Global hotspots and correlates of emerging zoonotic diseases

Nature Communications, Journal Year: 2017, Volume and Issue: 8(1)

Published: Oct. 18, 2017

Abstract Zoonoses originating from wildlife represent a significant threat to global health, security and economic growth, combatting their emergence is public health priority. However, our understanding of the mechanisms underlying remains rudimentary. Here we update database emerging infectious disease (EID) events, create novel measure reporting effort, fit boosted regression tree models analyze demographic, environmental biological correlates occurrence. After accounting for show that zoonotic EID risk elevated in forested tropical regions experiencing land-use changes where biodiversity (mammal species richness) high. We present new hotspot map spatial variation index, partial dependence plots illustrating relationships between events predictors. Our results may help improve surveillance long-term monitoring programs, design field experiments test emergence.

Language: Английский

---

Is my species distribution model fit for purpose? Matching data and models to applications

Global Ecology and Biogeography, Journal Year: 2015, Volume and Issue: 24(3), P. 276 - 292

Published: Jan. 8, 2015

Abstract Species distribution models ( SDM s) are used to inform a range of ecological, biogeographical and conservation applications. However, users often underestimate the strong links between data type, model output suitability for end‐use. We synthesize current knowledge provide simple framework that summarizes how interactions type sampling process (i.e. imperfect detection bias) determine quantity is estimated by . then draw upon published literature simulations illustrate evaluate information needs most common applications outputs. find that, while predictions fitted commonly available observational (presence records) suffice some applications, others require estimates occurrence probabilities, which unattainable without reliable absence records. Our review reveal converting continuous outputs into categories assumed presence or practice, it seldom clearly justified application's objective usually degrades inference. Matching s particular critical avoid poor scientific inference management outcomes. This paper aims help modellers assess whether their intended indeed fit purpose.

Language: Английский

---

Global priorities for an effective information basis of biodiversity distributions

Nature Communications, Journal Year: 2015, Volume and Issue: 6(1)

Published: Sept. 8, 2015

Abstract Gaps in digital accessible information (DAI) on species distributions hamper prospects of safeguarding biodiversity and ecosystem services, addressing central ecological evolutionary questions. Achieving international targets knowledge requires that gaps be identified actions prioritized. Integrating 157 million point records distribution maps for 21,170 terrestrial vertebrate species, we find outside a few well-sampled regions, DAI occurrences provides very limited spatially biased inventories species. Surprisingly, many large, emerging economies are even more under-represented global than species-rich, developing countries the tropics. Multi-model inference reveals completeness is mainly by distance to researchers, locally available research funding participation data-sharing networks, rather transportation infrastructure, or size Western data contributors as often assumed. Our results highlight urgent need integrating non-Western sources intensifying cooperation effectively address societal needs.

Language: Английский

---

Multidimensional biases, gaps and uncertainties in global plant occurrence information

Ecology Letters, Journal Year: 2016, Volume and Issue: 19(8), P. 992 - 1006

Published: June 2, 2016

Plants are a hyperdiverse clade that plays key role in maintaining ecological and evolutionary processes as well human livelihoods. Biases, gaps uncertainties plant occurrence information remain central problem ecology conservation, but these limitations largely unassessed globally. In this synthesis, we propose conceptual framework for analysing coverage, biases metrics along taxonomic, geographical temporal dimensions, apply it to all c. 370 000 species of land plants. To end, integrated 120 million point-occurrence records with independent databases on taxonomy, distributions conservation status. We find different data prevalent each dimension. Different coverage uncertainty uncorrelated, reducing spatial or by filtering out would usually come at great costs coverage. light multidimensional limitations, discuss prospects global biogeographical research, monitoring outline critical next steps towards more effective usage mobilisation. Our study provides an empirical baseline evaluating improving floristic knowledge, can be applied other clades.

Language: Английский

---

Bias correction in species distribution models: pooling survey and collection data for multiple species

Methods in Ecology and Evolution, Journal Year: 2014, Volume and Issue: 6(4), P. 424 - 438

Published: Oct. 10, 2014

Presence-only records may provide data on the distributions of rare species, but commonly suffer from large, unknown biases due to their typically haphazard collection schemes. Presence-absence or count collected in systematic, planned surveys are more reliable less abundant.We proposed a probabilistic model allow for joint analysis presence-only and survey exploit complementary strengths. Our method pools presence-absence many species maximizes likelihood, simultaneously estimating adjusting sampling bias affecting data. By assuming that is same all we can borrow strength across efficiently estimate improve our inference data.We evaluate model's performance 36 eucalypt south-eastern Australia. We find exhibit strong towards coast Sydney, largest city. data-pooling technique substantially improves out-of-sample predictive when amount available given scarceIf have only no both types several other spatial bias, then obtain an unbiased first species' geographic range.

Language: Английский

---

Essential biodiversity variables for mapping and monitoring species populations

Nature Ecology & Evolution, Journal Year: 2019, Volume and Issue: 3(4), P. 539 - 551

Published: March 11, 2019

Species distributions and abundances are undergoing rapid changes worldwide. This highlights the significance of reliable, integrated information for guiding assessing actions policies aimed at managing sustaining many functions benefits species. Here we synthesize types data approaches that required to achieve such an integration conceptualize 'essential biodiversity variables' (EBVs) a unified global capture species populations in space time. The inherent heterogeneity sparseness raw overcome by use models remotely sensed covariates inform predictions contiguous time extent. We define population EBVs as space-time-species-gram (cube) simultaneously addresses distribution or abundance multiple species, with its resolution adjusted represent available evidence acceptable levels uncertainty. essential enables monitoring single aggregate spatial taxonomic units scales relevant research decision-making. When combined ancillary environmental data, this fundamental directly underpins range ecosystem function indicators. concept present links disparate downstream uses informs vision which collection is closely infrastructure support effective assessment.

Language: Английский

---

Point process models for presence‐only analysis

Methods in Ecology and Evolution, Journal Year: 2015, Volume and Issue: 6(4), P. 366 - 379

Published: Feb. 23, 2015

Summary Presence‐only data are widely used for species distribution modelling, and point process regression models a flexible tool that has considerable potential this problem, when arise as events. In paper, we review models, some of their advantages common methods fitting them to presence‐only data. Advantages include (and not limited to) clarification what the response variable is modelled; framework choosing number location quadrature points (commonly referred pseudo‐absences or ‘background points’) objectively; clarity model assumptions tools checking them; handle spatial dependence between it present; ways forward regarding difficult issues such accounting sampling bias. Point related approaches which means variety different software can be fit these including maxent generalised linear modelling software.

Language: Английский

---

Modelling of species distributions, range dynamics and communities under imperfect detection: advances, challenges and opportunities

Ecography, Journal Year: 2016, Volume and Issue: 40(2), P. 281 - 295

Published: June 20, 2016

Building useful models of species distributions requires attention to several important issues, one being imperfect detection species. Data sets detections are likely suffer from false absence records. Depending on the type survey, positive records can also be a problem. Disregarding these observation errors may lead biases in model estimation as well overconfidence about precision. The severity problem depends intensity and how they correlate with environmental characteristics (e.g. where detectability strongly habitat features). A powerful modelling framework that accounts for has developed last 10–15 yr. Fundamental this is data must collected way informative process. For instance, such form multiple detection/non‐detection obtained visits/observers/detection methods at (at least) some sites, or times within survey visit. extend studying species’ range dynamics communities, approaches analysing abundance occupancy states (rather than binary presence/absence). This paper summarizes advances, discusses evidence effects difficulties working it, concludes current outlook future research application methods.

Language: Английский

---

Data Integration for Large-Scale Models of Species Distributions

Trends in Ecology & Evolution, Journal Year: 2019, Volume and Issue: 35(1), P. 56 - 67

Published: Nov. 2, 2019

With the expansion in quantity and types of biodiversity data being collected, there is a need to find ways combine these different sources provide cohesive summaries species' potential realized distributions space time. Recently, model-based integration has emerged as means achieve this by combining datasets that retain strengths each. We describe flexible approach using point process models, which convenient way translate across ecological currencies. highlight recent examples large-scale models based on outline conceptual technical challenges opportunities arise.

Language: Английский

---

Synthesizing multiple data types for biological conservation using integrated population models

Biological Conservation, Journal Year: 2017, Volume and Issue: 217, P. 240 - 250

Published: Nov. 20, 2017

Language: Английский

---