Elsevier

Applied Geography

Volume 44, October 2013, Pages 12-22
Applied Geography

Predicting geographic distribution and habitat suitability due to climate change of selected threatened forest tree species in the Philippines

https://doi.org/10.1016/j.apgeog.2013.07.005Get rights and content

Highlights

  • We modeled the potential distribution of threatened forest tree species in the Philippines due to climate change.

  • We found out that Maxent model performance is better than a random model.

  • Seven species were found to likely benefit from future climate due to the potential increase in their suitable habitat.

  • Results of the study could be used as basis in formulating appropriate adaptation policies, strategies and measures.

Abstract

Climate change is projected to alter the geographic distribution of forest ecosystems. This study aimed to evaluate the consequences of climate change on geographical distributions and habitat suitability of 14 threatened forest tree species in the Philippines. Based on the principle of maximum entropy, it utilized a machine algorithm called Maxent to estimate a target probability distribution and habitat suitability of the selected species. Threatened forest tree species occurrence records and sets of biophysical and bioclimatic variables were inputted to Maxent program to predict current and future distribution of the species. The Maxent models of the threatened species were evaluated using Receiver Operating Characteristics Area Under Curve (ROC AUC) and True Skill Statistics (TSS) tests which revealed that the models generated were better than random. The Maxent models ROC AUC values of the 14 species range from 0.70 to 0.972 which is higher than 0.5 of a null model. Based on TSS criteria, Maxent models performed good in two species, very good in ten species, and excellent in two species. Seven species (Afzelia rhomboidea; Koordersiodendron pinnatum; Mangifera altissima; Shorea contorta; Shorea palosapis; Shorea polysperma; Vitex parviflora) were found to likely benefit from future climate due to the potential increase in their suitable habitat while the other seven species (Agathis philippinensis; Celtis luzonica; Dipterocarpus grandiflorus; Shorea guiso; Shorea negrosensis; Toona calantas; Vatica mangachapoi) will likely experience decline in their suitable habitat. This study provided an initial understanding on how the distribution of threatened forest trees will be affected by climate change in the Philippines. The generated species distribution models and habitat suitability maps could be used as basis in formulating appropriate science-based adaptation policies, strategies and measures that could enhance the resilience of those threatened forest tree species and their natural ecosystems to current and future climate.

Introduction

The Philippines has very rich biodiversity in terms of number and percentage. It is regarded as one of 17 mega biodiversity countries accruing to its geographical isolation, diverse habitats and high rates of endemism (ASEAN Centre for Biodiversity, 2010, Protected Areas and Wildlife Bureau, 2009). It ranks 5th globally in terms of the number of plant species and maintains 5% of the world's flora. However, due to anthropogenic activities as well as natural disturbances, it continues to lose its rich biodiversity resources (Conservation International [CI], 2012). The presence of large number of endangered and threatened species brings down the country into one of the global biodiversity hotspots in the world, thus making it one of the top global conservation priority areas (ASEAN Centre for Biodiversity, 2010, International Union for Conservation of Nature, 2003, Myers et al., 2000). The official country listing of threatened plant species, Department of Environmental and Natural Resources Administrative Order Number [DAO] 2007-01 (2007), lists 176 vulnerable species, 99 critically endangered species, 187 endangered species and 64 other threatened plant species in the Philippines.

In addition to anthropogenic habitat alteration, climate change has been identified as well as one of the major threats facing biodiversity worldwide (ASEAN Centre for Biodiversity, 2010, Convention on Biological Diversity, 2010, Intergovernmental Panel on Climate Change, 2007, Millennium Ecosystem Assessment, 2005). It has been estimated that 20–30% of plant and animal species, globally, will be at higher risk of extinction due to global warming and that a significant proportion of endemic species may become extinct by 2050 or 2100 as a consequence as global mean temperatures exceed 2–3 °C above pre-industrial levels (Fischlin et al., 2007, Vié et al., 2008). As evidenced by the increasing mean temperature observed over time, climate change is said to be also occurring in the country (Philippines Atmospheric, Geophysical, and Astronomical Services Administration [PAGASA], 2011). Hence, there is a greater risk of extinction for species that are already vulnerable, particularly those with strict habitat requirements and restricted ranges (CBD, 2010).

An aspect that scientists and researchers look into to evaluate the potential impacts of climate change on natural systems is in the understanding of the likely shift in geographic distributions and habitat suitability of species due to future climate (Martinez-Meyer, 2005). In recent years, multiple modelling programs have been developed to predict the impacts of climate change on species distribution, even for areas that suffer from incomplete and biased samplings, or for areas where no collections have been made (Araujo and Guisan, 2006, Elith et al., 2006, Trisurat et al., 2011). Species-distribution models (SDMs) are based on the assumption that the relationship between a given pattern of interest (e.g. species abundance or presence/absence) and a set of factors assumed to control it can be quantified (Anderson et al., 2003, Anderson and Martinez-Meyer, 2004, Guisan and Zimmermann, 2000, Raxworthy et al., 2003). Maxent, one popular SDM based on presence-only modelling method involves maximum entropy modelling and has been used successfully to predict the distributions of different floral and faunal species (Baldwin, 2009, Trisurat et al., 2011, Weber, 2011). According to the study of Trisurat et al. (2011), from 26.6 °C 2008 to 28.7 °C in 2100 nine forest plant species have suitable distribution ranges in more than 15% of the region, 20 species have suitable ecological niches in less than 10% while the ecological niches of many Dipterocarpus species cover less than 1% of the region. Additionally, Maxent has been found to be robust to changes in sample size, and still have good predictive ability at low sample sizes, making it the ideal model for the prediction of distributions for rare species (Hernandez et al., 2006, Wisz et al., 2008).

Threats to vulnerable forest tree species and their habitat in the Philippines caused by anthropogenic activities and climate change could be exacerbated as well by modest research and conservation interest. Since 1997, only two studies on potential impacts of climate change on forests ecosystems in the Philippines have been conducted: Cruz (1997) and Lasco et al. (2008). Cruz (1997) hypothesized that tropical forest areas will likely expand as temperature and precipitation increase in many parts of the country. Temperature change may lead to a loss of a few species of plants and animals that may significantly erode the biodiversity of these forests. The study of Lasco et al. (2008) employed the Holdridge Life Zone system. Using GIS, they simulated how the forest types will change under increasing rainfall and temperature. The results demonstrated that dry forests are the most vulnerable terrestrial ecosystems in the Philippines which will disappear even with just 25% increase in rainfall. However, with regards to potential impacts of climate change on specific forest tree species, no research has been done yet in the country. In fact, a research gap identified by the Philippine Council for Agriculture, Forestry, and Natural Resources Research and Development (PCARRD) in 2010 that cross-cuts forest ecosystem, climate change and biodiversity is that future studies need to look at how climate change and the accompanying change in forest types will affect the biodiversity at the species level with special emphasis on rare, threatened, and endangered species.

With the research gap in mind, the main objective of this study was to assess the geographic distribution and habitat suitability due to climate change of selected threatened forest tree species in the Philippines. It is hoped that the results will be useful in developing science-based conservation strategies that could enhance the resilience of those threatened forest tree species and their natural ecosystems to current and future climate. This could provide options for the conservation of such species in areas that are continuously pressured by human activities and potentially by climate change. Information and maps describing habitat suitability and species distributions in such areas are essential for habitat conservation and species management.

Section snippets

Study area

The Philippines is located between 116° 40′, and 126° 34′ E. longitude and 4° 40′ and 21° 10′ N. latitude. It is bordered by the Philippine Sea to the east, the South China Sea to the west, and the Celebes Sea to the south. Relative to other countries in Southeast Asia, Philippines is located in east of Vietnam and north of Indonesia and Malaysia (Fig. 1). The Philippines has 7107 islands covering a total of 30 million ha. Owing to its archipelagic nature, topographic variations characterize

Variables with higher contribution to the geographical distribution of the 14 threatened forest tree species

Based on the results, the distributions of the threatened forest tree species are largely determined by biophysical variables (80%) more than bioclimatic variables (20%). Among the bioclimatic variables, Bio 2 (mean diurnal temperature range) has the highest average percent contribution (5.2%) in most of the 14 threatened forest tree species followed by Bio 6 (minimum temperature of coldest month) with average contribution of 4.6% (Fig. 3). This means that these two bioclimatic variables

Conclusions and recommendations

This study aimed to evaluate the consequences of climate change on geographical distributions and habitat suitability of selected 14 threatened forest tree species in the Philippines. Based on the principle of maximum entropy, it utilized a machine algorithm called Maxent to estimate a target probability distribution of the selected species. Three research questions were formulated as the framework of this study namely: 1) which variable has the highest contribution in the geographical

Acknowledgements

This research is part of the project Mainstreaming Climate Change in Biodiversity Planning and Management implemented by World Agroforestry Centre-Philippines and funded by the United States Agency for International Development.

References (56)

  • L. Beaumont et al.

    Predicting species distributions: use of climatic parameters in BIOCLIM and its impact on predictions of species' current and future distributions

    Ecological Modelling

    (2005)
  • B.M. Benito et al.

    Assessing extinction-risk of endangered plants using species distribution models: a case study of habitat depletion caused by the spread of greenhouses

    Biological Conservation

    (2009)
  • Conservation International

    Biodiversity hotspots: Philippines

    (2012)
  • Convention on Biological Diversity

    Global biodiversity outlook 3 (GBO-3)

    (2010)
  • R.V.O. Cruz

    Adaptation and mitigation measures for climate change: impacts on the forestry sector

  • Department of Environment and Natural Resources

    Establishing the national list of threatened Philippine plants and their categories, and the list of other wildlife species

    (2007)
  • J. Elith et al.

    Novel methods improve prediction of species' distributions from occurrence data

    Ecography

    (2006)
  • E.S. Fernando et al.

    Forest formations of the Philippines

    (2008)
  • A.H. Fielding et al.

    A review of methods for the measurement of prediction errors in conservation presence/absence models

    Environmental Conservation

    (1997)
  • A. Fischlin et al.

    Ecosystems, their properties, goods and services

  • Forest Management Bureau

    Philippine forestry statistics: forest cover within forest lands: 2003

    (2007)
  • C.H. Graham et al.

    A comparison of methods for mapping species ranges and species richness

    Global Ecology and Biogeography

    (2006)
  • A. Guisan et al.

    Predicting reptile distributions at the mesoscale: relation to climate and topography

    Journal of Biogeography

    (2003)
  • A. Guisan et al.

    GLM versus CCA spatial modelling of plant species distribution

    Plant Ecology

    (1999)
  • D.C.A. Guisan et al.

    Predictive habitat distribution models in ecology

    Ecological Modelling

    (2000)
  • P.A. Hernandez et al.

    The effect of sample size and species characteristics on performance of different species distribution modelling methods

    Ecography

    (2006)
  • C.J. Huberty

    Applied discriminant analysis

    (1994)
  • Intergovernmental Panel on Climate Change

    Summary for policymakers

  • Cited by (93)

    View all citing articles on Scopus
    View full text