Jayme Augusto Prevedello

The importance of scattered trees for biodiversity conservation: A global meta‐analysis

Summary 1.Scattered trees are thought to be keystone structures for biodiversity in landscapes worldwide. However, such trees have been largely neglected by researchers and their importance for biodiversity remains unclear. 2.We completed a global meta-analysis to quantify relationships between scattered trees and the species richness, abundance and composition of vertebrates, arthropods and plants. First, we tested whether areas near scattered trees support higher levels of species richness and abundance than nearby open areas. Second, we compared levels of species richness and abundance in matrix areas with scattered trees and areas embedded within nearby habitat patches. We also compared the composition of biological communities inhabiting habitat patches, open areas and areas with scattered trees. 3.A total of 62 studies contained suitable data for our quantitative analyses. The local abundance of arthropods, vertebrates and woody plants was 60–430% greater and overall species richness was 50–100% higher in areas with scattered trees than in open areas. Conversely, for herbaceous plants, there was no consistent relationship between species abundance and the occurrence of scattered trees, although species richness was, on average, 43% lower. 4.The abundance and richness of all taxonomic groups was similar in matrix areas supporting scattered trees and habitat patches, although the species richness of epiphytes was, on average, 50% higher in habitat patches. Communities inhabiting habitat patches were more similar in composition to the communities inhabiting areas with scattered trees, and less similar to the communities of open areas. 5.Synthesis and applications. Areas with scattered trees support greater levels of biodiversity than open areas, as well as communities that are more similar to those inhabiting habitat patches. Scattered trees can be regarded as keystone structures for vertebrates, arthropods and terrestrial plants in landscapes worldwide. The maintenance of scattered trees may be compatible with livestock grazing in some agricultural landscapes. Greater management effort and targeted, long-term policies are needed to retain or re-establish scattered trees in many farming landscapes in both forest and non-forest biomes around the world. This article is protected by copyright. All rights reserved.

Effects of cornfields on small mammal communities: a test in the Atlantic Forest hotspot

Croplands have expanded dramatically during the last century, frequently leading to severe biodiversity losses within occupied areas. In addition to such direct influences, croplands may also have affected biodiversity within adjacent natural habitats, yet such potential indirect effects have rarely been quantified. Here, we test for effects of cornfields on small mammal communities inhabiting forest fragments in the Atlantic Forest biodiversity hotspot, Brazil. We test the hypotheses that cornfields increase the abundance and dominance of granivorous small mammals in adjacent forest fragments but reduce the abundance of nongranivorous species and community richness. To test these hypotheses, we used a replicated randomized block design, comparing communities located at forest interiors, forest-cornfield edges, and also forest-road edges. The abundance of granivorous small mammals was significantly higher at forest-cornfield edges compared to the other treatments, whereas insectivorous—omnivorous species had an apparently neutral response to the edges. Forest-cornfield edges harbored fewer small mammal species and tended to have a greater dominance of granivores. Forestcornfield and forest-road edges harbored species characteristic of both forested and nonforested habitats. Our findings suggest that cornfields alter the structure of native small mammal communities, mostly by providing complementary resources for granivores.

Seasonality in metacommunity structure: an empirical test in the Atlantic Forest

ContextThe metacommunity concept helps to understand how local and regional processes regulate species distributions in landscapes. Metacommunity structure is often assumed as static, but may be rather dynamic, following temporal changes along environmental gradients.ObjectivesWe present an empirical test of the temporal dynamics of metacommunity structure, using small mammals in an Atlantic Forest landscape as a model system.MethodsWe analyzed incidence matrices using the Elements of Metacommunity Structure framework and evaluated whether local, landscape, and spatial factors structured the metacommunity during different climatic seasons (HS = humid; SHS = super-humid) and time periods (1 = 1999–2001; 2 = 2005–2009). We compared HS-1 and SHS-1 to evaluate if metacommunity structure varies between seasons, and HS-1 and HS-2 to evaluate if it varies between time periods.ResultsMetacommunity structure changed from Clementsian (HS-1) to random (SHS-1), but during HS-2 it was Clementsian again. This suggests that groups of species are responding similarly to the major gradient of variation during the HS only. Patch size structured the metacommunity during both humid periods, and local habitat structure only during HS-1. We suggest that during the SHS these gradients are lost due to increased matrix permeability to movement, which homogenizes local communities resulting in a random structure.ConclusionsSpecies habitat requirements and specializations determined metacommunity structure, but only during the HS. The Clementsian structure indicates that forest disturbances may result in the loss of whole groups of species during the HS. Alternating patterns of metacommunity structure may be associated to changes on matrix suitability between seasons.

Habitat fragmentation affects individual condition: evidence from small mammals of the Brazilian Atlantic Forest

Habitat loss and fragmentation are likely to affect individual condition as they usually alter the quality of habitat. However, responses of individuals of different species to these processes may be subtle and difficult to detect, despite that such responses may reveal interesting and diverse strategies of persistence of species in fragmented landscapes. Here, we used model selection to 1) compare the body condition (Scaled Mass Index) of 2 Neotropical marsupials (Didelphis aurita and Philander frenatus) from 6 continuous forest sites and 25 forest fragments, and 2) determine whether local habitat structure and landscape and temporal factors affect body condition in forest fragments. Body condition of both species was lower in forest fragments compared to continuous forest sites, but only in the super-humid season for D. aurita and in the humid season for P. frenatus. When only forest fragments were compared, body condition of D. aurita was higher in the humid season, in sites with fewer Cecropia trees and water courses, lower percent forest cover, and surrounded by a mixed matrix (plantation and cattle ranching), or located in small rural properties. Body condition of P. frenatus was slightly higher in the super-humid season, in sites surrounded by a mixed matrix, with fewer Cecropia trees and water courses, more lianas, and greater percent forest cover. Our results revealed negative effects of habitat loss and fragmentation on both species, but also different responses of each species to local extrinsic environmental factors. We suggest that body condition should not be used as a direct indicator of animal fitness or habitat quality. Nonetheless, body condition may still reveal subtle and previously undetected responses of species to habitat disturbance and fragmentation.

ATLANTIC MAMMAL TRAITS: a data set of morphological traits of mammals in the Atlantic Forest of South America

Measures of traits are the basis of functional biological diversity. Numerous works consider mean species-level measures of traits while ignoring individual variance within species. However, there is a large amount of variation within species and it is increasingly apparent that it is important to consider trait variation not only between species, but also within species. Mammals are an interesting group for investigating trait-based approaches because they play diverse and important ecological functions (e.g., pollination, seed dispersal, predation, grazing) that are correlated with functional traits. Here we compile a data set comprising morphological and life history information of 279 mammal species from 39,850 individuals of 388 populations ranging from -5.83 to -29.75 decimal degrees of latitude and -34.82 to -56.73 decimal degrees of longitude in the Atlantic forest of South America. We present trait information from 16,840 individuals of 181 species of non-volant mammals (Rodentia, Didelphimorphia, Carnivora, Primates, Cingulata, Artiodactyla, Pilosa, Lagomorpha, Perissodactyla) and from 23,010 individuals of 98 species of volant mammals (Chiroptera). The traits reported include body mass, age, sex, reproductive stage, as well as the geographic coordinates of sampling for all taxa. Moreover, we gathered information on forearm length for bats and body length and tail length for rodents and marsupials. No copyright restrictions are associated with the use of this data set. Please cite this data paper when the data are used in publications. We also request that researchers and teachers inform us of how they are using the data.