Assessing shortfalls and complementary conservation areas for national plant biodiversity in South Korea.

Hyeyeong Choe,Patrick R Huber,Dongkun Lee,James F Quinn,James H Thorne,Laura Scherer

doi:10.1371/journal.pone.0190754

Hyeyeong Choe, Patrick R Huber + Show 4 more

Open Access

https://doi.org/10.1371/journal.pone.0190754

Copy DOI

Abstract

Protected areas (PAs) are often considered the most important biodiversity conservation areas in national plans, but PAs often do not represent national-scale biodiversity. We evaluate the current conservation status of plant biodiversity within current existing PAs, and identify potential additional PAs for South Korea. We modeled species ranges for 2,297 plant species using Multivariate Adaptive Regression Splines and compared the level of mean range representation in South Korea’s existing PAs, which comprise 5.7% of the country’s mainland area, with an equal-area alternative PA strategy selected with the reserve algorithm Marxan. We also used Marxan to model two additional conservation scenarios that add lands to approach the Aichi Biodiversity Target objectives (17% of the country). Existing PAs in South Korea contain an average of 6.3% of each plant species’ range, compared to 5.9% in the modeled equal-area alternative. However, existing PAs primarily represent a high percentage of the ranges for high-elevation and small range size species. The additional PAs scenario that adds lands to the existing PAs covers 14,587.55 km2, and would improve overall plant range representation to a mean of 16.8% of every species’ range. The other additional PAs scenario, which selects new PAs from all lands and covers 13,197.35 km2, would improve overall plant range representation to a mean of 13.5%. Even though the additional PAs that includes existing PAs represents higher percentages of species’ ranges, it is missing many biodiversity hotspots in non-mountainous areas and the additional PAs without locking in the existing PAs represent almost all species’ ranges evenly, including low-elevation ones with larger ranges. Some priority conservation areas we identified are expansions of, or near, existing PAs, especially in northeastern and southern South Korea. However, lowland coastal areas and areas surrounding the capital city, Seoul, are also critical for biodiversity conservation in South Korea.

Highlights

Biological diversity contributes to food security, human health, the provision of clean air and water, and enhancement of local livelihoods and economic development [1]
We modeled species ranges for 2,297 plant species using Multivariate Adaptive Regression Splines and compared the level of mean range representation in South Korea’s existing Protected natural areas (PAs), which comprise 5.7% of the country’s mainland area, with an equal-area alternative PA strategy selected with the reserve algorithm Marxan
We modeled plant species distributions using climatic and topographic predictor variables for 2,297 plant species with the MARS (Multivariate Adaptive Regression Splines) multi-response Species Distribution Models (SDM) algorithm [24,25], and assessed biodiversity representation in the existing PAs as the mean proportion of plant species’ predicted ranges that were included in existing conservation areas

Summary

Introduction

Biological diversity contributes to food security, human health, the provision of clean air and water, and enhancement of local livelihoods and economic development [1]. Species extinction is progressing rapidly, and biodiversity continues to be lost [2,3,4,5] Efforts to halt this crisis are ongoing, but limited funds influence the conservation strategies and planning methods that can be used [6]. Gap analysis has been used to measure effectiveness of existing PAs in representing all elements of regional biodiversity, with the goal of identifying gaps in protection and optimizing new PA acquisitions [12]. Such analyses often use species occurrence records to assess representation [13], many such data collections have problems including incompleteness and geographical biases [14]. Examples include Bolliger et al [15], who identified highly probable areas of species occurrence based on species habitat suitability modeling for seven animal species and designated the upper 10% of species’ ranges suitability areas as core habitats; and Hall et al [16], who estimated the potential biodiversity values of patches based on land cover type, landscape position, patch size, and distance to mature forest patches

Methods

Results

Conclusion