Accounting for carbon stocks in models of land-use change: an application to Southern Yucatan

To assess the impact of land-use change on carbon stocks, we apply a new methodology, linking ecological and economic modeling, to southern Yucatan, Mexico. A spatial econometric multinomial logit model of ten land-cover classes is estimated (four primary forest categories, three secondary growth categories, an invasive species, and two agricultural land-cover categories), using satellite data on land cover, linked with census socioeconomic data and other biophysical spatial data from 2000. The analysis is novel in that it is the first attempt to link detailed satellite data on land use, with on-the-ground estimates of carbon stocks in a spatial econometric model of land use. The estimated multinomial logit model is then used with two scenarios of future economic growth (“low growth” and “high growth” changes in population, agricultural land use, market access, and education levels) in the region to predict land-cover changes resulting from the economic growth. The per hectare carbon (C) stocks in each land-cover class are derived from previously published estimates of biomass from field sampling across the study region. We consider aboveground-only, aboveground plus soil, transient and non-transient pools of carbon. These estimates are scaled up to the total area in each class according to the predictions of the model baseline and the two development scenarios. Subsequently, the changes in carbon stocks resulting from the predicted land-cover changes are calculated. Under the low growth scenario, carbon stocks declined by 5%; under the high growth scenario, losses were 12%. Including soil C, the proportional losses were lower, but the absolute amount lost was more than double (to 6 Tg C under the low and almost 15 Tg C under the high-growth scenario). This methodology could be further developed for applications in global change policy, such as payments for environmental services (PES) or reduction in emissions from deforestation and degradation (REDD).     

Biophysical suitability,economic pressure and land-cover change: a global probabilistic approach and insights for REDD+

There has been a concerted effort by the international scientific community to understand the multiple causes and patterns of land-cover change to support sustainable land management. Here, we examined biophysical suitability, and a novel integrated index of “Economic Pressure on Land” (EPL) to explain land cover in the year 2000, and estimated the likelihood of future land-cover change through 2050, including protected area effectiveness. Biophysical suitability and EPL explained almost half of the global pattern of land cover (R ² = 0.45), increasing to almost two-thirds in areas where a long-term equilibrium is likely to have been reached (e.g. R ² = 0.64 in Europe). We identify a high likelihood of future land-cover change in vast areas with relatively lower current and past deforestation (e.g. the Congo Basin). Further, we simulated emissions arising from a “business as usual” and two reducing emissions from deforestation and forest degradation (REDD) scenarios by incorporating data on biomass carbon. As our model incorporates all biome types, it highlights a crucial aspect of the ongoing REDD + debate: if restricted to forests, “cross-biome leakage” would severely reduce REDD + effectiveness for climate change mitigation. If forests were protected from deforestation yet without measures to tackle the drivers of land-cover change, REDD + would only reduce 30 % of total emissions from land-cover change. Fifty-five percent of emissions reductions from forests would be compensated by increased emissions in other biomes. These results suggest that, although REDD + remains a very promising mitigation tool, implementation of complementary measures to reduce land demand is necessary to prevent this leakage.     

Agroforest’s growing role in reducing carbon losses from Jambi (Sumatra), Indonesia

This paper examines the size and intensity of changes among five land categories during the two time intervals in a region of Indonesia that is pioneering negotiations concerning reducing emissions from deforestation and forest degradation (REDD). Maps at 1973, 1993, and 2005 indicate that land-cover change is accelerating, while carbon loss is decelerating in Jambi Province, Sumatra. Land dynamics have shifted from Forest loss during 1973–1993 to Agroforest loss during 1993–2005. Forest losses account for most reductions in aboveground carbon during the both time intervals, but Agroforest plays an increasingly important role in carbon reductions during the more recent interval. These results provide motivation for future REDD policies to count carbon changes associated with all influential land categories, such as Agroforests.     

Land-use portfolios and the management of private landholdings in south-central Indiana

In the study of land-use and land-cover change as an important contributor to regional environmental change, linking household level land-use decisions to particular land-cover patterns has been an enduring challenge. The frequent conflation of land use and land cover has been appropriate and fruitful in regions where extractive activities are common, but the decoupling of household characteristics from land-cover choices in exurban landscapes may require that the two be treated separately. This research employs the concept of a land-use portfolio, defined as a unique combination of land use types, and land cover derived from remote sensing to examine the relationship between land use and land cover at the parcel scale. Data on the type and spatial organization of land use were collected for individual parcels through sketch maps constructed by land owners and then described quantitatively using GIS and spatial metrics from landscape ecology. The results of this analysis suggest that in the naturally forested region of south-central Indiana, parcels are frequently managed as multiple types of land use thereby supporting the portfolio approach. Generally, land-management complexity is related to land-cover fragmentation, but the strength of this relationship varies across portfolio types. In addition, substantial amounts of forest land cover occur in non-forest land uses further supporting the need to treat the two separately to accurately link land use intentions and land cover outcomes.     

Quantifying disturbance effects on vegetation carbon pools in mountain forests based on historical data

Although the terrestrial carbon budget is of key importance for atmospheric CO₂ concentrations, little is known on the effects of management and natural disturbances on historical carbon stocks at the regional scale. We reconstruct the dynamics of vegetation carbon stocks and flows in forests across the past 100 years for a valley in the eastern Swiss Prealps using quantitative and qualitative information from forest management plans. The excellent quality of the historical information makes it possible to link dynamics in growing stocks with high-resolution time series for natural and anthropogenic disturbances. The results of the historical reconstruction are compared with modelled potential natural vegetation. Forest carbon stock at the beginning of the twentieth century was substantially reduced compared to natural conditions as a result of large scale clearcutting lasting until the late nineteenth century. Recovery of the forests from this unsustainable exploitation and systematic forest management were the main drivers of a strong carbon accumulation during almost the entire twentieth century. In the 1990s two major storm events and subsequent bark beetle infestations significantly reduced stocks back to the levels of the mid-twentieth century. The future potential for further carbon accumulation was found to be strongly limited, as the potential for further forest expansion in this valley is low and forest properties seem to approach equilibrium with the natural disturbance regime. We conclude that consistent long-term observations of carbon stocks and their changes provide rich information on the historical range of variability of forest ecosystems. Such historical information improves our ability to assess future changes in carbon stocks. Further, the information is vital for better parameterization and initialization of dynamic regional scale vegetation models and it provides important background for appropriate management decisions.     

The influence of socioeconomic, environmental, and demographic factors on municipality-scale land-cover change in Mexico

Land-cover change is the result of complex multi-scale interactions between socioeconomic, demographic, and environmental factors. Demographic change, in particular, is thought to be a major driver of forest change. Most studies have evaluated these interactions at the regional or the national level, but few studies have evaluated these dynamics across multiple spatial scales within a country. In this study, we evaluated the effect of demographic, environmental, and socioeconomic variables on land-cover change between 2001 and 2010 for all Mexican municipalities (n?=?2,443) as well as by biome (n?=?4). We used a land-cover classification based on 250-m MODIS data to examine the change in cover classes (i.e., woody, mixed woody, and agriculture/herbaceous vegetation). We evaluated the trends of land-cover change and identified the major factors correlated with woody vegetation change in Mexico. At the national scale, the variation in woody vegetation was best explained by environmental variables, particularly precipitation; municipalities where woody cover increased tended to be in areas with low average annual precipitation (i.e., desert and dry forest biomes). Demographic variables did not contribute much to the model at the national scale. Elevation, temperature, and population density explained the change in woody cover when municipalities were grouped by biome (i.e., moist forest, dry forest, coniferous forest, and deserts). Land-cover change at the biome level showed two main trends: (1) the tropical moist biome lost woody vegetation to agriculture and herbaceous vegetation, and (2) the desert biome increased in woody vegetation within more open-canopy shrublands.     

Regional carbon cycle responses to 25 years of variation in climate and disturbance in the US Pacific Northwest

Variation in climate, disturbance regime, and forest management strongly influence terrestrial carbon sources and sinks. Spatially distributed, process-based, carbon cycle simulation models provide a means to integrate information on these various influences to estimate carbon pools and flux over large domains. Here we apply the Biome-BGC model over the four-state Northwest US region for the interval from 1986 to 2010. Landsat data were used to characterize disturbances, and forest inventory data were used to parameterize the model. The overall disturbance rate on forest land across the region was 0.8 % year^?1, with 49 % as harvests, 28 % as fire, and 23 % as pest/pathogen. Net ecosystem production (NEP) for the 2006–2010 interval on forestland was predominantly positive (a carbon sink) throughout the region, with maximum values in the Coast Range, intermediate values in the Cascade Mountains, and relatively low values in the Inland Rocky Mountain ecoregions. Localized negative NEPs were mostly associated with recent disturbances. There was large interannual variation in regional NEP, with notably low values across the region in 2003, which was also the warmest year in the interval. The recent (2006–2010) net ecosystem carbon balance (NECB) was positive for the region (14.4 TgC year^?1). Despite a lower area-weighted mean NECB, public forestland contributed a larger proportion to the total NECB because of its larger area. Aggregated forest inventory data and inversion modeling are beginning to provide opportunities for evaluating model-simulated regional carbon stocks and fluxes.     

Land-cover changes in and around a National Park in a mountain landscape in the Pyrenees

The current state of Mediterranean mountain areas has been driven by two main factors: intense traditional human activity and the dynamics of the ecosystem itself. In this study, we examine land-cover changes in a National Park in the Pyrenees mountains (NE Iberian Peninsula), which was designated a protected area 55 years ago. First, we have analyzed spatio-temporal changes in land-cover pattern and forest dynamics from 1957 to 2005. During this period, land-cover dynamics consisted of two main processes: (i) expansion of the forest area and (ii) increasing cover of forests already present in 1957. To analyze the role of the conservation level of the park, we have also compared the results obtained within the park with those of unprotected, peripheral areas. In the two areas with different protection level, dense forests increased throughout the period because of the reduction in forestry activities. The peripheral area showed a higher rate of forest-cover change from 1957 to 2005 compared to the National Park. This higher increase in forest cover in the peripheral area could be related to a higher proportion in the National Park of screes and rocky areas and to the decline and transformation of forest activities in these peripheral, lower elevation areas.     

基于卫星夜间灯光数据的中国分省碳排放时空模拟

中国能源统计数据"横向不可比,纵向不可加"现象依然突出,尤其是分省能源消费统计千差万别,给分省碳排放评估带来了较大困难,如何利用卫星遥感数据科学合理地估算中国分省碳排放是当前亟须研究的问题。本文运用DMSP/OLS全球稳定夜间灯光数据,在通过相互校正、年内融合和年际间校正等系列处理得到中国分省稳定夜间灯光数据的基础上,首先分别构建中国分省稳定夜间灯光亮度DN值与人均碳排放和单位面积碳排放之间的时空地理加权回归模型,两个模型整体效果均较好,拟合优度分别高达96.74%和99.24%;其次运用稳定夜间灯光亮度DN值对分省人均碳排放和单位面积碳排放进行时空模拟;最后运用人口规模和土地面积对分省碳排放进行估算。估算结果显示:(1)整体来看,2000—2013年年均碳排放模拟值与实际值6.3349×109t较为接近,两个模型的相对误差均在0.5%以内。(2)分年度来看,所有年份的相对误差均在5%以内,2006年分省加总碳排放模拟值与实际碳排放6.2036×109t最为接近,绝对误差和相对误差均较小,两个模型模拟值的相对误差均为0.04%。(3)分省域来看,2000—2013年年均碳排放模拟值与实际碳排放均非常接近,除海南和宁夏外,其余28个省区市的相对误差均在1%以内。(4)分年度分省域来看,以2013年为例,40%省份的相对误差在2%以内,70%省份的相对误差在5%以内。从整体、分年度、分省域、分年度分省域的估算结果来看,基于稳定夜间灯光数据的中国分省碳排放时空模拟效果良好。因此,运用卫星夜间灯光数据可以较为准确地对中国分省碳排放进行估算和预测,为卫星遥感影像数据服务分省碳排放监测和评估提供一种补充性参考。     

Modeling spatio-temporal change patterns of forest cover: a case study from the Himalayan foothills (India)

The present study used temporal remote sensing data for 1990, 2001 and 2006 to assess spatio-temporal patterns of forest cover changes in Shiwalik range of the Himalaya, Dehradun forest division. Forests are innately associated to human well-being. However, with the increasing anthropogenic activities, deforestation has increased. Quantitative change analysis of the forest cover for the past two decades provides valuable insight into the forest conservation vis-à-vis anthropogenic activities in the region. Spatio-temporal datasets along with biotic and abiotic variables provide opportunities to model the forest cover change further. The present study investigates forest cover change and predicts status of forest cover in the Dehradun forest division. Land Change Modeller (LCM) was used to predict status of forest cover for 2010 and 2015 using current disturbance scenarios. Comparing actual LULC of 2006 with the predicted LULC of 2006 validated change prediction model and agreement was 61.03%. The forested areas are getting degraded due to anthropogenic activities, but deforestation/degradation does not contribute much in expanding urban area. Agricultural areas and fallow lands are the main contributors to increased urban area. The study demonstrates the potential of geospatial tools to understand spatio-temporal forest cover change and generate the future scenarios.     

Historical carbon fluxes in the expanding deforestation frontier of Southern Brazilian Amazonia (1985–2012)

In tropical areas, pioneer occupation fronts steer the rapid expansion of deforestation, contributing to carbon emissions. Up-to-date carbon emission estimates covering the long-term development of such frontiers depend on the availability of high spatial–temporal resolution data. In this paper, we provide a detailed assessment of carbon losses from deforestation and potential forest degradation from fragmentation for one expanding frontier in the Brazilian Amazon. We focused on one of the Amazonia’s hot-spots of forest loss, the BR-163 highway that connects the high productivity agricultural landscapes in Mato Grosso with the exporting harbors of the Amazon. We used multi-decadal (1984–2012) Landsat-based time series on forested and non-forested area in combination with a carbon book-keeping model. We show a 36% reduction in 1984s biomass carbon stocks, which led to the emission of 611.5 TgCO₂ between 1985 and 1998 (43.6 TgCO₂ year⁻¹) and 959.8 TgCO₂ over 1999–2012 (68.5 TgCO₂ year⁻¹). Overall, fragmentation-related carbon losses represented 1.88% of total emissions by 2012, with an increasing relevance since 2004. We compared the Brazilian Space Agency deforestation assessment (PRODES) with our data and found that small deforestation polygons not captured by PRODES had increasing importance on estimated deforestation carbon losses since 2000. The comparative analysis improved the understanding of data-source-related uncertainties on carbon estimates and indicated disagreement areas between datasets that could be subject of future research. Furthermore, spatially explicit, annual deforestation and emission estimates like the ones derived from this study are important for setting regional baselines for REDD+ or similar payment for ecosystem services frameworks.

     

Global estimates of soil carbon sequestration via livestock waste: a STELLA simulation

It has become increasingly well documented that human activities are enhancing the greenhouse effect and altering the global climate. Identifying strategies to mitigate atmospheric carbon dioxide emissions on the national level are therefore critical. Fossil fuel combustion is primarily responsible for the perturbation of the global carbon cycle, although the influence of humans extends far beyond the combustion of fossil fuels. Changes in land use arising from human activities contribute substantially to atmospheric carbon dioxide; however, land use changes can act as a carbon dioxide sink as well. A soil carbon model was built using STELLA to explore how soil organic carbon sequestration (SOC) varies over a range of values for key parameters and to estimate the amount of global soil carbon sequestration from livestock waste. To obtain soil carbon sequestration estimates, model simulations occurred for 11 different livestock types and with data for eight regions around the world. The model predicted that between 1980 and 1995, United States soils were responsible for the sequestration of 444–602 Tg C from livestock waste. Model simulations further predicted that during the same period, global soil carbon sequestration from livestock waste was 2,810–4,218 Tg C. Our estimates for global SOC sequestration are modest in proportion to other terrestrial carbon sinks (i.e. forest regrowth); however, livestock waste does represent a potential for long-term soil carbon gain. SOC generated from livestock waste is another example of how human activities and land use changes are altering soil processes around the world. Readers should send their comments on this paper to: BhaskarNath@aol.com within 3 months of publication of this issue.     

Role of remote sensing and community forestry to manage forests for the effective implementation of REDD+ mechanism: a case study on Cambodia

In this study, we have shown the importance of remote sensing applications and community forestry for forest management, discussed as a case study on Cambodian forest management. Curbing deforestation is necessary for the effective implementation of Reducing Emissions from Deforestation and forests Degradation (REDD+) mechanism and management of forest resources to support sustainable forest management plans. The updated information of the forest cover and forest biomass using advanced remote sensing techniques can be useful for selecting the suitable sites for planned thinning, reforestation, community forestry, and concession land, which eventually will help in controlling the deforestation in Cambodia. To overcome the limitations of remote sensing, an integrated approach of remote sensing and community forestry to monitor forests from local to national level has also been discussed.     

Land-cover changes and recent hydrological evolution in the Duero Basin (Spain)

Land-use and land-cover changes have attracted substantial scientific interest in recent years because of their marked influence on hydrological cycles. In developed countries of the Mediterranean basin, the generalized revegetation and forest growth in mountainous areas that occurred during the last five decades are negatively affecting the evolution of water resources in headwaters. In this study, changes in land cover in the Duero River Basin (northern Spain) during the last 50 years were analyzed and their role in hydrological evolution was estimated. For this purpose, step-wise linear regressions were developed to estimate the evolution of runoff as a function of climate (precipitation and temperatures). The results show a significant expansion of forest cover in the headwaters, although it has been more extensive in the mountains to the north of the basin than to the south. River discharges in the headwaters underwent a generalized decline during the study period (1961–2006), but precipitation over the same period did not show an appreciable trend. In the absence of noticeable trends in removal of water for human consumption, our results indicate that revegetation is contributing to the observed hydrological decline. Our hypothesis is confirmed because of the greater divergence in the evolution of precipitation and runoff in the northern headwaters (more forest growth) than in the south headwaters (less forest growth). Results suggest that further increases in forest area will enhance hydrological decline and highlight the importance of integrating land-cover information in water availability assessments in a region where water is a strategic resource.     

Agricultural adjustment,population dynamics and forests redistribution in a subtropical watershed of NW Argentina

Patterns of land-use and land-cover change are usually grouped into one of two categories defined by the dominant trend: (1) deforestation resulting from expanding agriculture and (2) forest expansion, usually related to the abandonment of marginal lands. At regional scale, however, both processes can occur simultaneously even in the absence of net change. Given the focus on net change, such redistribution of agricultural and natural and seminatural lands has been generally overlooked. The interaction between agriculture modernization, human demography and complex topographic gradients of northwestern Argentina has resulted in processes of both forest recovery and deforestation, thus providing the opportunity to analyze patterns and driving forces of land-cover redistribution. We analyzed 20 years (1986–2006) of land-cover change in a subtropical watershed in relation to topographic and demographic variables. Although net forest change represented <1 %, forests redistribution affected 7 % of forest lands. There was a consistent geographic segregation of deforestation and forest recovery, with forests expanding over steep highlands and agriculture expanding over lowland irrigated areas. Population trends were not associated to forest expansion in lowlands but they explained 32 % of forest recovery in highlands. Highland forest expansion and lowland deforestation, respectively, imply conservation opportunities for humid montane forests and the environmental services they provide (e.g., watershed conservation) and threats for the conservation of dry forests and its biodiversity. Our study exemplifies the importance of land-use redistribution (rather than net change) with relevant environmental consequences at regional scale.     

Land change in the southern Yucatán: case studies in land change science

The southern Yucatán Peninsular Region project was designed from the outset as an integrative, multidisciplinary program of study examining tropical deforestation in the largest track of seasonal tropical forest remaining in Mexico and in which smallholder agriculture and a major biosphere reserve are juxtaposed in regard to land uses and covers. Treating land as a coupled human–environment system, the project joins the remote sensing, environmental, social, and modeling sciences in a way that is now recognized as land change science. This paper introduces the project, the study region, and six papers that explore some of the coupled system dynamics in the region. These include the sub-regional variation in deforestation, the pan-regional adoption or anticipation of cattle ranching, the emergence of divergent household agricultural and overall livelihood strategies, the roles of cultural and household histories in agricultural livelihood choices, the temporal intensification of swidden cultivation and its implications for forest species, and carbon stocks across cultivation units, including a new econometric modeling application to forecast changes in these stocks.     

基于时空数据融合的江汉平原水稻种植信息提取

及时、准确监测水稻种植面积,对区域粮食政策制定、粮食安全以及农业发展具有重要意义。然而我国南方地区水稻生长期内降水充沛的气候特点使得遥感影像"云污染"现象严重,为解决水稻种植信息遥感提取存在可用数据不足的问题,以江汉平原为例,利用时空数据融合模型(Spatial and Temporal Data Fusion Approach,STDFA)将Landsat 8 OLI与时序MODIS数据融合,重构出具有高时-空分辨率特征数据,然后采用面向对象的SVM分类方法对研究区内水稻种植信息进行提取,结果如下:融合后的红与近红外波段反射率与真实反射率的相关系数分别为0.84和0.81,研究区水稻提取精度为94.46%,Kappa系数为0.91。说明时空融合模型能够较好地重构出高时空分辨率数据,从而实现多云雨地区农作物种植信息遥感提取。     

Characteristics and drivers of forest cover change in the post-socialist era in Croatia: evidence from a mixed-methods approach

Extensive forests in Croatia represent an important biological and economic resource in Europe. They are characterised by heterogeneity in forest management practices dating back to the socialist planned economy of the pre-1991 era. In this study we investigated the difference in rates of deforestation and reforestation in private- and state-owned forests during the post-socialist period and the causal drivers of change. The selected region of Northern Croatia is characterised by a high percentage of privately owned forests with minimal national monitoring and control. We used a mixed-methods approach which combines remote sensing, statistical modelling and a household-based questionnaire survey to assess the rates of forest cover change and factors influencing those changes. The results show that predominantly privately owned forests in Northern Croatia have recorded a net forest loss of 1.8 % during the 1991–2011 period, while Croatia overall is characterised by a 10 % forest cover increase in predominantly state-owned forests. Main factors influencing forest cover changes in private forests are slope, altitude, education structure, population age and population density. The results also show that the deforestation in private forests is weakening overall, mostly due to the continuation of the de-agrarisation and de-ruralisation processes which began during socialism.     

武汉市湖泊水域利用转变及其碳排放影响

考虑到湖泊水域面积与土地面积的相互变迁进程,以及土地利用变化已成为仅次于化石能源燃烧的第二大温室气体排放源的客观现实,首次尝试分析特定区域湖泊水域的变化以及间接产生的碳源碳汇影响。根据2005、2010和2015年遥感影像获取武汉市湖泊水域利用转变的数据,采用国际通用的土地碳排放测算模型,评估武汉市湖泊水域转变为4类用地产生的碳排放量及碳强度变化情况。研究显示:(1)2005～2015年武汉市湖泊水域面积整体减少超10 km~2,且2010年后呈现加速萎缩趋势, 2010～2015年的减少值约为2005～2010年减少值的2.04倍;(2)虽然部分耕地、林地和草地转化为湖泊导致湖泊水域的增加,但湖泊大面积转变为建设用地导致水域面积的显著较少;(3)从碳排放总量来看,变动面积区域内呈碳源区,后5年碳源值是前5年的1.45倍;前后5年湖泊水域变化的碳排放强度分别为0.03和0.04 t/km~2,碳排放强度明显增加。针对武汉市湖泊水域利用的现实状况,提出把"山水林田湖草"放到统一的管理机制中进行考虑、规范土地资源利用、限制建设用地过度扩张等措施。     

Forest dynamics in Mississippi, USA: a hybrid statistical and geospatial analysis

Understanding forest changes and its trajectory is important to develop policy options and future scenarios for climate analysis. This research is conducted to gain insights on secondary forests change using Mississippi, USA, as a case study. We investigate the spatial patterns and temporal dynamics of secondary forests at high resolution and examine the forces driving their changes. An extensive literature review is conducted to refine the conceptual framework of forest changes and identify the underlying key factors. Forest changes are quantified at high spatial (30-m) and temporal (biennial) resolutions, using time series remotely sensed data between 1984 and 2007. A number of geospatial and socioeconomic data were compiled to analyze the spatial variations of forest disturbances and their linkages to various socioeconomic, political, and biogeophysical factors. The results show that the secondary forests are highly dynamic and variable. Disturbances and regeneration occur continuously everywhere in a systematic and coordinated fashion. This pattern prevents an extensive disturbance and increases total forest cover. Market conditions (i.e., timber price) are the key predictor of the level and overall trend of forest disturbances. However, spatial patterns of forest dynamics cannot be explained by location-specific biophysical, socioeconomic, and policy factors identified in the literature. They can best be described by the ecological characteristics of the forests (i.e., the forest type and age distribution), which have a clear economic linkage. The research shows that regenerated forests frequently experience loss and gain of their extent, and their ecological characteristics change drastically on a short-term basis. These results point out challenges and opportunities in forest management and policy with regard to reforestation.