Introduction

The island of Borneo suffers from some of the world's highest deforestation rates (Damayanti & Prasetyo, 2015; Margono, Potapov, Turubanova, Stolle, & Hansen, 2014), imperiling its thousands of endemic species and ranking it among the world's biodiversity hotspots (Corlett, 2014; Myers, Mittermeier, Mittermeier, da Fonseca, & Kent, 2000). Over half the island's forests have been converted to other land uses, including oil palm plantations and croplands (Fitzherbert et al., 2008; Zamzani, Onda, Yoshino, & Masuda, 2009), and forest loss is rampant even within protected areas (Curran et al., 2004; Margono et al., 2014). The long-term survival of Borneo's biodiversity thus depends on reforestation as a supplement to the creation of protected areas (Chazdon, 2008; International Tropical Timber Organization [ITTO], 1990; Kettle et al., 2011; Normile, 2009).

Much of Borneo's land area is potentially available for reforestation. Following widespread forest clearing and subsequent wildfires, hundreds of thousands of square kilometers of the island have become dominated by alang-alang (Imperata cylindrica) grassland (Garrity et al., 1997; ITTO, 1990; Kuusipalo et al., 1995). These areas have little value for native species and are unused economically (Garrity et al., 1997; Kuusipalo et al., 1995; Ohta, 1990), but if protected from fire can revert to natural forest (Yassir, van der Kamp, & Buurman, 2010). They thereby offer substantial reforestation opportunities with few costs to local communities. Tropical reforestation projects, however, may not replicate natural conditions or may fail altogether (Chazdon, 2008; Wuethrich, 2007). Our knowledge of Bornean reforestation progress, and how it benefits native species, remains especially limited (Kettle et al., 2011).

Here, we evaluate two community reforestation projects in Gunung Palung National Park in Indonesian Borneo. The park is a 108,000-ha protected area in West Kalimantan, consisting mostly of lowland dipterocarp rainforest (∼3,000 mm annual precipitation). The buffer zones around the park have lost nearly all their forest cover, and at least 10% of the park itself has been deforested due to illegal logging and conversion to agriculture (Curran et al., 2004; Zamzani et al., 2009). Open lands continue to encroach into the park, aided by frequent wildfires. In partnership with the park administration, the conservation and human health NGO Yayasan Alam Sehat Lestari (ASRI) has worked with local communities to reforest two areas on the park border (Pohnan, Ompusunggu, & Webb, 2015). We evaluate the success of those efforts using bird communities as indicators of restoration progress. Borneo is home to over 400 bird species, many of which play vital roles in ecological processes like seed dispersal, and their diversity is a useful indicator of forest quality (Edwards, Ansell, Ahmad, Nilus, & Hamer, 2009; Lambert & Collar, 2002; Maas et al., 2016). Bird species respond predictably to disturbance and vegetation structure (Lack, 1933), and their species richness increases as degraded areas are restored to natural vegetation (Edwards et al., 2009; Pei et al., 2018). Moreover, the small size of the reforestation sites (<20 ha) allows us to investigate impacts at a scale relevant to the conservation capacity of local communities and nonprofit organizations. Our results highlight the conservation value of secondary rainforest (Chazdon et al., 2009; Takano et al., 2014) and suggest that even small reforestation projects can provide rapid benefits for Borneo's native biodiversity.

Methods

Reforestation Sites

We monitored bird communities at two reforestation sites within Gunung Palung National Park (Figure 1(a)). The first site (1°22′15′′S, 110°13′31′′E, elevation 27 m) is a 23-ha area near the village of Laman Satong that was clearcut in the 1990s (Figure 1(b)). It lies on the park border, surrounded by selectively logged forest within the park and a road and agricultural land outside. After clearcutting, repeated wildfires destroyed the soil seed bank and remaining woody vegetation. By 2009, it was dominated by alang-alang grass and bracken ferns (Pteridium aquilinum), fire-tolerant natives that colonize disturbed areas and slow forest succession (Kuusipalo et al., 1995; Yassir et al., 2010). The land was then targeted for reforestation by ASRI and the national park. Each year from 2009 to 2015, at the beginning of the wet season from November to December, 2 to 7.5 ha were planted with a mix of ∼30 native tree species at a density of 1,300 to 4,000 seedlings per hectare. Seeds and seedlings were collected from forests in the national park or from community-owned forests, or from local citizens in exchange for health care subsidies at ASRI's private clinic. Replanted areas were weeded 3 to 4 times per year until seedlings were tall enough to overtop competing grass and ferns, and firebreaks were constructed to protect the site from fire. Site preparation, planting, and maintenance were conducted by over 400 local employees and community members, sometimes through outreach events aimed at informing communities about forest loss (Pohnan et al., 2015). The replanted area reached a peak of 19.8 ha (out of the 23 ha available) before being burned by a wildfire in 2013, leaving only 0.9 unburned hectares that had been planted in 2009. Planting efforts continued, with intensified fire prevention, and by 2015, the replanted area had recovered to 8.2 ha.

Figure 1.

(a) Our reforestation sites were in Gunung Palung National Park in West Kalimantan, Indonesia. (b) Aerial view of Laman Satong in 2009 before planting and (c) the same for Sedahan Jaya in 2012.

The second site (1°13′7.26′′S, 110°2′2.00′′E, elevation 15 m), ∼27 km away from the first and near the village of Sedahan Jaya, is a 6.5-ha freshwater swamp and peat forest (Figure 1(c)). It is surrounded by a mixture of rice paddies, fallow farmland, slash-and-burn gardens dominated by bananas and fruit trees, and selectively logged forest within the park. The site was cleared by farmers in 2010 for conversion to rice paddies. Parts of it were burned and farmed for 1 year, but then abandoned and left fallow. The soil seed bank was still partly intact in some areas and natural secondary succession began immediately, driven primarily by fast-growing Macaranga species. Other areas remained dominated by bracken ferns. ASRI and the national park then targeted the site for reforestation as part of a dispersal corridor for Bornean orangutans (Pongo pygmaeus). In 2013, they planted the understory of all 6.5 ha using a mix of ∼20 species at a density of ∼1,000 seedlings per hectare, followed by targeted enrichment plantings in 2014. Labor and maintenance were conducted by several dozen local employees and community members.

Results

Bird species richness increased rapidly at both sites in response to reforestation. We recorded a total of 101 bird species, 81 from Laman Satong and 69 from Sedahan Jaya (Table 1). As anthropogenic grassland at Laman Satong was succeeded by native rainforest (Figure 2(a) and (b)), bird species richness increased by 94% from 2011 to 2016 (from 36 to 70 species, Figure 3(a)), despite the limited scale of reforestation (<8 ha planted per year and <20 ha maximum extent). Annual increases in species richness were initially high (67% from 2011 to 2012), then declined due to the 2013 fire (−47% from 2012 to 2013) before rising again (59% from 2013 to 2014, average of 17% per year from 2014 to 2016). At the less degraded Sedahan Jaya site, species richness increased by 29% over a 3-year period beginning shortly after planting (from 49 to 63 species, Figure 3(b)), as early successional Macaranga groves and bracken fern began to mature into more diverse rainforest. Annual increases ranged from 4% (2013–2014) to 11% per year (average from 2014 to 2016). Across both sites and excluding the decline due to fire at Laman Satong, species richness increased by an average of 27 (±23, n = 7)% per year.

Table 1.

Birds Observed at Two Reforestation Sites in Gunung Palung National Park, West Kalimantan, Indonesian Borneo.

Figure 2.

Vegetation recovery at Laman Satong. (a) Before planting, the site was dominated by fire-tolerant grasses and ferns that impeded forest succession. (b) After several years of planting and fire prevention, unburnt areas were covered in young secondary forest.

Figure 3.

Bird species richness increased following reforestation efforts at (a) Laman Satong and (b) Sedahan Jaya. Open circles and dotted line show species richness. Solid circles and line show the area planted with native seedlings over the preceding year, and bars show cumulative reforested area at the time of the bird surveys. The 2013 dip in (a) is from a wildfire that burned most of the site.

The fire in 2013 at Laman Satong interrupted recovery and resulted in the disappearance of nearly half the observed bird species. The bird community recovered quickly, however, as unburnt forest continued to mature, surviving trees regrew, and new areas were planted. Just 3 years after the fire, species richness had surpassed its prefire peak in 2012 by 17% (70 vs. 60 species). Bird species richness initially tracked reforested area but then continued to rise despite lower planting rates and reforested area in postfire years. There was thus no relationship between species richness and reforested area (p = .42, r² = .22, n = 5) or site age (p = .24, r² = .42, n = 5), even when combining surveys from both sites (reforested area p = .33, r² = .16, n = 8; site age p = .28, r² = .19, n = 8). Recovery was instead a function of fire incidence and the resulting successional stage of the site.

Discussion

Reforestation can provide rapid benefits to native bird communities in Borneo. Our reforestation sites saw rises in bird species richness of up to 94% within a few years, with average annual increases of 27% in nonfire years. Final species richness estimates were comparable to those in 20-year-old secondary forests elsewhere in Borneo (Edwards et al., 2009), despite our sites being three orders of magnitude smaller (<20 vs. >10,000 ha). The rise in species richness is accompanied by other indicators of restoration success. Older trees at the sites regularly flower and produce fruits, and seedlings now germinate naturally from seeds dispersed by birds or primates. Feeding guild composition of the most recent surveys approached that of forest or mixed forest/agricultural bird communities at other tropical sites, with a predominance of insectivores and frugivores (Maas et al., 2016). Finally, in addition to birds, several mammals have been photographed at the sites, including orangutans at Sedahan Jaya. These positive responses were maintained despite temporary setbacks from a fire at one site, thanks to the protection of unburnt patches, regrowth of surviving trees, intensified fire prevention, and regular planting. Degraded and unused landscapes in Borneo can thus be restored to habitat for many native species, provided they receive appropriate long-term maintenance and protection from fire.

While rapid recovery of the bird community is heartening, it is worth noting that our sites contained fewer than the >200 species expected from intact primary rainforest (Edwards et al., 2009; Lambert & Collar, 2002), and not all the birds were forest specialists. We detected only two Borneo endemics (blue-headed Pitta Hydrornis baudii and the abundant generalist Dusky Munia Lonchura fuscans, compared to > 50 endemics across the island, Phillips & Phillips, 2009), and only two species of conservation concern (blue-headed Pitta and Greater Green Leafbird Chloropsis sonnerati, both listed as vulnerable). The sites also benefited from being in a nationally protected area and from bordering primary forest that served as a source of dispersing migrants. Our results may not hold for sites that are located far from natural forest remnants or are not strictly protected, or for threatened or endemic birds. Likewise, our failure to detect area effects on species richness may be an artifact of small size, lack of isolation, or presence of fire at our sites. Area effects may be more apparent at regional or landscape scales (Pei et al., 2018), in more isolated sites (Stratford & Stouffer, 2015), or in areas less impacted by fire. We thus emphasize the importance of protecting Borneo's remnant primary forests and suggest that reforestation is best viewed as a valuable supplement to the creation of protected areas.