Ecological coherence of marine protected area networks: a spatial assessment using species distribution models

2010 ◽  
Vol 48 (1) ◽  
pp. 112-120 ◽  
Author(s):  
Göran Sundblad ◽  
Ulf Bergström ◽  
Alfred Sandström
Keyword(s):  
Species Distribution ◽  
Protected Area ◽  
Species Distribution Models ◽  
Marine Protected Area ◽  
Distribution Models ◽  
Spatial Assessment ◽  
Ecological Coherence

scholarly journals Integrating distribution models and habitat classification maps into marine protected area planning.

2018 ◽  
Author(s):  
Renata Ferrari ◽  
Hamish Malcolm ◽  
Joe Neilson ◽  
Vanessa Lucieer ◽  
Alan Jordan ◽  
...  
Keyword(s):  
Functional Groups ◽  
Species Distribution ◽  
Conservation Planning ◽  
Species Distribution Models ◽  
Classification Scheme ◽  
Marine Protected Area ◽  
Sensitivity Analyses ◽  
Habitat Types ◽  
Distribution Models ◽  
Habitat Classification

Effective conservation planning requires biotic data across an entire region. In data-poor ecosystems conservation planning is informed by using environmental surrogates (e.g. temperature) predominantly in two ways: to develop habitat classification schemes (1) or develop species distribution models (2). We test the utility of both approaches for conservation planning of marine ecosystems, and rank environmental surrogates, such as depth and distance from shore, according to their power to predict the distribution and abundance of biotic species. Specifically, we compared a habitat classification scheme; based on coarse levels of habitat types derived from depth and distance from shore; against species distribution models, which predict fish abundance and prevalence as a function of environmental surrogates (depth, distance from shore, latitude, reef area, zoning, and several metrics of habitat structural complexity). We consistently set conservation target levels to 21% of each conservation feature, following global standards and a sensitivity analyses. Thus when running scenarios to protect fish species we aimed to protect at least 21% of each species, and when running scenarios of habitat classes, we aimed to protect at least 21% of each habitat class. We found that when aiming to protect 21% of the chosen conservation targets, distribution models protected 21% of the predicted abundance/occurrence of all modelled species and functional groups, but did not protect most habitats. Contrastingly, using a habitat classification scheme protected 21% of all habitat types and 34% of all species and functional groups, but required protecting three times more area. Thus, using only distribution models as targets in data-poor ecosystems could be a risky conservation planning strategy. Ultimately the best conservation outcomes were achieved by incorporating local knowledge to synthesize the conservation outcomes of both scenarios.

scholarly journals Protected Area Networks Do Not Represent Unseen Diversity

2021 ◽  
Author(s):  
Angel Delso ◽  
Jesús Muñoz ◽  
Javier Fajardo
Keyword(s):  
Protected Areas ◽  
Species Distribution ◽  
Conservation Planning ◽  
Protected Area ◽  
Species Distribution Models ◽  
Systematic Conservation Planning ◽  
Knowledge Gaps ◽  
Distribution Models ◽  
Current Area ◽  
Planning Approaches

Abstract Most existing protected area networks are usually biased to protect charismatic species or showy landscapes. We hypothesized that conservation networks designed including unseen diversity –groups usually species-rich but consisting of inconspicuous taxa, or affected by knowledge gaps– would be more efficient than networks ignoring those groups. To test this hypothesis, we created species distribution models for 3,006 species of arthropods and determined which were represented in three networks of different size and biogeographic origin. We assessed the efficiency of each network using spatial prioritization to measure its completeness –increment needed to achieve conservation targets– and specificity –how much overlap the priority areas based on unseen diversity with existing networks. We find that representativeness of unseen diversity in existing protected areas –extrinsic representativeness– was low, as ~40% of unseen diversity species were unprotected. We also find that existing networks should be expanded by an additional ~26-46% of their current area to complete targets, and that specificity can be as low as 8.8%, meaning that existing networks are not efficient to conserve unseen diversity. We conclude that information on unseen diversity must be included in systematic conservation planning approaches to design more efficient and ecologically representative protected areas.

scholarly journals A Robust Prediction Model for Species Distribution Using Bagging Ensembles with Deep Neural Networks

2021 ◽  
Vol 13 (8) ◽  
pp. 1495
Author(s):  
Jehyeok Rew ◽  
Yongjang Cho ◽  
Eenjun Hwang
Keyword(s):  
Neural Networks ◽  
Species Distribution ◽  
Best Practice ◽  
Deep Neural Networks ◽  
Species Distribution Models ◽  
Species Distribution Model ◽  
Environmental Data ◽  
Distribution Model ◽  
Distribution Models ◽  
Robust Prediction

Species distribution models have been used for various purposes, such as conserving species, discovering potential habitats, and obtaining evolutionary insights by predicting species occurrence. Many statistical and machine-learning-based approaches have been proposed to construct effective species distribution models, but with limited success due to spatial biases in presences and imbalanced presence-absences. We propose a novel species distribution model to address these problems based on bootstrap aggregating (bagging) ensembles of deep neural networks (DNNs). We first generate bootstraps considering presence-absence data on spatial balance to alleviate the bias problem. Then we construct DNNs using environmental data from presence and absence locations, and finally combine these into an ensemble model using three voting methods to improve prediction accuracy. Extensive experiments verified the proposed model’s effectiveness for species in South Korea using crowdsourced observations that have spatial biases. The proposed model achieved more accurate and robust prediction results than the current best practice models.

Sign in / Sign up

Export Citation Format

Share Document