Global Mountain Biodiversity Assessment (GMBA)


Altitude referenced biological databases – a tool for understanding mountain biodiversity, part II

GMBA-GBIF workshop, Copenhagen, 26-28 Sept. 2007


Mary T. Kalin Arroyo et al. (Universidad de Chile, Chile)
Altitudinal range of plant species and colonizing ability in the South American Andes

Plant species distributed along the altitudinal gradient above treeline are adapted to a wide range of temperature conditions over a short physical distance, these telescoping conditions found over many degrees of latitude.Given that species with wider altitudinal ranges exhibit broader temperature niches than altitudinally narrowly distributed species, the probability of successful colonization/migration should be greatest in species with broad altitudinal ranges (Increased colonization probability hypothesis). Using presence and abundance records for 100 m altitudinal intervals on four mountains located at increasing distances eastward of the Southern Patagonian Icefield at the same latitude, we tested the hypothesis that alpine species with broader altitudinal ranges on the eastern source mountain for postglacial colonization are more likely to be successful colonizers on mountains closer to the Patagonian icefield exposed for less time since the retreat of the ice. Altitudinal range size was significantly correlated with number of mountains recolonized, but local abundance on the putative source mountain was not. These results suggest that niche breathe rather than source size determines colonizing ability. With these results in hand, we began investigating the increased colonization probability hypothesis at a biogeographical scale using georeferenced data base herbarium records for the genus Senecio, the largest genus in the alpine zone in the South American Andes. In Senecio, available data suggests that latitudinal range (a proxy of colonization success) is positively correlated with altitudinal range in puna species. The results in Senecio are interesting given that altitudinal and latitudinal ranges obtained from herbarium data base records are imperfect. That the tendency in Senecio was found in a pool of species which necessarily represent different evolutionary ages suggests the possibility of a hitherto undetected ecological rule in the alpine. Studies on many other genera and for a random sample of the high elevation flora are needed to confirm the generality of these findings, a goal that is attainable given the existence of electronic data bases such as the Chilean flora data base, and TROPICOS. Financed by Millennium Science Initiative Contract (ICM) No. P05-002, Chile, NGS and Fondecyt Chile grants.

Jingyun Fang (Peking University, China)
Exploring altitudinal patterns of plant diversity in China's mountains

China is a mountainous country with rich biodiversity resources. To explore altitudinal patterns and geographic differentiation of plant species diversity in China's mountains, Peking University has conducted a multiyear project on survey of plant species diversity of China's Mountains since the mid-1990s. I will present the protocol of the project, including the goals, research topics, mountain sites, and methodological standards for field surveys and data analysis. I will also show some achievements of the project.

Robert P. Guralnik (University of Colorado, Boulder, USA)
Databases as tool to model climate change effects on montane and alpine species distributions

Mountain species distributions are likely strongly shaped by past and current climate changes. During warming events, such as the major warming from the Last Glacial Maximum to middle Holocene, and the likely accelerating warming occurring now, populations have shifted from lower elevations to higher elevations in order to track favorable conditions within physiological tolerances. In so doing, it has been assumed that distributions likely became fragmented as low-lying valleys and basin became unsuitable habitats, forcing populations of colder adapted species onto isolated mountain peaks. Determining exactly how distributions shifted from cooler to warmer periods has proven to be a core historical biogeographical task, and workers have used inferences based directly on the fossil record and from patterns of population genetic variation to examine changes in distributional extent through time. Here I summarize continuing work that combines together fossil data from databases like Faunmap, population genetic data, and a novel approach, retrodicted ecological niche modeling using modern species occurrence data collected from global federated databases (e.g. the GBIF data portal), in order to document past potential distributions of species compared with modern ones. I aim to show, using a variety of molluscan and mammal exemplars, that simple patterns of vicariance caused by fragmentation of habitats from the Last Glacial Maximum to present may not be correct for many species and that species responses likely varied depending on the unique physiological characteristics of montane and alpine species. The results also show the value of global biodiversity databases for answering exciting mountain biodiversity and biogeography research questions.

Christian Körner (University of Basel, Switzerland)
Exploring the evolution and ecology of mountain biodiversity by linking organismic data bases with geophysical information systems

Geo-referenced biodiversity data bases can be linked with geo-statistical information systems to answer basic questions of the causes and trends of biodiversity. This article presents some ideas on the clever application of these novel tools with a mountain biodiversity focus. It is supposed that mountains represent a globally highly replicated 'experiment by nature' that offers testing a broad spectrum of hypotheses related to global biodiversity patterns. As 'islands in the sky' mountains permit exploring both space and time (season length) constrains to species richness and diversification within a broad matrix of climatic conditions. A tool kit of globally applicable climatic reference lines is presented that permits meaningful large scale comparisons. Examples are provided for the geostatistical analysis of climatic zones, seasonality, shrinking of land area with altitude and topographic steepness, all in support of such global comparisons.

Sandra Lavorel et al. (Laboratoire d’Ecologie Alpine CNRS, Grenoble, France)
A plant functional traits data base for the Alps – Application to the understanding of functional effects of changed grassland management

Plant functional traits are powerful tools to identify generic responses of biodiversity and ecosystem function to environmental change and to understand their underlying mechanisms. A collaborative initiative was launched to collect under a single data base information for vegetation composition, environmental variables, and plant traits of alpine ecosystems. This data base makes it possible to gather under a flexible structure data collected by naturalists, conservation managers and scientists, and to use this data to address both fundamental research and management questions. As a first test for the applicability of this data base we asked the following questions:(1) Knowing that plants show phenotypic variation for traits known as variable (e.g. height, leaf structural and chemical traits), to what extent are trait data collected under specific conditions in a given massif applicable to other alpine areas? This question is addressed by comparing trait values for a few common grassland species along a climatic gradient and across grassland management states.(2) Are there generic functional responses that can be detected using traits? This question is addressed by comparing responses of community-level traits to grassland management gradients within three different massifs.(3) Can traits be used to quantify the effects of environmental change on ecosystem services that are relevant to local stakeholders? This question is addressed by linking functional diversity to ecosystem functions associated with services identified by local stakeholders.

David Noguès-Bravo and JP Martinez-Rica (National Museum of Natural Sciences, CSIC, Madrid, Spain)
Are broad-scale e-datasets useful to model the distribution of mountain species? The case of White-Backed Woodpecker in the Spanish Pyrenees

Herein, I evaluate whether a broad-scale biological dataset in a mountain environment, Western Pyrenees, is useful to assess the effects of habitat loss and fragmentation in the presence of White-backed Woodpecker (Dendrocopos leucotos lifordi), a montane endangered bird in Spain. Thus, I assess the effect of habitat loss and fragmentation in the presence of White-backed Woodpecker using firstly an Atlas dataset (species presences/absences at 100 km2 of resolution) and secondly a detailed dataset based on detailed census records (species presences/absences at 1 km2 of resolution). Results obtained from both datasets agree, showing that large areas of suitable habitat and a scarcely fragmented habitat favored the viability of the species. Meanwhile more accurate digital datasets and fine-scale studies in mountain areas are being developed in the coming years, as proposed by GMBA and GBIF, these results suggest the usefulness of coarse e-datasets, species distribution at a 100 km2 resolution, to test broad ecological relationships between the distribution of species and their driving factors.

Peter Pearmann et al. (University of Lausanne, Switzerland)
Climate threat to alpine plant diversity in Europe: the EUROMONT initiative

A recent paper by Thuiller and co-authors predicted species loss up to 60% for mountain species in Europe. These projections were derived from niche-based species distribution models built at a coarse scale over Europe and projected at a slightly finer resolution (cells of 16x16 km). Although these broad scale projections can provide valuable scenarios for many species, we question their accuracy for mountain species. Previous studies at finer scale in local study areas suggest diverging trends between broad and local scale projections. As these projections are then used to assess management issues, such as the role of parks and natural reserves for preserving future biodiversity as climate changes, it has become urgent to assess whether reliable local trends can be predicted from global projections, or if local projections only should be used. A workshop was recently held in Lausanne to gather plant distribution data at the fine scale for 11 study areas across six mountain ranges, this way allowing the local assessment of climate threat to mountain floras in Europe. Predictive models were built and climate change projections were derived. Preliminary results show on average smaller extinction rates at the local scale, due to some species finding refugees at high elevation, but even higher extinction rates than predicted at the European scale can also be observed for some species in some study areas. Hence, responses differ largely between study areas, depending on their species pool and climate sensitivity. Whatever the scale, important extinctions remain predicted per pixel. More advanced analyses are still going on, which will provide more informed conclusions.

Roberto Pizzolotto and Pietro Brandmayr (Università della Calabria, Italy)
A georeferenced databank for evaluating the impact of climate change in Southern Italian mountains

The Ecology Department of Calabria University (Italy), developed a database aimed to make available data collected in many faunal studies about several species harboured by the National Park of Pollino Mount. This Habitat Biodiversity Data Base (HBDB) allows an interaction between vegetation map and faunal data. The HBDB at present is a databank set on Ground Beetles and Lepidoptera Ropalocera collected since seventies until now in the Pollino National Park. The access to data is possible through queries, which can provide several information as well as community structure of considered habitat, detailed file of each species and related information about their own biology and range zone. These outcomes can give temporal data about the considered species, allowing to get information about possible changes occurred in time on spatial use. In order to evaluate the effects of the climate change on the fauna of the Mediterranean Basin, the HBDB has been used to get data about carabids and butterflies which have been sampled in 1977, then re-sampled after a quarter of century (2004) on the top of Pollino Mount. The period comprised between sample and re-sample was marked by a significant temperature increase for the entire globe. Many changes in species composition occurred in stands at altitude lower than 1,900m asl, where orophilous species have gone locally extinct or at least strongly reduced. Some Mediterranean or thermophilous species have increased populations size and have colonised areas at higher altitude than 25 years ago. These changes are probably linked to the increase in temperature.Many species reach the boundaries of their southern range on Pollino’s mountains, and some of these are present with endemic subspecies. As a consequence of temperature increase, it is likely that some species became totally ol locally extinct.

Carsten Rahbek (University of Copenhagen, Denmark)
Is there a general altitudinal gradient of species richness?

China is a mountainous country with rich biodiversity resources. To explore altitudinal patterns and geographic differentiation of plant species diversity in China's mountains, Peking University has conducted a multiyear project on survey of plant species diversity of China's Mountains since the mid-1990s. I will present the protocol of the project, including the goals, research topics, mountain sites, and methodological standards for field surveys and data analysis. I will also show some achievements of the project.

Tom Romdal and J-A. Grytnes (University of Copenhagen, Denmark and University of Bergen, Norway)
An indirect area effect on elevational species richness patterns

Studies of elevational species richness variation attempt to neutralize area effects through standardized sampling. We propose that regional-local richness relations cause the area of suitable habitat in elevational zones or bands to influence point diversity measured along elevational transects. We test this hypothesis by a meta-analysis using previously published studies containing elevational richness data from local scale. The analysis is restricted to studies in forest habitat that are standardised for area and effort. Contiguous forest area for individual studies was defined and measured through the global MODIS VCF 500m resolution dataset. Elevational band forest area was subsequently tested as determinant for local species richness. Results show that regional habitat area has a considerable and robust effect on local-scale elevational richness pattern. Habitat area contributes to both monotonically decreasing and hump-shaped species richness patterns. A number of factors influence the strength of the effect. Different taxa have different dispersal abilities and habitat restrictions. The definitions of forest distribution and forest contiguity, grain size, the total amount of forest extent and the latitude are also necessary to consider. We conclude that habitat area should be incorporated as a potentially important factor for local-scale elevational species richness patterns.
This paper was printed in: Tom S. Romdal and John-Arvid Grytnes (2007) An indirect area effect on elevational species richness patterns. Ecography 30: 440 448, 2007

Nathan Sanders (University of Tennessee, USA)
Independent elevational gradients in, and the controls of, decomposer, producer, consumer, and predator diversity

Interactions between trophic levels can shape diversity within local assemblages. In addition, regional-scale factors such as climatic conditions, spatial constraints, and habitat diversity can also limit diversity within local assemblages. Untangling the relative contributions of local interactions and regional scale determinants on local diversity has been challenging, largely because few ecologists work at local and regional scales and because empirical data have been unavailable. Here, we argue that elevational gradients are useful tools for understanding the relative effects of local and regional scale factors on local diversity. Using a growing, georeferenced database in the southern Appalachian Mountains, we ask whether there is evidence that trophic interactions to positive correlations among decomposers, producers, consumers, and predators or whether regional-scale factors (climatic conditions, spatial factors) shape diversity within assemblages. The database consists of georeferenced records for >10,0000 specimens in each trophic group along an elevational gradient from 300m to 2100m. Using spatially autoregressive models, we found that climatic factors, particularly actual evapotranspiration, rather than spatial factors or trophic interactions shape local diversity gradients. Diversity was not correlated across trophic groups, suggesting that each trophic group responded to environmental variation in subtly different ways. Taken together, our results indicate that systematic environmental variation along elevational gradients, coupled with robust georeferenced collections for a variety of taxa across trophic groups, can uncover the relative effects of local and regional factors on the structure of local communities.

Eva Spehn (University of Basel, Switzerland)
GMBA activities and aim of the mountain biodiversity database project

Yuan Zhu (Beijing Normal University, China)
Altitudinal pattern of vascular plant species richness based on equal-area belts in Mt. Helan

Altitudinal pattern of plant species richness along an elevational gradient has often been studied by dividing the mountains into equal-elevation belts. However, comparisons of species richness among different belts with different areas is not appropriate. Based on a Digital Elevation Model (DEM) and Geographic Informational System (GIS), we divided Mt. Helan (1,300–3,500 m) into several equal-area belts along an altitudinal gradient, and compared the plant species richness among them. The results were as followed: (1) The altitudinal pattern of species richness in Mt. Helan showed a unimodal shape, peaking around 2,000 m altitude. (2) The slope heterogeneity was the primary variable to explain the altitudinal pattern of species richness. High values of slope heterogeneity could reflect the complex topographic features and diversified habitats within a belt, implying a favorable condition for more species to coexist. (3) The unimodal pattern of species richness might result from a combined effect of evolutionary history of vegetation, climates, topographic features, ecotone, and the mid-domain effect. (4) Compared with equal-elevation belts, equal-area belts could directly eliminate the influence of area upon species richness, leading to a more reliable analytic result, especially when the species data were accurate at all altitudinal ranges.
This paper was printed in: Zhu, et al. 2007. Altitudinal pattern of vascular plant species richness based on equal-area belts in Mt. Helan. Biodiversity Science, 15(4): 408~418. (in Chinese)