• Open Access

Cryptic destruction of India's native forests

Authors


  • Editor
    Corey Bradshaw

Correspondence
William F. Laurance, School of Marine and Tropical Biology, James Cook University, Cairns, Queensland 4870, Australia. Tel: +61-7-4042-1819; fax: +61-7-4781-5511. E-mail: bill.laurance@jcu.edu.au

Abstract

India sustains some of the world's most imperiled forests. The Forest Survey of India recently announced that forest cover in India had expanded by nearly 5% over the past decade. This result, while technically accurate, is misleading. The Forest Survey estimates forest cover by using automated algorithms to analyze satellite imagery—an approach that fails to distinguish native forests from tree plantations, which are often monocultures of exotic species that have limited value for endangered biodiversity. Since the early 1990s, tree plantations have expanded in India at an estimated rate of roughly 15,400 km2/year. Subtracting plantations from total forest cover shows that native forests in India have declined by 1.5%–2.7% per year. The limited precision of our estimate highlights a paucity of data on native forest cover in India—a problem requiring urgent attention. Forest cutting for fuelwood has been the biggest driver of forest loss and thinning in India. Like India, many nations now rely on satellite imagery to discern changes in vegetation cover, and these frequently lump native, exotic, and degraded forest types. Without sufficiently high-resolution imagery and adequate safeguards, such approaches could paint a misleading picture of the fate of the world's native forests.

Introduction

India is among the world's most biologically important nations, sustaining parts of four recognized biodiversity hotspots within its boundaries. These include the wet tropical forests of the Western Ghats and Sri Lanka, the monsoonal forests of Indo-Burma, the montane forests of the Himalayas, and the Sundaland forests of the southern Nicobar Islands (Mittermeier et al. 2004). Other areas of biologically rich forest are found in central India and the Eastern Ghats (Rodgers & Panwar 1988). Many endangered and locally endemic species rely on these ecosystems for survival (Ramesh et al. 1997; Alfred 1998; Karanth et al. 2010).

How are these crucial forests faring? In 2009, the Forest Survey of India (FSI) announced ostensibly good news: Indian forests had expanded by nearly 5% over the preceding decade (FSI 2009). Unfortunately, this result is biologically misleading because it fails to discriminate natural or native forests from large expanses of exotic vegetation, thereby obscuring the fate of the former. Here we critically evaluate forest trends in India and argue that native forests are actually being lost, degraded, or transformed at an alarming pace. We also suggest strategies to improve the monitoring of native forests and to slow forest disruption and its resulting atmospheric carbon emissions.

Evidence for native forest decline

We base our argument on three lines of evidence. First, regional and local studies suggest that native forests are declining in many parts of India. For instance, native forests in the Western Ghats declined at a mean rate of 0.99% per year from 1973 to 1995 (Jha et al. 2000), whereas those of the Himalayas shrank at 1.18% per year between 1990 and 2000 (Pandit et al. 2007). Forests in both regions are relatively well-protected. Native forests in other areas, especially in central India, are less-effectively protected, and thus likely to suffer even higher rates of forest loss. Some ecosystems, such as the tropical dry evergreen forest along the coast of Tamil Nadu, have nearly vanished from historic and contemporary degradation (Meher-Homji 1996).

Second, the fate of native forests in India is being obscured by the rapid expansion of plantations, which are included in estimates of total forest cover. The best estimates of forest cover in India were generated by the FSI (FSI 2009), whereby any land exceeding 1 ha with ≥10% tree cover is classified as “forest cover.” Advances in remote-sensing technology and computerized algorithms are facilitating increasingly accurate forest-cover assessments (FSI 2009). The FSI (2009) data suggest that total forest cover in India has increased almost linearly over time, at a mean rate of 2,990 km2/year from 1994 to 2006 (Figure 1).

Figure 1.

Trends in total forest cover and plantation area in India. Forest cover increased at a mean rate of nearly 3,000 km2/year from 1994 to 2006 (forest cover = 2990 × year − 5305; R2= 90.0%, P= 0.004; linear regression) but plantations expanded far more rapidly, suggesting a marked decline in the area of native forest.

However, the FSI assessment does not distinguish native forests from tree plantations—often monocultures of mostly exotic species such as eucalyptus, acacia, rubber, teak, bamboo, or pine trees (according to FAO (2001), nearly 45% of India's forest plantations are fast-growing, short-rotation species, such as Eucalyptus grandis, E. tereticornis, Acacia auriculiformis, A. mearnsii, and A. nilotica). Since 1992, plantations have expanded rapidly in India via state-sponsored programs designed to help meet burgeoning demands for timber and fuelwood and to reforest denuded hillsides to limit downslope flooding (Laurance 2007a; ITTO 2009). Unfortunately, plantation monocultures generally have limited value for conserving endangered biodiversity (Cannell 1999; Barlow et al. 2007; Gardner et al. 2007; Shahabuddin & Rao 2010).

India now ranks second globally in the total land area under plantation (ITTO 2009), but determining the precise rate of plantation expansion is challenging because data are incomplete and scattered among different Indian states. The best available data on Indian plantations were collated by the FAO (1997, 2005, 2007) based on country reports and other sources (under FAO criteria, “forest plantation” includes any type of forest except natural forest, including the predominant wood-production plantations as well as conservation plantations and planted forests). The FAO data, which are less accurate than are the FSI data on forest cover, suggest that plantations in India increased at a mean annual rate of ∼15,400 km2/year from 1995 to 2005 (Figure 1).

By subtracting the estimated rates of plantation expansion from that of forest-cover increase, it is possible broadly to infer the rate at which native forests are declining in India. Our linear regression (Figure 1) suggests that total forest cover rose from 660,337 km2 in 1995 to 690,250 km2 in 2005, whereas plantations expanded from 146,200 km2 to 300,280 km2 over this same interval according to FAO data. Consequently, native forests are estimated to have declined from 514,137 km2 in 1995 to 389,970 km2 in 2005, which translates into a mean percentage loss of 2.42% per year. This rate is somewhat lower than that derived by Puyravaud et al. (2010) because FAO estimates of plantation areas were gleaned from the original FAO (1997, 2005, 2007) reports rather than secondary sources (ITTO 2009).

Our final strategy for assessing native forests is to evaluate recent changes in forest biovolume. Biovolume, comprised by wood and other aboveground forest material, is estimated from intensive field observations (FSI 2009). From 2002 to 2006, the total area of forest increased by 0.60% whereas forest biovolume declined by 5.92% (FSI 2003, 2005, 2009), at an average rate of 70 million m3/year. Assuming a typical biovolume (67.14 m3) for forests in India (averaged from values in FSI 2003, 2005, 2009), the loss is equivalent to clearing 10,426 km2 of forest per year. If one assumes this clearing affected native and nonnative forest (collectively encompassing 690,250 km2 in 2005) at equivalent intensity, then this corresponds to a net deforestation rate of 1.51% per year. However, if the decline in biovolume arose solely from the loss or thinning of native forests (because plantations continued to expand over this period), then this translates into a mean clearing rate of 2.67% per year for native forests (assuming 389,970 km2 of native forest was present in 2005, as estimated above). This second value accords more closely with our estimated rate of native forest loss (2.42% per year) calculated by subtracting estimated plantation area from total forest cover.

Forest cutting for fuelwood (Figure 2) is clearly the greatest driver of forest loss and attrition in India (Rath 2002; Sagar & Singh 2004; Davidar et al. 2007, 2010). About 247 million people in the country rely mostly on natural forests for part of their subsistence or cash livelihoods, and two-thirds of these use fuelwood as an energy source, primarily for cooking and heating (World Bank 2006). Rural households generally require 1.5–2.0 kg of dry fuel per person each day (Pandey 2002). In total, an estimated 94.6 million metric tons of fuelwood are being harvested annually in India (Pandey 2002), and additional wood is consumed for building construction, fencing, and other uses.

Figure 2.

Montane rainforest and fuelwood-gathering in the Western Ghats (photos by W. Laurance). Wild tigers in Bandhavgarh forest (photo by P. Davidar).

Policy implications

India has already lost nearly 80% of its original native forest cover (Laurance 2007b). Using the best available datasets, our analyses suggest that remaining native forests in India are declining at a rapid pace—from 1.5% to 2.7% per year. The relative imprecision of our deforestation estimate is caused by a serious paucity of reliable data about the extent and condition of native forests in India. In addition to outright forest loss, large expanses of Indian forests are being thinned or degraded in various ways, such as by the overharvesting of fuelwood and nontimber products, overgrazing by livestock, illegal logging, and exotic-species invasions (Sinha & Bawa 2001; Sagar & Singh 2004; Davidar et al. 2007, 2010).

We urge progress in two key areas. First, available remote-sensing data could be used far more effectively to monitor the condition of Indian native forests (e.g., Sagar & Singh 2004). Broad forest-cover assessments, such as those of the FAO (2005), often merge cover from native forests and plantations because they rely on coarse-resolution imagery (e.g., ∼1.1 km pixels from the Advanced Very High Resolution Radiometer [AVHRR] satellites). The FSI, however, uses data from Indian Remote Sensing (IRS) satellites that provide far greater resolution: 23.5 m in the multispectral mode and 5.8 m in the panchromatic mode (Roy et al. 1996). Given this high resolution, total forest cover in India is likely estimated with impressive accuracy. The IRS sensors can, in fact, be used to distinguish most plantations from native forests (e.g., FSI 2009: 24) but this requires additional effort, including more refined algorithms that employ spectral modeling data, spatially explicit analyses of time-series data, and careful validation of forest classifications with ground truthing (cf. Turner et al. 2003; Chambers et al. 2007). Our plea for improved forest classification echoes the earlier conclusions of Menon & Bawa (1998), who also found large disparities in forest definitions and estimated deforestation rates in India, and who likewise highlighted the need for improved remote-sensing and ground-truthing data to discriminate native and plantation forests.

Second, urgent attention should be focused on providing alternative energy sources to fuelwood, such as natural gas, biogas, and electricity for rural populations in India (Agoramoorthy & Hsu 2008, 2009). Providing more fuel-efficient, cleaner-burning stoves (e.g., http://www.bioenergylists.org/taxonomy/term/100) might also reduce dependence on forests. This would not only help to alleviate pressures on some of the world's most biologically important forests, but would reduce the massive emissions of heat-absorbing black soot and greenhouse gases from millions of traditional wood-burning fires in India. Such emissions likely contribute substantially to global warming (Venkataraman et al. 2005).

The cryptic destruction of India's native forests highlights a key challenge confronting those attempting to understand current trends in forest ecosystems. Data from satellites are increasingly used to monitor changes in net forest cover (e.g., FSI 2003, 2005, 2009; FAO 2005, 2006), but such estimates usually pool old-growth native forests, secondary regrowth, and exotic plantations, among others. More generally, the distinction between native and modified forests is increasingly being blurred. In Peninsular Malaysia, for example, species-rich native forests are rapidly being replaced by monocultures of exotic rubberwood trees, yet these still are legally classified as permanent forest reserves (Aziz et al. 2010). A failure to discern such changes could paint a highly misleading picture of the fate of the world's native forests.

Acknowledgements

We thank G. R. Clements, S. G. Laurance, and two anonymous referees for helpful comments on the manuscript.

Ancillary