The upper timberline is a conspicuous natural vegetation limit in many upper montane settings, but the causes determining its elevation and structure are still poorly understood. Modern forest distributions and particular characteristics such as species distributions, growth habits and fitness, need to be seen in the context of historical human land use, natural disturbances and site conditions (Foster et al., ABSTRACTAim To determine whether the cover and growth habit of the main forest forming species (Polylepis australis BITT.) in a mountain range with low human population density is mainly affected by anthropogenic activities or by environmental influences. Location Central Argentina.Methods Using GIS and field surveys we established 146 plots of 30 · 30 m located in five river basins differing in human impact. We measured P. australis cover, growth habit of each individual (number of basal ramifications), index of long term human impact (percentage of rock exposed by soil erosion due to livestock and fires), evidence of logging, fire scars, local relief, percentage of rock outcrops and altitude above sea level. We analysed the influence of independent variables on P. australis cover and growth habit (average number of basal ramifications per plot) using correlations and General Linear Models.Results Polylepis australis cover was greater at intermediate altitudes above sea level and in areas with reduced long term human impact. Contrastingly local relief, percentage of rock outcrops and logging in the recent past did not have a major influence on P. australis abundance. Growth habit varied in complex patterns. Individuals with fewer ramifications were found in valley bottoms and more disturbed basins, while more ramifications were found at mid-and upper slopes and well preserved basins. In valley bottoms, ramifications decreased with increasing altitude whereas the opposite trend was observed for mid-slopes. Ramifications were positively related to fires in two river basins and in mid-and upper slopes but not in valley bottoms. Fire impact was always less in valley bottoms than at mid-and upper slopes.Main conclusions Human impact had a major role on P. australis cover, while growth habit was determined by complex combinations of potentially cumulative natural and anthropogenic factors. Even in sparsely populated mountains, both human impact and their interaction with natural environmental gradients influence plant communities and need to be understood for effective management.

Interactions between resource and consumer species are organized in ecological networks. Species interactions in these networks are influenced by the functional traits of the interacting partners, but the generality of trait-based interaction rules and the relationship between functional traits and a species' specialization on specific interaction partners are not yet understood. Here we combine data on eight interaction networks between fleshy-fruited plants and frugivorous birds sampled across the tropical and subtropical Andean range. We test which combinations of morphological plant and animal traits determine trait matching between resource and consumer species in these networks. In addition, we test which of the morphological traits influence functional specialization of plant and bird species. In a meta-analysis across network-specific fourth-corner analyses, we found that plant-animal trait pairs related to size matching (fruit size-beak size) and avian foraging behavior (plant height-wing shape and crop mass-body mass) were positively related in these networks. The degree of functional specialization on specific interaction partners was positively related to crop mass in plants and to the pointedness of the wing in birds. Our findings show that morphological trait matching between fleshy-fruited plants and frugivorous birds is a general phenomenon in plant-frugivore networks across the Andes and that specific plant and bird traits can be used to approximate the degree of functional specialization. These insights into the generality of interaction rules are the base for predictions of species interactions in ecological networks, for instance in novel communities in the future, and can be applied to identify plant and animal species that fulfill specialized functional roles in ecological communities.

No abstract

Altitudinal gradients are expected to heavily influence the general performance of mountain tree species. For this study we evaluated vitality, radial growth and reproductive efficiency in Polylepis australis (Rosaceae) throughout its complete altitudinal range in central Argentina. We selected seven forest fragments ranging from 900 to 2700 m a.s.l. In the field, we subjectively assigned vitality and seed productivity indexes to 12 P. australis trees per fragment. In addition, we cored and assessed for radial growth 10–18 trees per altitudinal level. In the laboratory, we individually weighed seeds from four or five trees from each altitudinal level and divided into two seed mass classes (light and heavy). We then tested for germination and monitored their seedlings for survival. Seedlings of three trees per altitudinal level were harvested after 40 days and assessed for dry mass. Seedlings of one to two trees per altitudinal level were monitored for survival until both leaves and roots had died. Our main results showed two types of responses to altitude: a unimodal relationship with an optimum at intermediate altitudes of around 1800 m a.s.l. for tree vitality, radial growth, seed productivity and seed mass; and an increase with altitude for in‐vitro seed germination and seedling survival. A rise in temperature due to climate change could restrict P. australis to the upper most altitudes, where conservation of these forests will be a priority.

South American high-mountain ecosystems are greatly influenced by human disturbance. In the mountains of Có rdoba, Argentina, Polylepis australis (Rosaceae) woodlands are currently highly fragmented and subject to extensive burning and livestock grazing, resulting in severe changes of habitat characteristics, which hamper natural regeneration. In order to find out how to achieve successful reforestation, we compared P. australis seedling survival and growth and the development of a shrubby habit for two seed provenances and different planting microsites. Survival of planted seedlings after 5 years was 70%, with most deaths (19%) in the first year and declining mortality with ongoing establishment. Survival did not show any relationship with seed provenance or microsite characteristics. Height growth averaged 34.6 ± 1.2 cm in 5 years. Seedlings produced from seeds collected in a well-preserved woodland grew taller and showed a higher tendency for development of shrubby habit than those produced from seeds collected in a degraded woodland. Seedlings planted in more degraded microsites with exposed soil or rock due to past grazing pressure grew less and developed a more shrubby habit than those planted in better preserved microsites. Our results show that restoration of degraded areas with P. australis is possible and that there is potential to improve restoration success with a careful selection of seed provenance and planting microsites.

Interactions between resource and consumer species result in complex ecological networks. The overall structure of these networks is often stable in space and time, but little is known about the temporal stability of the functional roles of consumer species in these networks. We used a trait-based approach to investigate whether consumers (frugivorous birds) show similar degrees of functional specialisation on resources (plants) in ecological networks across seasons. We additionally tested whether closely related bird species have similar degrees of functional specialisation and whether birds that are functionally specialised on specific resource types within a season are flexible in switching to other resource types in other seasons. We analysed four seasonal replicates of two species-rich plant-frugivore networks from the tropical Andes. To quantify fruit preferences of frugivorous birds, we projected their interactions with plants into a multidimensional plant trait space. To measure functional specialisation of birds, we calculated a species' functional niche breadth (the extent of seasonal plant trait space utilised by a particular bird) and functional originality (the extent to which a bird species' fruit preference functionally differs from those of other species in a seasonal network). We additionally calculated functional flexibility, i.e. the ability of bird species to change their fruit preference across seasons in response to variation in plant resources. Functional specialisation of bird species varied more among species than across seasons, and phylogenetically similar bird species showed similar degrees of functional niche breadth (phylogenetic signal λ = 0·81) and functional originality (λ = 0·89). Additionally, we found that birds with high functional flexibility across seasons had narrow functional niche breadth and high functional originality per season, suggesting that birds that are seasonally specialised on particular resources are most flexible in switching to other fruit resources across seasons. The high flexibility of functionally specialised bird species to switch seasonally to other resources challenges the view that consumer species rely on functionally similar resources throughout the year. This flexibility of consumer species may be an important, but widely neglected mechanism that could potentially stabilise consumer-resource networks in response to human disturbance and environmental change.

Regeneration is known to be limited at many temperate tree lines, but very little data is available on the impacts of altitude and anthropogenic disturbance on regeneration patterns along tropical tree lines. The study focused on the reproductive traits of two Polylepis species in the Páramo de Papallacta in Ecuador along an altitudinal gradient, and involved different intensities of cattle trampling within subsequent altitudinal ranges. We analyzed flowering, fruit set, seed viability, germination, and seedling establishment as well as stand structure of Polylepis incana and P. pauta. The numbers of P. incana inflorescences and seedlings per m 2 showed a marginally significant decrease with increasing altitude. Mean tree height was significantly lower at higher altitudes, while stem number increased. The number of P. pauta inflorescences also decreased significantly upslope. In both forest types, trampling was found to have a positive impact on seedling abundance, presumably due to the removal of the litter layer. Thus, there was no evidence of negative effects of moderate cattle grazing on both tree line species. However, sapling establishment was minimal inside the forest stands at all altitudes and grazing levels, and we consequently observed a low proportion of narrow stems within all investigated forests. Our results show that, along with vegetative growth limitations of adult trees, important regeneration traits such as seedling and inflorescence numbers are also influenced by altitude, which might contribute to the formation of the upper tree line. Nevertheless, recruitment in the forest interior was low overall indicating that further factors, such as light conditions, affect regeneration of the studied species.

Aims: The high plant species diversity of tropical mountain forests is coupled with high habitat heterogeneity along gradients in elevation and topography. We quantified the effects of elevation, topography and forest edge on habitat conditions and woody plant diversity of tropical montane forest fragments.Location: Tropical montane forest fragments, 'Yungas', Bolivia. Methods:We measured microclimate and sampled soil properties and woody vegetation at forest edges and in the forest interior on ridges and in gorges along an elevational gradient of 600 m. We analysed effects of elevation, topography and forest edge on habitat conditions (i.e. microclimate, soil properties and forest structure), species richness, evenness and composition with linear mixed effects models and detrended correspondence analysis (DCA).Results: Changes in habitat conditions were weaker along the elevational gradient than between forest interior and forest edge and between different topographies. Species richness was not affected by any gradient, while species evenness was reduced at forest edges. All three gradients affected species composition, while effects of topography and forest edge were stronger than that of elevation. Conclusions:In general, effects of the 600-m elevational gradient were weak compared to effects of forest edge and topography. Edge effects shifted species composition towards pioneer species, while topographical heterogeneity is particularly important for generating high diversity in montane forests. These results underscore that edge effects have severe consequences in montane forest remnants and that small-scale variation between topographical microhabitats should be considered in studies that predict monotonous upslope migrations of plant species in tropical montane forests due to global warming.