Concern about biodiversity loss has led to increased public investment in conservation. Whereas there is a widespread perception that such initiatives have been unsuccessful, there are few quantitative tests of this perception. Here, we evaluate whether rates of biodiversity change have altered in recent decades in three European countries (Great Britain, Netherlands and Belgium) for plants and flower visiting insects. We compared four 20-year periods, comparing periods of rapid land-use intensification and natural habitat loss (1930–1990) with a period of increased conservation investment (post-1990). We found that extensive species richness loss and biotic homogenisation occurred before 1990, whereas these negative trends became substantially less accentuated during recent decades, being partially reversed for certain taxa (e.g. bees in Great Britain and Netherlands). These results highlight the potential to maintain or even restore current species assemblages (which despite past extinctions are still of great conservation value), at least in regions where large-scale land-use intensification and natural habitat loss has ceased.

With the expansion in the quantity and types of biodiversity data being collected, there is a need to find ways to combine these different sources to provide cohesive summaries of species' potential and realized distributions in space and time. Recently, model-based data integration has emerged as a means to achieve this by combining datasets in ways that retain the strengths of each. We describe a flexible approach to data integration using point process models, which provide a convenient way to translate across ecological currencies. We highlight recent examples of large-scale ecological models based on data integration and outline the conceptual and technical challenges and opportunities that arise. Species Distribution Models in EcologyLarge-scale ecological models of how species distributions and abundances vary over space and time are a critical tool in macroecology, biogeography, and conservation biology. They underpin our understanding of how biodiversity is shaped, how it is responding to anthropogenic activities, and how it might change in the future [1][2][3]. There is now a substantial literature on statistical tools for building species distribution models (SDMs) (see Glossary) and best practice in how to fit them [4][5][6][7]. SDMs also form a building block upon which more complex models, incorporating occupancy and/or abundance in space and time, can be built [8,9].

B iodiversity and the many ecosystem functions and services it underpins are undergoing significant and often rapid changes worldwide 1. A range of global initiatives and policy frameworks, including the Convention on Biological Diversity (CBD) and Sustainable Development Goals (SDGs), have aimed to reduce this change and to halt the loss of biodiversity, with limited progress to date 2. Appropriately gauging the impact of such policies or the progress toward international biodiversity goals has a key requirement: the availability of information on the status and trends of biodiversity in a form that is easily understood, timely, scientifically rigorous, standardized, relevant, global and representative of species populations across taxa and regions over time. Such information is particularly crucial in assessments, such as those carried out by the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) 3 , and is needed to construct 'indicators' , which are aggregate measures that often address specific conservation targets 4,5. Underpinning such metrics are core, essential measurements known as EBVs, which capture key constituent components of biodiversity change 6,7 , akin and complementary to the 'essential climate variables' supporting climate change assessment and policy 8. Facilitated by the Group on Earth Observations Biodiversity Observation Network (GEO BON, http://geobon.org) and related efforts, the biodiversity science and observation community is now engaging in an effort to conceptualize and formulate these essential biodiversity components to enable more focused, integrated, and effective biodiversity monitoring in support of assessment and policy within a unified framework. This study represents the formal outcome of a process undertaken from 2015 through 2018 by the founding members of the GEO BON Species Populations Working Group 9 , which includes the authors of this Perspective, charged with providing the formal definitions, conceptualizations and recommendations addressing species distribution and abundance EBVs. Changes in species distribution and abundance affect all biodiversity facets 10 , including the loss of potentially significant traits and functions 1,11 and associated ecosystem consequences 12,13. Patterns of spatial distribution and changes to these patterns inform us about the commonness, rarity and potential extinction risk for species 14-16 , determine the national and regional stewardship of species and are key to ensuring effective monitoring 17 , protection 18,19 and population

Despite the broad conceptual and applied relevance of how the number of species or endemics changes with area (the species-area and endemics-area relationships (SAR and EAR)), our understanding of universality and pervasiveness of these patterns across taxa and regions has remained limited. The SAR has traditionally been approximated by a power law, but recent theories predict a triphasic SAR in logarithmic space, characterized by steeper increases in species richness at both small and large spatial scales. Here we uncover such universally upward accelerating SARs for amphibians, birds and mammals across the world’s major landmasses. Although apparently taxon-specific and continent-specific, all curves collapse into one universal function after the area is rescaled by using the mean range sizes of taxa within continents. In addition, all EARs approximately follow a power law with a slope close to 1, indicating that for most spatial scales there is roughly proportional species extinction with area loss. These patterns can be predicted by a simulation model based on the random placement of contiguous ranges within a domain. The universality of SARs and EARs after rescaling implies that both total and endemic species richness within an area, and also their rate of change with area, can be estimated by using only the knowledge of mean geographic range size in the region and mean species richness at one spatial scale.

In ecology, the true causal structure for a given problem is often not known, and several plausible models and thus model predictions exist. It has been claimed that using weighted averages of these models can reduce prediction error, as well as better reflect model selection uncertainty. These claims, however, are often demonstrated by isolated examples. Analysts must better understand under which conditions model averaging can improve predictions and their uncertainty estimates. Moreover, a large range of different model averaging methods exists, raising the question of how they differ in their behaviour and performance. Here, we review the mathematical foundations of model averaging along with the diversity of approaches available. We explain that the error in model‐averaged predictions depends on each model's predictive bias and variance, as well as the covariance in predictions between models, and uncertainty about model weights. We show that model averaging is particularly useful if the predictive error of contributing model predictions is dominated by variance, and if the covariance between models is low. For noisy data, which predominate in ecology, these conditions will often be met. Many different methods to derive averaging weights exist, from Bayesian over information‐theoretical to cross‐validation optimized and resampling approaches. A general recommendation is difficult, because the performance of methods is often context dependent. Importantly, estimating weights creates some additional uncertainty. As a result, estimated model weights may not always outperform arbitrary fixed weights, such as equal weights for all models. When averaging a set of models with many inadequate models, however, estimating model weights will typically be superior to equal weights. We also investigate the quality of the confidence intervals calculated for model‐averaged predictions, showing that they differ greatly in behaviour and seldom manage to achieve nominal coverage. Our overall recommendations stress the importance of non‐parametric methods such as cross‐validation for a reliable uncertainty quantification of model‐averaged predictions.

Hansen et al. (Reports, 15 November 2013, p. 850) published a high-resolution global forest map with detailed information on local forest loss and gain. We show that their product does not distinguish tropical forests from plantations and even herbaceous crops, which leads to a substantial underestimate of forest loss and compromises its value for local policy decisions.

No abstract

No abstract