Remote sensing of native and invasive species in Hawaiian forests

Detection and mapping of invasive species is an important component of conservation and management efforts in Hawai'i, but the spectral separability of native, introduced, and invasive species has not been established. We used high spatial resolution airborne imaging spectroscopy to analyze the canopy hyperspectral reflectance properties of 37 distinct species or phenotypes, 7 common native and 24 introduced tree species, the latter group containing 12 highly invasive species. Airborne Visible and Infrared Imaging Spectrometer (AVIRIS) reflectance and derivative-reflectance signatures of Hawaiian native trees were generically unique from those of introduced trees. Nitrogen-fixing trees were also spectrally unique from other groups of non-fixing trees. There were subtle but significant differences in the spectral properties of highly invasive tree species in comparison to introduced species that do not proliferate across Hawaiian ecosystems. The observed differences in canopy spectral signatures were linked to relative differences in measured leaf pigment (chlorophyll, carotenoids), nutrient (N, P), and structural (specific leaf area; SLA) properties, as well as to canopy leaf area index. These leaf and canopy properties contributed variably to the spectral separability of the trees, with wavelength-specific reflectance and absorption features that overlapped, but which were unique from one another. A combination of canopy reflectance from 1125–2500 nm associated with leaf and canopy water content, along with pigment-related absorption features (reflectance derivatives) in the 400–700 nm range, was best for delineating native, introduced, and invasive species. There was no single spectral region that always defined the separability of the species groups, and thus the full-range (400–2500 nm) spectrum was highly advantageous in differentiating these groups. These results provide a basis for more detailed studies of invasive species in Hawai'i, along with more explicit treatment of the biochemical properties of the canopies and their prediction using imaging spectroscopy.     

Identification of gaps in mangrove forests with airborne LIDAR

Mangrove forests change frequently due to disturbances from tropical storms, frost, lightning, and insects. It has been suggested that the death and regeneration of trees in small gaps due to lightning may play a critical role in mangrove forest turnover; however, the large-scale quantification of spatial pattern and areas of gaps is lacking for investigating this issue. Airborne light detection and ranging (LIDAR) technology provides an effective way for identifying gaps by remotely obtaining direct measurements of ground and canopy elevations. A method based on an alternative sequential filter and black top-hat mathematical morphological transformation was developed to extract gap features. Comparison of identified gap polygons with raw LIDAR measurements and field surveys shows that the proposed method successfully extracted gap features in mangrove forests in Everglades National Park. There are 400–500 lightning gaps per square kilometer in mangrove forests at the study sites. The distribution of gap sizes follows an exponential form and the area of gaps with sizes larger than 100 m² account for 55–61% of the total area of gaps. The area of gaps in the mangrove forest in Everglades National Park is about 4–5% of the total forest area and the average gap formation rate is about 0.3% of the total forest area per year, indicating that lightning gaps play an important role in mangrove forest dynamics.     

Classification of Australian forest communities using aerial photography, CASI and HyMap data

Within Australia, the discrimination and mapping of forest communities has traditionally been undertaken at the stand scale using stereo aerial photography. Focusing on mixed species forests in central south-east Queensland, this paper outlines an approach for the generation of tree species maps at the tree crown/cluster level using 1 m spatial resolution Compact Airborne Spectrographic Imager (CASI; 445.8 nm–837.7 nm wavelength) and the use of these to generate stand-level assessments of community composition. Following automated delineation of tree crowns/crown clusters, spectral reflectance from pixels representing maxima or mean-lit averages of channel reflectance or band ratios were extracted for a range of species including Acacia, Angophora, Callitris and Eucalyptus. Based on stepwise discriminant analysis, classification accuracies of dominant species were greatest (87% and 76% for training and testing datasets; n = 398) when the mean-lit spectra associated with a ratio of the reflectance (ρ) at 742 nm (ρ₇₄₂) and 714 nm (ρ₇₁₄) were used. The integration of 2.6 m HyMap (446.1 nm–2477.8 nm) spectra increased the accuracy of classification for some species, largely because of the inclusion of shortwave infrared wavebands. Similar increases in accuracy were achieved when classifications of field spectra resampled to CASI and HyMap wavebands were compared. The discriminant functions were applied subsequently to classify crowns within each image and produce maps of tree species distributions which were equivalent or better than those generated through aerial photograph interpretation. The research provides a new approach to tree species mapping, although some a priori knowledge of the occurrence of broad species groups is required. The tree maps have application to biodiversity assessment in Australian forests.     

Prospects for quantifying structure,floristic composition and species richness of tropical forests

Airborne spectral and light detection and ranging (lidar) sensors have been used to quantify biophysical characteristics of tropical forests. Lidar sensors have provided high-resolution data on forest height, canopy topography, volume, and gap size; and provided estimates on number of strata in a forest, successional status of forests, and above-ground biomass. Spectral sensors have provided data on vegetation types, foliar biochemistry content of forest canopies, tree and canopy phenology, and spectral signatures for selected tree species. A number of advances are theoretically possible with individual and combined spectral and lidar sensors for the study of forest structure, floristic composition and species richness. Delineating individual canopies of over-storey trees with small footprint lidar and discrimination of tree architectural types with waveform distributions is possible and would provide scientists with a new method to study tropical forest structure. Combined spectral and lidar data can be used to identify selected tree species and identify the successional status of tropical forest fragments in order to rank forest patches by levels of species richness. It should be possible in the near future to quantify selected patterns of tropical forests at a higher resolution than can currently be undertaken in the field or from space.     

Modeling forest songbird species richness using LiDAR-derived measures of forest structure

Conservation of biodiversity requires information at many spatial scales in order to detect and preserve habitat for many species, often simultaneously. Vegetation structure information is particularly important for avian habitat models and has largely been unavailable for large areas at the desired resolution. Airborne LiDAR, with its combination of relatively broad coverage and fine resolution provides existing new opportunities to map vegetation structure and hence avian habitat. Our goal was to model the richness of forest songbirds using forest structure information obtained from LiDAR data. In deciduous forests of southern Wisconsin, USA, we used discrete-return airborne LiDAR to derive forest structure metrics related to the height and density of vegetation returns, as well as composite variables that captured major forest structural elements. We conducted point counts to determine total forest songbird richness and the richness of foraging, nesting, and forest edge-related habitat guilds. A suite of 35 LiDAR variables were used to model bird species richness using best-subsets regression and we used hierarchical partitioning analysis to quantify the explanatory power of each variable in the multivariate models. Songbird species richness was correlated most strongly with LiDAR variables related to canopy and midstory height and midstory density (R² = 0.204, p < 0.001). Richness of species that nest in the midstory was best explained by canopy height variables (R² = 0.197, p < 0.001). Species that forage on the ground responded to mean canopy height and the height of the lower canopy (R² = 0.149, p < 0.005) while aerial foragers had higher richness where the canopy was tall and dense and the midstory more sparse (R² = 0.216, p < 0.001). Richness of edge-preferring species was greater where there were fewer vegetation returns but higher density in the understory (R² = 0.153, p < 0.005). Forest interior specialists responded positively to a tall canopy, developed midstory, and a higher proportion of vegetation returns (R² = 0.195, p < 0.001). LiDAR forest structure metrics explained between 15 and 20% of the variability in richness within deciduous forest songbird communities. This variability was associated with vertical structure alone and shows how LiDAR can provide a source of complementary predictive data that can be incorporated in models of wildlife habitat associations across broad geographical extents.     

Estimating field-plot data of forest stands using airborne laser scanning and SPOT HRG data

Airborne laser scanner systems provide detailed forest information that can be used for important improvements in forest management decisions. Planning systems under development use plot-survey data to represent forest stands in large forest holdings which enables new flexible methods to model the forest and optimize selection of silvicultural treatments. In Sweden today, only averages of forest stand variables are used, and the survey methods used do not provide plot-survey data for all stands in large forest holdings. This is a task possibly solved using airborne laser scanner data. Various measures can be derived from laser data, each describing different forest variables, such as tree height distribution, vegetation density and vertical tree crown structure. Here, imputation of field plot (10 m radius) data using measures derived from airborne laser scanner data (TopEye) and optical image data (SPOT 5 HRG satellite sensor) were evaluated as a method to provide data for new long-term management planning systems. In addition to commonly applied measures, the semivariogram of laser measurements was evaluated as a new measure to extract spatial characteristics of the forest. The study used data from 870, 10 m radius field plots (0 to 812 m³ ha^− 1) surveyed for a 1200 ha large forest estate in the south of Sweden. At the best, combining measures derived from laser scanner data and SPOT 5 data, stand mean volume was estimated with a root mean square error (RMSE) of 20% of the sample mean and stem density with 22% RMSE. Bias of stem density estimates was 5%, and stand stem volume 4%. Although these accuracies are sufficient for operational application, estimates of tree species proportions and within-stand variation were clearly not.     

Using AVIRIS data and multiple-masking techniques to map urban forest tree species

Tree type and species information are critical parameters for urban forest management, benefit cost analysis and urban planning. However, traditionally, these parameters have been derived based on limited field samples in urban forest management practice. In this study we used high-resolution Airborne Visible Infrared Imaging Spectrometer (AVIRIS) data and multiple-spectral masking techniques to identify and map urban forest trees. Trees were identified based on their spectral character difference in AVIRIS data. The use of multiple-masking techniques shift the focus to the target land cover types only, thus reducing confounding noise during spectral analysis. The results were checked against ground reference data and by comparison to tree information in an existing geographical information system (GIS) database. At the tree type level, mapping was accomplished with 94% accuracy. At the tree species level, the average accuracy was 70% but this varied with both tree type and species. Of the four evergreen tree species, the average accuracy was 69%. For the 12 deciduous tree species, the average accuracy was 70%. The relatively low accuracy for several deciduous species was due to small tree size and overlapping among tree crowns at the 3.5 m spatial resolution of AVIRIS data. This urban forest tree species mapping method has the potential to increase data update intervals and accuracy while reducing costs compared to field sampling or other traditional methods.     

Mapping vegetation in Yellowstone National Park using spectral feature analysis of AVIRIS data

Knowledge of the distribution of vegetation on the landscape can be used to investigate ecosystem functioning. The sizes and movements of animal populations can be linked to resources provided by different plant species. This paper demonstrates the application of imaging spectroscopy to the study of vegetation in Yellowstone National Park (Yellowstone) using spectral feature analysis of data from the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). AVIRIS data, acquired on August 7, 1996, were calibrated to surface reflectance using a radiative transfer model and field reflectance measurements of a ground calibration site. A spectral library of canopy reflectance signatures was created by averaging pixels of the calibrated AVIRIS data over areas of known forest and nonforest vegetation cover types in Yellowstone. Using continuum removal and least squares fitting algorithms in the US Geological Survey's Tetracorder expert system, the distributions of these vegetation types were determined by comparing the absorption features of vegetation in the spectral library with the spectra from the AVIRIS data. The 0.68 μm chlorophyll absorption feature and leaf water absorption features, centered near 0.98 and 1.20 μm, were analyzed. Nonforest cover types of sagebrush, grasslands, willows, sedges, and other wetland vegetation were mapped in the Lamar Valley of Yellowstone. Conifer cover types of lodgepole pine, whitebark pine, Douglas fir, and mixed Engelmann spruce/subalpine fir forests were spectrally discriminated and their distributions mapped in the AVIRIS images. In the Mount Washburn area of Yellowstone, a comparison of the AVIRIS map of forest cover types to a map derived from air photos resulted in an overall agreement of 74.1% (kappa statistic=0.62).     

基于机载激光雷达数据的森林结构参数反演

机载激光雷达(Light Detection And Ranging,LiDAR)技术对植被空间结构和地形的探测能力较强,在植被参数定量测量和反演方面具有显著优势。首先利用野外调查并结合高分辨率Geoeye-1影像数据,对黑河上游天涝池流域植被类型进行分类,提取研究区森林分布,然后结合0.5m×0.5m机载激光雷达(LiDAR)数据对森林结构参数(树高、冠幅、胸径和叶面积指数)进行反演,最后利用实际观测数据对反演结果进行验证。结果表明:机载激光雷达数据能够精确地反演森林结构参数,树高、冠幅、胸径和叶面积指数的实测值与估测值决定系数分别为0.98、0.84、0.57和0.73。本研究获得流域森林覆盖区域高精度树冠高度和叶面积指数空间分布图,同时分析了冠层高度和叶面积指数随高度的变化。本研究的结果为该流域分布式生态水文模型提供了重要的输入参数。     

Integrating waveform lidar with hyperspectral imagery for inventory of a northern temperate forest

It has been suggested that attempts to use remote sensing to map the spatial and structural patterns of individual tree species abundances in heterogeneous forests, such as those found in northeastern North America, may benefit from the integration of hyperspectral or multi-spectral information with other active sensor data such as lidar. Towards this end, we describe the integrated and individual capabilities of waveform lidar and hyperspectral data to estimate three common forest measurements - basal area (BA), above-ground biomass (AGBM) and quadratic mean stem diameter (QMSD) - in a northern temperate mixed conifer and deciduous forest. The use of this data to discriminate distribution and abundance patterns of five common and often, dominant tree species was also explored. Waveform lidar imagery was acquired in July 2003 over the 1000 ha. Bartlett Experimental Forest (BEF) in central New Hampshire (USA) using NASA's airborne Laser Vegetation Imaging Sensor (LVIS). High spectral resolution imagery was likewise acquired in August 2003 using NASA's Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). Field data (2001-2003) from over 400 US Forest Service Northern Research Station (USFS NRS) plots were used to determine actual site conditions.Results suggest that the integrated data sets of hyperspectral and waveform lidar provide improved outcomes over use of either data set alone in evaluating common forest metrics. Across all forest conditions, 8-9% more of the variation in AGBM, BA, and QMSD was explained by use of the integrated sensor data in comparison to either AVIRIS or LVIS metrics applied singly, with estimated error 5-8% lower for these variables. Notably, in an analysis using integrated data limited to unmanaged forest tracts, AGBM coefficients of determination improved by 25% or more, while corresponding error levels decreased by over 25%. When data were restricted based on the presence of individual tree species within plots, AVIRIS data alone best predicted species-specific patterns of abundance as determined by species fraction of biomass. Nonetheless, use of LVIS and AVIRIS data - in tandem - produced complementary maps of estimated abundance and structure for individual tree species, providing a promising adjunct to traditional forest inventory and conservation biology planning efforts.     

Quantifying the spatial properties of forest canopy gaps using LiDAR imagery and GIS

The spatial properties of gaps have an important influence upon the regeneration dynamics and species composition of forests. However, such properties can be difficult to quantify over large spatial areas using field measurements. This research considers how we conceptualize and define forest canopy gaps from a remote sensing point of view and highlights the inadequacies of passive optical remotely sensed data for delineating gaps. The study employs the analytical functions of a geographical information system to extract gap spatial characteristics from imagery acquired by an active remote sensing device, an airborne light detection and ranging instrument (LiDAR). These techniques were applied to an area of semi-natural broadleaved deciduous forest, in order to map gap size, shape complexity, vegetation height diversity and gap connectivity. A vegetation cover map derived from imagery from an airborne multispectral scanner was used in combination with the LiDAR data to characterize the dominant vegetation types within gaps. Although the quantification of these gap characteristics alone is insufficient to provide conclusive evidence on specific processes, the paper demonstrates how such information can be indicative of the general status of a forest and can provide new perspectives and possibilities or further ecological research and forest monitoring activities.     

机载脉冲激光雷达剖面测量技术的进展及应用

机载脉冲激光雷达(LiDAR)剖面测量技术是一种先进的主动遥感测量技术,可以快速、大面积地直接获取地表地物、森林和水下地貌等的三维信息,具有机动性强、高效和实时等优点。该技术可用于海岸线海水深度测量,海岸生态状况监测,森林资源调查及地震等突发事件响应,具有巨大的应用潜力和广阔的发展前景。首先全面介绍了国外机载脉冲激光雷达剖面测量技术的发展历史,评述了各个发展阶段,并介绍了国内该技术的发展状况。然后分析了机载脉冲激光雷达剖面测量技术的主要应用领域。最后对机载脉冲激光雷达剖面测量技术的未来发展趋势作了展望。     

Delineation of individual deciduous trees in plantations with low-density LiDAR data

Delineation of individual deciduous trees with Light Detection and Ranging (LiDAR) data has long been sought for accurate forest inventory in temperate forests. Previous attempts mainly focused on high-density LiDAR data to obtain reliable delineation results, which may have limited applications due to the high cost and low availability of such data. Here, the feasibility of individual deciduous tree delineation with low-density LiDAR data was examined using a point-density-based algorithm. First a high-resolution point density model (PDM) was developed from low-density LiDAR point cloud to locate individual trees through the horizontal spatial distribution of LiDAR points. Then, individual tree crowns and associated attributes were delineated with a 2D marker-controlled watershed segmentation. Additionally, the PDM-based approach was compared with a conventional canopy height model (CHM) based delineation. The results demonstrated that the PDM-based approach produced an 89% detection accuracy to identify deciduous trees in our study area. The tree attributes derived from the PDM-based algorithm explained 81% and 83% of tree height and crown width variations of forest stands, respectively. The conventional CHM-based tree attributes, on the other hand, could explain only 71% and 66% of tree height and crown width, respectively. Our results suggest that the application of the PDM-based individual tree identification in deciduous forests with low-density LiDAR data is feasible and has relatively high accuracy to predict tree height and crown width, which are highly desired in large-scale forest inventory and analysis.     

Mapping canopy gaps in an indigenous subtropical coastal forest using high-resolution WorldView-2 data

Invasive species usually colonize canopy gaps in tropical and subtropical forests, which results in a loss of native species. Therefore, an understanding of the location and distribution of canopy gaps will assist in predicting the occurrence of invasive species in such canopy gaps. We tested the utility of WorldView-2 (WV-2) with eight spectral bands at 2 m spatial resolution to delineate forest canopy gaps in a subtropical Dukuduku coastal forest in South Africa. We compared the four conventional visible-near-infrared bands with the eight-band WV-2 image. The eight-band WV-2 image yielded a higher overall accuracy of 86.90% (kappa coefficient = 0.82) than the resampled conventional four-band image that yielded an overall accuracy of 74.64% (kappa coefficient = 0.63) in pixel-based classification. We further compared the vegetation indices that were derived from four conventional bands with those derived from WV-2 bands. The enhanced vegetation index yielded the highest overall accuracy in the category of conventional indices (85.59% at kappa coefficient = 0.79), while the modified plant senescence reflectance index involving the red-edge band showed the highest overall accuracy (93.69%) in the category of indices derived from eight-band WV-2 imagery in object-based classification. Overall, the study shows that the unique high-resolution WV-2 data can improve the delineation of canopy gaps as compared to the conventional multispectral bands.     

基于机载激光雷达数据识别单株木的新方法

在林业应用中,机载激光雷达技术较被动式遥感技术有着独特优势。以从机载激光雷达数据中分离单株木,提取单株木树高和树冠大小信息为目的,融合有标记约束的分水岭分割和流域跟踪分割两种图像分割方法,提出一种新的单株木识别思路。以美国某地区实地采集激光雷达数据为例验证本文提出方法,实验结果表明该方法通过增加边缘检测范围的约束条件,能够有效避免过分割现象,并通过使用约束条件,减少在其检测范围内的目标数量,从而避免不必要的检测干扰,较传统方法能快速准确地识别单株木。     

Strengths and limitations of assessing forest density and spatial configuration with aerial LiDAR

Changes in the structural state of forests of the semi-arid U.S.A., such as an increase in tree density, are widely believed to be leading to an ecological crisis, but accurate methods of quantifying forest density and configuration are lacking at landscape scales. An individual tree canopy (ITC) method based on aerial LiDAR has been developed to assess forest structure by estimating the density and spatial configuration of trees in four different height classes. The method has been tested against field measured forest inventory data from two geographically distinct forests with independent LiDAR acquisitions. The results show two distinct patterns: accurate, unbiased density estimates for trees taller than 20 m, and underestimation of density in trees less than 20 m tall. The underestimation of smaller trees is suggested to be a limitation of LiDAR remote sensing. Ecological applications of the method are demonstrated through landscape metrics analysis of density and configuration rasters.     

A systematic method for 3D mapping of mangrove forests based on Shuttle Radar Topography Mission elevation data, ICEsat/GLAS waveforms and field data: Application to Ciénaga Grande de Santa Marta, Colombia

Mangrove forests are found within the intertropical zone and are one of the most biodiverse and productive wetlands on Earth. We focus on the Ciénaga Grande de Santa Marta (CGSM) in Colombia, the largest coastal lagoon–delta ecosystem in the Caribbean area with an extension of 1280 km², where one of the largest mangrove rehabilitation projects in Latin America is currently underway. Extensive man-made hydrological modifications in the region caused hypersaline soil (> 90 g kg^− 1) conditions since the 1960s triggering a large dieback of mangrove wetlands (~ 247 km²). In this paper, we describe a new systematic methodology to measure mangrove height and aboveground biomass by remote sensing. The method is based on SRTM (Shuttle Radar Topography Mission) elevation data, ICEsat/GLAS waveforms (Ice, Cloud, and Land Elevation Satellite/Geoscience Laser Altimeter System) and field data. Since the locations of the ICEsat and field datasets do not coincide, they are used independently to calibrate SRTM elevation and produce a map of mangrove canopy height. We compared height estimation methods based on waveform centroids and the canopy height profile (CHP). Linear relationships between ICEsat height estimates and SRTM elevation were derived. We found the centroid of the canopy waveform contribution (CWC) to be the best height estimator. The field data was used to estimate a SRTM canopy height bias (− 1.3 m) and estimation error (rms = 1.9 m). The relationship was applied to the SRTM elevation data to produce a mangrove canopy height map. Finally, we used field data and published allometric equations to derive an empirical relationship between canopy height and biomass. This relationship was used to scale the mangrove height map and estimate aboveground biomass distribution for the entire CGSM. The mean mangrove canopy height in CGSM is 7.7 m and most of the biomass is concentrated in forests around 9 m in height. Our biomass maps will enable estimation of regeneration rates of mangrove forests under hydrological rehabilitation at large spatial scales over the next decades. They will also be used to assess how highly disturbed mangrove forests respond to increasing sea level rise under current global climate change scenarios.     

An evaluation of imaging spectrometry for estimating forest canopy chemistry

Abstract

High spectral resolution Airborne Imaging Spectrometer (AIS) data were acquired over 20 well-studied Wisconsin forest sites to evaluate the potential of remote sensing for estimating forest canopy chemistry. Intensive nutrient cycling research in these forests demonstrates that canopy lignin content is strongly related to measured annual nitrogen mineralization at the undisturbed sites and may serve as an accurate index for nitrogen cycling rates. Ground measurements were made of foliar biomass and canopy nitrogen and lignin content, the latter within two weeks of the AIS overflight. The spectral data were transformed using derivative techniques modified from laboratory spectroscopy. Stepwise regression assisted in determining combinations of wavelengths most highly correlated with canopy chemistry and biomass. Strong correlations between AIS data and total canopy lignin content in deciduous forests and canopy lignin concentration (total lignin/biomass) in both deciduous and coniferous stands indicate that imaging spectrometry can be used to estimate canopy lignin content and, from that, the spatial distribution of annual nitrogen mineralization rates.     

基于激光雷达的大兴安岭典型森林生物量制图技术研究

森林生物量作为森林生态系统基本的数量表征,表明了森林的经营水平和开发利用价值,并能反映其与环境在物质循环和能量流动方面的复杂关系。同时,森林生物量也是林业问题和生态问题研究的基础。以内蒙古大兴安岭国家野外生态站为研究区域,通过对机载激光雷达(LiDAR)点云数据的预处理,利用计算机编程提取LiDAR点云数据的结构参数,以植被分位数高度变量与密度变量为自变量,结合地面调查数据,建立生物量与LiDAR结构参数的回归模型(决定系数为0.69,均方根误差为0.34)。运用IDL编程对LiDAR点云块数据进行运算并生成分辨率为20m×20m的栅格图像,拼接后得到整个区域的地上生物量分布图,对生成的地上生物量分布图进行验证的R2为0.78,RMSE为23.09t/hm2,平均估测精度达83%。     

Integrating Landsat TM and SRTM-DEM derived variables with decision trees for habitat classification and change detection in complex neotropical environments

Remotely sensed images and processing techniques are a primary tool for mapping changes in tropical forest types important to biodiversity and environmental assessment. Detailed land cover data are lacking for most wet tropical areas that present special challenges for data collection. For this study, we utilize decision tree (DT) classifiers to map 32 land cover types of varying ecological and economic importance over an 8000 km² study area and biological corridor in Costa Rica. We assess multivariate QUEST DTs with unbiased classification rules and linear discriminant node models for integrated vegetation mapping and change detection. Predictor variables essential to accurate land cover classification were selected using importance indices statistically derived with classification trees. A set of 35 variables from SRTM-DEM terrain variables, WorldClim grids, and Landsat TM bands were assessed.

Of the techniques examined, QUEST trees were most accurate by integrating a set of 12 spectral and geospatial predictor variables for image subsets with an overall cross-validation accuracy of 93% ± 3.3%. Accuracy with spectral variables alone was low (69% ± 3.3%). A random selection of training and test set pixels for the entire landscape yielded lower classification accuracy (81%) demonstrating a positive effect of image subsets on accuracy. A post-classification change comparison between 1986 and 2001 reveals that two lowland forest types of differing tree species composition are vulnerable to agricultural conversion. Tree plantations and successional vegetation added forest cover over the 15-year time period, but sometimes replaced native forest types, reducing floristic diversity. Decision tree classifiers, capable of combining data from multiple sources, are highly adaptable for mapping and monitoring land cover changes important to biodiversity and other ecosystem services in complex wet tropical environments.