Elsevier

Ecological Complexity

Volume 20, December 2014, Pages 307-314
Ecological Complexity

Original Research Article
Prediction of climate warming impacts on plant species could be more complex than expected. Evidence from a case study in the Himalaya

https://doi.org/10.1016/j.ecocom.2014.02.003Get rights and content

Highlights

  • Our findings here show that prospective distributions of plant species may diverge to a great extent from potential ones.

  • We provide a framework for the quantitative estimation of such divergence.

  • We show that predictions about climate change impacts on plant species are at risk of being excessively optimistic.

  • We conclude that predicting climate change impacts on plant species could result more complex than expected at present.

Abstract

Many studies have investigated the possible impact of climate change on the distributions of plant species. In the present study, we test whether the concept of potential distribution is able to effectively predict the impact of climate warming on plant species.

Using spatial simulation models, we related the actual (current species distribution), potential (modelled distribution assuming unlimited dispersal) and predicted (modelled distribution accounting for wind-limited seed dispersal) distributions of two plant species under several warming scenarios in the Sagarmatha National Park (Nepal). We found that the two predicted distributions were, respectively, seven and nine times smaller than the potential ones. Under a +3 °C scenario, both species would likely lose their actual and predicted distributions, while their potential distributions would remain partially safe. Our results emphasize that the predicted distributions of plant species may diverge to a great extent from their potential distributions, particularly in mountain areas, and predictions of species preservation in the face of climate warming based on the potential distributions of plant species are at risk of producing overoptimistic projections.

We conclude that the concept of potential distribution is likely to lead to limited or inefficacious conservation of plant species due to its excessively optimistic projections of species preservation. More robust strategies should utilize concepts such as “optimal reintroduction”, which maximizes the benefit–cost ratio of conservation activities by limiting reintroduction efforts to suitable areas that could not otherwise be reached by a species; moreover, such strategies maximize the probability of species establishment by excluding areas that will be endangered under future climate scenarios.

Introduction

Concern has escalated in recent years regarding the potential effects of climate change on species and ecosystems (Parmesan and Yohe, 2003, Gilman et al., 2010, Araújo et al., 2011). The IPCC (2007) remarked that future climate change is estimated to exacerbate the loss of species, especially of those taxa with strict climate requirements and limited migratory capabilities (Vittoz et al., 2009). Mountainous areas with cold, alpine climates have received particular interest in terms of changes in species distribution (Körner, 1999). Mountain ecosystems are susceptible to the impacts of a rapidly changing climate and provide interesting locations for the early detection and study of the signals of climatic change (Beniston, 2003). Nogués-Bravo et al. (2007) predicted that mountainous areas will experience unprecedented rates of warming during this century, two to three times greater than the rate observed during the previous century. The midlatitude mountains of Asia are expected to show the greatest increase in average temperature among the mid-latitude mountain systems of the world (Nogués-Bravo et al., 2007).

Although few data are available for the Himalayas (e.g., Giam et al., 2010), several works have focused their attention on the impacts of climatic variation on the plants of other alpine areas (Thuiller et al., 2005, Parolo and Rossi, 2008). In a study of approximately 85 subalpine and alpine nonwoody plants in the Austrian Alps, Dirnböck et al. (2003) predicted that 40–50% of these species could become extinct due to climate change in the next 50 years. Guisan and Theurillat (2000) predicted that nearly 40% of the 63 alpine and nival plant species in their study could lose more than 90% of their suitable habitat. Reductions in distribution or extinction are particularly likely for species with weak dispersal capacity (Engler et al., 2009), while efficiently dispersed species have a greater chance of rapidly responding to climate warming (Vittoz et al., 2009).

The major limitation of these models is that they tend to ignore dispersal restrictions by referring to the potential distributions of plant species (Engler et al., 2009). In an ideal environment, a species is expected to occupy a geographical area that strictly corresponds to its potential niche, thus occurring everywhere that environmental conditions are suitable (Pulliam, 2000). In reality, this potential distribution is unlikely to be observed, and the realized distribution is reduced from the potential due to abiotic (e.g., topographic barriers) and biotic factors (e.g., competition) (Scherrer and Körner, 2011). Studies based on the assumption of universal dispersal (i.e., a species has unlimited dispersal, its future distribution being the entire projection of its potential niche; Thomas et al., 2001) might provide good approximations for plants with high dispersal ability, but they likely overestimate the future distributions of many other species. For example, in the alpine environment, wind is one of the major factor influencing seed dispersal (Tackenberg and Stocklin, 2008); parts of a plant's potential distribution may therefore remain uninhabited, despite their local suitability, as a consequence of dispersal limitations due to topographic barriers and wind behaviour (Pulliam, 2000, Parolo et al., 2008). While the unlimited dispersal assumption represents an optimistic best-case scenario, some studies have also provided a worst case no-dispersal scenario (Thuiller et al., 2005) to establish a lower bound for their projections. As noted by Bellard et al. (2012), this scenario is clearly convenient for practical purposes, but most species fall somewhere between these two extremes. In addition, the difference between these extreme projections can generate heavy uncertainties (Thuiller et al., 2004). Reducing these uncertainties requires the consideration of dispersal processes, but few studies to date have included dispersal limitations when projecting species distribution under climate change scenarios (Dullinger et al., 2004, Midgley et al., 2006, Engler et al., 2009).

Accordingly, this work has the following goals: (1) to emphasize the empirical (not theoretical) differences among the actual (realized), predicted (dispersal-restricted) and potential (dispersal-unlimited) distributions of plant species; (2) to propose a methodology based on spatial simulation modelling for the individuation of the predicted distributions of wind-dispersed species; (3) to quantify the deviation between plant species survival probabilities in the face of climate warming as estimated using both potential and predicted distributions; (4) to test our approach using two plant species in the Himalayan mountain system.

Overall, we aim to test if the concept of potential distribution is able to effectively predict the impacts of climate warming on the plant species in our case study. The answer to this question has great implications for conservation.

Section snippets

Study area and study species

The study area (Fig. 1) corresponds to the Khumbu Valley, which lies in the Sagarmatha National Park (SNP; northeastern Nepal, Solukhumbu District). Data on the climate of this region has been collected at the Pyramid Meteorological Station since 1994 (Bollasina et al., 2002, Diodato et al., 2011). The study area covers a 30.55 km2 portion of the national park and ranges from 4907 to 5913 m a.s.l. in altitude. The area is divided into a lower alpine belt, dominated by shrubs of dwarf

Field sampling and basic statistics

Floristic surveys were conducted in autumn 2010. Although the field accessibility of these mountainous areas is very limited, we were able to collect at approximately 150 locations for each species with systematic, extensive sampling. The coordinates of each site were measured with a global positioning system (GPS) device using differential correction techniques to improve the accuracy of the data locations (error < 1 m). The actual species distribution (ACT) was estimated as the convex hull (area

Results

The climatic statistics are summarized in Fig. 2. Based on the dominating winds, the PRED assessment for each species required five dispersal simulations (Fig. 3), two for the southern part (Pyramid station) of the study area and three for the northern one (Kala Pattar). The resulting regression equation between elevation (X) and the 99th percentiles of T° (Y) was as follows: Y = 33.9599  0.0053 × X.

The two species have an identical ACT due to the very spatially proximate coordinates of their

Discussion

We found that the two species have the same ACT. This result is not surprising, as both species: (a) occupy a similar ecological niche; (b) are completely dependent on winds for their dispersal; and (c) have the same anatomical dispersal structures (achene-pappus units). During the calibration phase, we found a Vrel of 1.4 m/s, which is in agreement with the experimental and simulated results of non-tree species (Tackenberg, 2003), although reported release velocities range between 0.1 and 6.9 

Conclusions

Climate change ecology is still in its early stages, and enormous improvements are rapidly being made in virtually all aspects of this emerging field. Among the critical requirements for predicting future trends is the need to overcome several model limitations.

Accordingly, we have proposed here a conceptual and operative framework based on spatial simulation models for determining the future spatial distributions of wind-dispersed plant species. We have also stressed the empirical differences

Acknowledgments

This work was supported by the Ev-K2-CNR (Bergamo, Italy) SHARE project. The authors would like to express their thanks to Sanu Raja Maharjan (Kathmandu) and Dr. Oliviero Spettoli (Parco Regionale Oglio Sud) for their help in field activities. The authors also thank NAST (Nepal Academy of Science and Technology) for the collaboration. We acknowledge the anonymous referees for their constructive and helpful comments.

References (48)

  • M. Beniston

    Climatic change in mountain regions: a review of possible impacts

    Clim. Change

    (2003)
  • M. Bollasina et al.

    Meteorological observations at high altitude in the Khumbu Valley, Nepal, Himalayas, 1994–1999

    Bull. Glaciol. Res.

    (2002)
  • Y.S. Chen et al.

    Gnaphalieae

  • C.G. Davis et al.

    Atmospheric transport models for complex terrain

    J. Clim. Appl. Meteorol.

    (1984)
  • N. Diodato et al.

    How do Himalayan areas respond to global warming?

    Int. J. Climatol.

    (2011)
  • T. Dirnböck et al.

    A regional impact assessment of climate and land-use change on alpine vegetation

    J. Biogeogr.

    (2003)
  • S. Dullinger et al.

    Modelling climate change-driven treeline shifts: relative effects of temperature increase, dispersal and invisibility

    J. Ecol.

    (2004)
  • R. Engler et al.

    Predicting future distributions of mountain plants under climate change: does dispersal capacity matter?

    Ecography

    (2009)
  • K.J. Feeley et al.

    Biotic attrition from tropical forests correcting for truncated temperature niches

    Global Change Biol.

    (2010)
  • A. Guisan et al.

    Equilibrium modelling of alpine plant distribution: how far can we go?

    Phytocoenologia

    (2000)
  • S.I. Higgins et al.

    Predicting plant migration rates in a changing world: the role of long-distance dispersal

    Am. Nat.

    (1999)
  • B. Holzinger et al.

    Changes in plant species richness over the last century in the eastern Swiss Alps: elevational gradient, bedrock effects and migration rates

    Plant Ecol.

    (2008)
  • IPCC

    Climate Change 2007: Synthesis Report. An Assessment of the Intergovernmental Panel on Climate Change

    (2007)
  • L.R. Iverson et al.

    Predicting abundances of 80 tree species following climate change in the eastern United States

    Ecol. Monogr.

    (1998)
  • Cited by (0)

    View full text