A global system of harmonized observations is needed to inform scientists and policy-makers.

Aim To develop a novel global spatial framework for the integration and analysis of ecological and environmental data. Location The global land surface excluding Antarctica. Methods A broad set of climate‐related variables were considered for inclusion in a quantitative model, which partitions geographic space into bioclimate regions. Statistical screening produced a subset of relevant bioclimate variables, which were further compacted into fewer independent dimensions using principal components analysis (PCA). An ISODATA clustering routine was then used to classify the principal components into relatively homogeneous environmental strata. The strata were aggregated into global environmental zones based on the attribute distances between strata to provide structure and support a consistent nomenclature. Results The global environmental stratification (GEnS) consists of 125 strata, which have been aggregated into 18 global environmental zones. The stratification has a 30 arcsec resolution (equivalent to 0.86 km2 at the equator). Aggregations of the strata were compared with nine existing global, continental and national bioclimate and ecosystem classifications using the Kappa statistic. Values range between 0.54 and 0.72, indicating good agreement in bioclimate and ecosystem patterns between existing maps and the GEnS. Main conclusions The GEnS provides a robust spatial analytical framework for the aggregation of local observations, identification of gaps in current monitoring efforts and systematic design of complementary and new monitoring and research. The dataset is available for non‐commercial use through the GEO portal (http://www.geoportal.org).

A new map of standardized, mesoscale (tens to thousands Ecosystems provide many benefits to human societies of hectares) terrestrial ecosystems for the conterminous United as goods (such as food, fuel, and fiber) and services (such as States was developed by using a biophysical stratification water purification, maintenance of soil fertility, and pollinaapproach. The ecosystems delineated in this top-down, deduction). Sustained provision of these goods and services is tive modeling effort are described in NatureServe's classificaimportant for human societies, but recent studies like that by tion of terrestrial ecological systems of the United States. The the Millennium Ecosystem Assessment (2005) have shown ecosystems were mapped as physically distinct areas and were that degraded ecosystems are compromised in their ability associated with known distributions of vegetation assemblages to provide these benefits. In order for existing ecosystems to by using a standardized methodology first developed for South persist on the planet, they will need to be well managed. Many America. This approach follows the geoecosystems concept of U.S. agencies and international organizations now promote R.J. Huggett and the ecosystem geography approach of R.G. ecosystem-based management approaches; these groups Bailey. include the U.S. Bureau of Land Management (1994), the U.S. Unique physical environments were delineated through Department of Agriculture (USDA) Forest Service (2006a, p. a geospatial combination of national data layers for biogeog-22), the U.S. Geological Survey (as described by Myers and raphy, bioclimate, surficial materials lithology, land surface others (2007)), the United Nations Environment Programme forms, and topographic moisture potential. Combining these (which supports the Secretariat of the Convention on Biologilayers resulted in a comprehensive biophysical stratification cal Diversity, 2000), and the Group on Earth Observations of the conterminous United States, which produced 13,482 (2005). Management of ecosystems necessarily requires an unique biophysical areas. These were considered as fundamenunderstanding of the types, spatial pattern and scales, and tal units of ecosystem structure and were aggregated into 419 distributions of ecosystems that occur within the management potential terrestrial ecosystems. jurisdiction. The ecosystems classification effort preceded the map-Ecosystem definition, classification, and mapping have ping effort and involved the independent development of received considerable attention since ecosystems were first diagnostic criteria, descriptions, and nomenclature for describchampioned by Tansley (1935). Odum (1953), in his seminal ing expert-derived ecological systems. The aggregation and textbook on ecology, recognized ecosystems as systems of labeling of the mapped ecosystem structure units into the biotic communities interacting with their physical environecological systems classification was accomplished in an iterament. These early definitional concepts have evolve...

No abstract

Projected climate and environmental change are expected to increase the pressure on global freshwater resources. To prepare for and cope with the related risks, stakeholders need to devise plans for sustainable management of river systems, which in turn requires the identification of management-appropriate operational units, such as groups of rivers that share similar environmental and biological characteristics. Ideally, these units are of a manageable size, and are biotically or abiotically distinguishable across a variety of river types. Here, we aim to address this need by presenting a new global river classification framework (GloRiC) to establish a common vocabulary and standardized approach to the development of globally comprehensive and integrated river classifications that can be tailored to different goals and requirements. We define the GloRiC conceptual framework based on five categories of variables: (1) hydrology; (2) physiography and climate; (3) fluvial geomorphology; (4) water chemistry; and (5) aquatic biology. We then apply the framework using hydroenvironmental attributes provided by a seamless high-resolution river reach database to create initial instances of three sub-classifications (hydrologic, physio-climatic, and geomorphic) which we ultimately combine into 127 river reach types at the global scale. These supervised classifications utilize a mix of statistical analyses and expert interpretation to identify the classifier variables, the number of classes, and their thresholds. In addition, we also present an unsupervised, multivariable k-means statistical clustering of all river reaches into 30 groups. These first-of-their-kind global river reach classifications at high spatial resolution provide baseline information for a total of 35.9 million kilometers of rivers that have been assessed in this study, and are expected to be particularly useful in remote or data-poor river basins. The GloRiC framework and associated data are primarily designed for broad and rapid applicability in assessments that require stratified analyses of river ecosystem conditions at global and regional scales; smaller-scale applications could follow the same conceptual framework yet use more detailed data sources.