition was considered, the results revealed similar patterns. Because these traits previously have been described as targets of bee- and hummingbird-mediated selection in M. luteus, our results have implications for understanding the processes that determine flower diversification in Chilean Mimulus. Although we cannot rule out ecological sorting as an explanation for the geographical association between pollinators and flower phenotypes, changes in the prevalence and importance of bees and hummingbirds across populations appear to account, at least in part, for the flower phenotypic variation across populations. The extent to which insect and hummingbird pollination in M. luteus produces pollinator-mediated divergence among populations needs to be examined in future studies.

Floral phenotype and pollination system of a plant may be influenced by the abiotic environment and the local pollinator assemblage. This was investigated in seven plant–hummingbird assemblages on the West Indian islands of Grenada, Dominica and Puerto Rico. We report all hummingbird and insect pollinators of 49 hummingbird-pollinated plant species, as well as six quantitative and semi-quantitative floral characters that determine visitor restriction, attraction and reward. Using nonmetric multidimensional scaling analysis, we show that hummingbird-pollinated plants in the West Indies separate in floral phenotypic space into two gradients—one associated with the abiotic environment and another with hummingbird size. Plants pollinated by large, long-billed hummingbirds had flowers with long corolla tube, large amounts of nectar and showy orange-red colouration. These attracted few or no insect species, whereas plants pollinated by small, short-billed hummingbirds were frequently pollinated by insects, particularly lepidopterans. The separation of plants related to environmental factors showed that species in the wet and cold highlands produced large amounts of dilute nectar, possessed no or a weak odour, and were associated with few insects, particularly few hymenopterans, compared to plants in the dry and warm lowlands. The most specialised hummingbird-pollinated plants are found in the West Indian highlands where they are pollinated by mainly large, long-billed hummingbirds. At the other extreme, highly generalised plants growing in the dry and warm lowlands are pollinated by small, short-billed hummingbirds and numerous insect species. This illustrates that, even within the hummingbird-pollinated flora, pollination syndrome and the degree of specialisation may vary tremendously depending on pollinator morphology and environment.