Carbon emissions and potential emissions reductions from low-intensity selective logging in southwestern Amazonia
Introduction
Globally, approximately 11% of annual net greenhouse gas emissions and 14% of carbon emissions are from forestry and other land uses, mostly in developing tropical and subtropical countries (Goodman and Herold, 2014). Reducing tropical deforestation and degradation have long been considered important to reduce global carbon emissions, but the contribution of forest degradation has only recently been quantified over large scales (Berenguer et al., 2014, Baccini et al., 2017, Erb et al., 2017). Forest degradation reportedly accounts for one quarter (Pearson et al., 2017) to over two thirds (Baccini et al., 2017) of all forest emissions in tropical countries. In Central and South America, half (Hosonuma et al., 2012) to two thirds (Pearson et al., 2017) of degradation emissions are from logging. Both reducing carbon emissions from forestry operations and ensuring sustainability of forest management are important to address given that over half of all remaining tropical forests are dedicated to wood production (Blaser et al., 2011).
Working with forest management enterprises (e.g., concessions and community-based forest management) has the potential to improve conservation outcomes by reducing degradation through improved harvest practices and reducing deforestation by creating profitable business models based on retaining forests as forests (Griscom and Goodman, 2015). Natural forest management and avoided forest conversion are both high-potential and low-cost “natural climate solutions” to mitigate climate change (Griscom et al., 2017). At the same time, unnecessarily destructive logging is a major form of degradation that generally precedes and can even promote deforestation (Asner et al., 2006).
Due to high diversity and the marketability of timber from only a few tree species, logging in the tropics is typically selective but can nevertheless result in substantial carbon emissions. Carbon losses from selective logging come from clearing for haul roads and log yards, collateral damage around felled trees and on skid trails, crop tree residuals (i.e., branches, stumps, and sections of the stem left in the forest, etc), and the life cycle of the extracted wood itself. By employing what have become known as reduced-impact logging (RIL) practices (Putz and Pinard, 1993), these emissions can be substantially reduced (Johns et al., 1996, Pinard and Putz, 1996, West et al., 2014). Recommended RIL practices include planning harvest operations (e.g., mapping and marking commercial trees; planning roads, log yards, and skid trails; and using directional felling techniques to avoid damage to future crop trees and streams), cutting lianas on trees to be harvested at least 6 months before felling, and following guidelines on tree felling and skidding (Pinard et al., 1995, Dykstra and Heinrich, 1996). RIL guidelines typically also include practices intended to reduce the biodiversity and hydrological impacts of logging and to improve worker safety, but we here focus on only those practices likely to reduce carbon emissions (RIL-C; Griscom et al., 2014). In this study we also disregard the substantial post-logging carbon benefits of RIL such as increased rates of carbon stock recovery (Lincoln, 2008, Vidal et al., 2016).
Despite the large amount of research conducted on RIL across the tropics (FAO, 2004), we are aware of no studies on RIL in Peru. Given that logging contributes substantially to Peru’s carbon emissions, which the country has committed to reduce (MINAM, 2015), we conducted this study to establish a baseline from which improvements can be measured. Furthermore, while Peru is globally known for its deforestation (Asner et al., 2010, Hansen et al., 2013, Robiglio et al., 2014) and problems with illegal logging (Gutierrez-Velez and MacDicken, 2008, Finer et al., 2014), the country also hosts efforts to improve forest management practices that deserve attention.
Among the interventions intended to promote responsible forest management in general and RIL in particular, Forest Stewardship Council (FSC) certification looms large. Since its founding in 1993, numerous studies reported on FSC’s impacts but few were designed to avoid positive selection biases, ignored contextual changes of likely importance, and suffered from other deficiencies such as small sample sizes (Romero et al., 2017, Komives et al., 2018). In a meta-analysis of this literature, Blackman et al. (2017) found no conclusive evidence that forest certification has positive environmental outcomes. In another meta-analysis that disregarded the quality of the included studies, Burivalova et al. (2017) reported that the environmental outcomes from certified and RIL forest management (including C emissions) were better than conventional management in 76% of case studies (worse in 6% and no difference in 18%). However, this and other studies also concluded that the benefits of RIL decline when logging intensity is taken into account because certified concessions and those that claim to employ RIL practices tend to harvest at lower intensities than their uncertified and conventionally logged counterparts (Medjibe et al., 2013, Griscom et al., 2014, Martin et al., 2015, Burivalova et al., 2017).
The Peruvian Amazon is an important focal geography where many global issues are currently at play. Peru has fourth highest area of forest cover in tropics, and nearly half of the country’s 68 million ha are classified as permanent production forest (Blaser et al., 2011). The southwestern Amazon and specifically the MAP region (Madre de Dios, Peru–Acre, Brazil–Pando, Bolivia) is undergoing rapid deforestation and forest degradation following completion of the Interoceanic Highway (Baraloto et al., 2015, Alarcón et al., 2016), and deforestation and degradation of natural forests represents almost half of Peru’s greenhouse gas emissions (MINAM, 2013). Legal and illegal logging has already degraded much of the natural forest in the Peruvian Amazon (Asner et al., 2010), and the legal timber industry is expanding (Cossio et al., 2014).
We employed the methods of Griscom et al. (2014) to estimate carbon emissions from forestry concessions in Madre de Dios, Peru. Specifically, our objectives were to: (i) establish a baseline for carbon emissions from selective logging in the region; (ii) assess whether RIL training associated with FSC certification reduced carbon emissions; and (iii) estimate potential emissions reductions through RIL-C. Because of the low logging intensity (Cossio et al., 2014) and distinctive architecture of trees in southern Peruvian forests (relatively short stems and large crowns; Goodman et al., 2014), we hypothesized baseline emissions per ha to be lower and emissions per unit timber extracted to be higher than the tropical average. As field staff in all certified concessions were trained in RIL while those employed by non-certified concessions were not, we expected emissions from certified concessions to be lower while fully recognizing that any differences cannot be attributed to certification due to other differences among the concessions and lack of a counterfactual design.
Section snippets
Site and socio-economic description
Our study was conducted within Tahuamanu Province of Madre de Dios, Peru. Forests here are broadly classified as lowland, moist, terra firme forest (Whitmore, 1998, Achard et al., 2002), and “bamboo-dominated” forests are common in this region (Carvalho et al., 2013). Mean annual temperature is 24.5 °C; mean annual precipitation is 1811 mm with a 3–4 month dry season (Hijmans et al., 2005). The area is relatively flat with medium gradient hills and elevation ca. 250–375 masl (FAO et al., 1998).
Logging practices
All FMUs were logged at low intensities (2.9–8.1 m3 ha−1) and harvested only 15–62% of government-authorized volumes (Table 2). All five certified FMUs harvested at lower intensities than the four non-certified FMUs in terms of trees (p = 0.045) and volumes (p = 0.017) per ha. Felled trees were on average 101.8 (standard error 2.3) cm dbh and 38.6 (0.6) m tall. Mean extracted logs were 15.6 (0.3) m long, 11.3 (0.6) m3, and 4.3 (0.3) Mg C. Certified concessions were much larger than
Logging practices and emissions baselines
Harvest intensities for all concessions in our study were much lower than reported for other tropical forests, with the exception of other parts of the southwestern Amazon (Rutishauser et al., 2015) and Gabon (Medjibe et al., 2013). As found elsewhere (e.g., Blackman et al., 2018), certified concessions were much larger than non-certified concessions, which is no surprise given that the costs of certification (Ruslandi et al., 2014) would be prohibitive for small concessions.
Like prior studies,
Conclusions and recommendations
From a carbon perspective, Madre de Dios in the southwestern Amazon has among the best carbon outcomes from selective logging of all tropical countries studied. Nonetheless, three times more carbon is emitted from logging operations than is extracted, and this amount could be reduced by half through RIL-C harvesting practices. FSC certification was not created specifically to reduce carbon emissions, and we find little evidence that it does so in Peru. We call for a clearer link between
Acknowledgements
We thank our forestry technicians and research assistants, Robert Nishida, Nilton Martinez, and Angel Balarezo for their hard work. Fieldwork was supported by WWF–Peru, and we thank Cecilia Alvarez, Rafael Venegas, Juan Carlos Riveros, Karen Mo, Alonso Gonzales, and Karina Salas for coordination of field activities. RCG and MHA were funded by WWF-Peru during the field campaign. All analysis, interpretation, and preparation of the manuscript were conducted independently.
References (81)
- et al.
Effects of road infrastructure on forest value across a tri-national Amazonian frontier
Biol. Conserv.
(2015) - et al.
Improved timber harvest techniques maintain biodiversity in tropical forests
Curr. Biol.
(2014) - et al.
Does eco-certification stem tropical deforestation? Forest Stewardship Council certification in Mexico
J. Environ. Econ. Manage.
(2018) - et al.
The Mexican model of community forest management: the role of agrarian policy, forest policy and entrepreneurial organization
For. Pol. Econ.
(2006) - et al.
Thresholds of logging intensity to maintain tropical forest biodiversity
Curr. Biol.
(2014) - et al.
Mapping selective logging impacts in Borneo with GPS and airborne lidar
For. Ecol. Manage.
(2016) - et al.
Quantifying the direct social and governmental costs of illegal logging in the Bolivian, Brazilian, and Peruvian Amazon
For. Pol. Econ.
(2008) - et al.
Logging damage during planned and unplanned logging operations in the eastern Amazon
For. Ecol. Manage.
(1996) - et al.
The sustainability of timber production from Eastern Amazonian forests
Land Use Policy
(2012) - et al.
Logging in bamboo-dominated forests in southwestern Amazonia: caveats and opportunities for smallholder forest management
For. Ecol. Manage.
(2014)
Rapid tree carbon stock recovery in managed Amazonian forests
Curr. Biol.
Fates of trees damaged by logging in Amazonian Bolivia
For. Ecol. Manage.
Sustainability of reduced-impact logging in the Eastern Amazon
For. Ecol. Manage.
Large trees as key elements of carbon storage and dynamics after selective logging in the Eastern Amazon
For. Ecol. Manage.
Recovery of biomass and merchantable timber volumes twenty years after conventional and reduced-impact logging in Amazonian Brazil
For. Ecol. Manage.
Forest biomass recovery after conventional and reduced-impact logging in Amazonian Brazil
For. Ecol. Manage.
Tropical forest cover monitoring in the humid tropics: TREES project
Trop. Ecol.
Deforestación en el sureste de la amazonia del Perú entre los años 1999–2013; caso Regional de Madre de Dios (Puerto Maldonado – Inambari)
Revista de Investigaciones Altoandinas
Condition and fate of logged forests in the Brazilian Amazon
PNAS
High-resolution forest carbon stocks and emissions in the Amazon
PNAS
Tropical forests are a net carbon source based on aboveground measurements of gain and loss
Science
Variation in wood density determines spatial patterns in Amazonian forest biomass
Global Change Biol.
A large-scale field assessment of carbon stocks in human-modified tropical forests
Glob. Change Biol.
Does forest certification in developing countries have environmental benefits? Insights from Mexican corrective action requests
Int. Forest. Rev.
A critical comparison of conventional, certified, and community management of tropical forests for timber in terms of environmental, economic, and social variables
Conserv. Lett.
Bamboo-dominated forests of the southwest Amazon: detection, spatial extent, life cycle length and flowering waves
PLoS ONE
Regional deforestation trends within local realities: land-cover change in southeastern Peru 1996–2011
Land
Synergisms among fire, land use, and climate change in the Amazon
SPIE
Logging Practices: Principles of Timber Harvesting Systems
FAO Model Code of Forest Harvesting Practice
Timing of carbon emissions from global forest clearance
Nat Clim Change
Unexpectedly large impact of forest management and grazing on global vegetation biomass
Nature
Logging concessions enable illegal logging crisis in the Peruvian Amazon
Sci. Rep.
Part 1: Analysis of data and methodologies of the biomass plots installed in Madre de Dios
Consultancy to establish the technical guidelines to complete a carbon baseline and design a measurement and monitoring system for carbon stocks in the region of Madre de Dios, Peru
Cited by (30)
Aboveground biomass allometric models for large trees in southwestern Amazonia
2022, Trees, Forests and PeopleCitation Excerpt :As applied in Brazil, commercial trees with DBH ≥ 50 cm diameter are measured in order to obtain information on the quantity and quality of forest resources for wood production (Husch et al., 2003; Romero et al., 2020a, 2020b). The measurement unit used for management calculations and for log sales is volume in cubic meters (m3) (Husch et al., 2003; Goodman et al., 2019; Romero et al., 2020b, 2021). The stock potential of managed commercial species can be obtained from forest inventories that provide the data for estimating volume, biomass and forest carbon stocks (Husch et al., 2003; Chave et al., 2014; Vidal et al., 2016).
Carbon payments can cost-effectively improve logging sustainability in the Amazon
2022, Journal of Environmental ManagementCitation Excerpt :Whilst forest management in such areas can be sustainable and profitable (Humphries et al., 2012), and rates of degradation are often reduced (Sze et al., 2021), our economic and yield data do not allow us to address the potential of REDD + to work in these areas. Finally, we considered harvest emissions through loss of live biomass carbon, but did not account for associated emissions from heavy machinery and vehicles (Pearson et al., 2017), additional damage through skid trails construction (Goodman et al., 2019), or changes to forest soil carbon (Riutta et al., 2021). These would likely increase carbon emissions saved under restricted harvests, meaning our breakeven carbon prices may be slightly overestimated.
An unsupervised domain adaptation approach for change detection and its application to deforestation mapping in tropical biomes
2021, ISPRS Journal of Photogrammetry and Remote SensingCitation Excerpt :In this respect, deforestation of natural forests constitutes one of the largest sources of greenhouse gas emissions, being responsible for the reduction of carbon storage and for invaluable losses in biodiversity (De Sy et al., 2015). The Amazon forest, for instance, which contains approximately 10% of all the biomass on the planet, and a similar proportion of all living species, has been facing serious threats as a result of the impact of unsustainable economic activities (Goodman et al., 2019; Malingreau et al., 2012; Nogueron et al., 2006). Therefore, monitoring Earth’s surface changes, especially those that adversely impact the environment, has become a priority for a number of authorities around the world.
Prospective carbon balance of the wood sector in a tropical forest territory using a temporally-explicit model
2021, Forest Ecology and ManagementCitation Excerpt :Griscom et al. (2014) and Feldpausch et al. (2005) have proposed such a method accounting for emission due to timber extraction, logging damage and logging infrastructure, which has since been widely used (Ellis et al., 2019). Carbon emissions are expressed relatively to the amount of wood produced (using the Total Emission Factor (Griscom et al., 2014; Pearson et al., 2014; Zalman et al., 2019) and the Carbon Impact factor (Ellis et al., 2019; Goodman et al., 2019; Umunay et al., 2019)), which is particularly interesting to compare the relative importance of these different sources of emissions and the efficiency of different logging practices for climate change mitigation (see the recent special issue on RIL-C (Ellis and Putz, 2019)). However, both remote sensing and previously cited field-based approaches use a committed emissions approach, meaning that they only account for fluxes at the time of logging (Fearnside, 1997), whereas carbon fluxes continue long after logging operations.