Hyperspectral data as a biodiversity screening tool can differentiate among diverse Neotropical fishes.

Hyperspectral data encode information from electromagnetic radiation (i.e., color) of any object in the form of a spectral signature; these data can then be used to distinguish among materials or even map whole landscapes. Although hyperspectral data have been mostly used to study landscape ecology, floral diversity and many other applications in the natural sciences, we propose that spectral signatures can be used for rapid assessment of faunal biodiversity, akin to DNA barcoding and metabarcoding. We demonstrate that spectral signatures of individual, live fish specimens can accurately capture species and clade-level differences in fish coloration, specifically among piranhas and pacus (Family Serrasalmidae), fishes with a long history of taxonomic confusion. We analyzed 47 serrasalmid species and could distinguish spectra among different species and clades, with the method sensitive enough to document changes in fish coloration over ontogeny. Herbivorous pacu spectra were more like one another than they were to piranhas; however, our method also documented interspecific variation in pacus that corresponds to cryptic lineages. While spectra do not serve as an alternative to the collection of curated specimens, hyperspectral data of fishes in the field should help clarify which specimens might be unique or undescribed, complementing existing molecular and morphological techniques.

Nitrous oxide, methane and carbon dioxide patterns and dynamics from an experimental pig mass grave.

The objective of the three-year study was to examine spatial and temporal patterns of fluxes and soil pore air concentrations of methane (CH4), carbon dioxide (CO2) and nitrous oxide (N2O) from an experimental mass grave located in a temperate environment. The mass grave (5×10m) contained twenty pig carcasses at a depth of approximately 1m was compared to a plot of the same dimensions containing only disturbed soil, as well as an undisturbed plot. Soil pore air CH4 concentrations were sub-ambient (<1.8ppm) except at 75 and 100cm depths at the mass grave in years 1 and 2 but decreased in year 3. The consumption of CH4 within the aerobic soil resulted in small negative fluxes at the soil surface. Soil pore air CO2 concentration showed an increase with depth in all three plots, with the largest increase (>100,000ppm at 1m) in the mass grave, though there was a marked decrease from years 1 to 3. Surface fluxes of CO2 showed strong seasonal variations, peaking in summer. Soil pore air N2O concentration showed major increases in the mass grave, compared to the other two plots with the pattern maintained over the three years, resulting in larger surface fluxes of N2O. To establish the role of the carcasses in N2O dynamics, we incubated a soil sample containing carcass material which resulted in fast rates of N2O production and consumption. The maintenance of elevated pore air concentration and surface flux of N2O throughout the 3 years suggests that this is a long-term pattern and likely the best of the three gases to use to detect graves. Thus, we suggest that measurement of soil pore air concentrations, especially of N2O, could be a simple and effective approach to help determine the location of clandestine graves.

Nitrous oxide, methane and carbon dioxide dynamics from experimental pig graves.

Twelve pig carcasses were buried in single, shallow and deep (30 and 90 cm, respectively) graves at an experimental site near Ottawa, Ontario, Canada, with three shallow and three deep wrapped in black plastic garbage bags. An additional six carcasses were left at the surface to decompose, three of which were bagged. Six reference pits without remains were also dug. The objective of this three-year study was to examine the biogeochemistry and utility of nitrous oxide (N2O), methane (CH4) and carbon dioxide (CO2) in grave detection and whether grave depth or cadaver condition (bagged versus bare) affected soil pore air concentrations and emission of the three gases. Graves showed significantly higher (α=0.05) concentrations and surface fluxes of N2O and CO2 than reference pits, but there was no difference in CH4 between graves and reference pits. While CH4 decreased with depth in the soil profiles, N2O and CO2 showed a large increase compared to reference pits. Shallow graves showed significantly higher emissions and pore air concentrations of N2O and CO2 than deep graves, as did bare versus bagged carcasses.

Detection of single graves by airborne hyperspectral imaging.

Airborne hyperspectral imaging (HSI) was assessed as a potential tool to locate single grave sites. While airborne HSI has shown to be useful to locate mass graves, it is expected the location of single graves would be an order of magnitude more difficult due to the smaller size and reduced mass of the targets. Two clearings were evaluated (through a blind test) as potential sites for containing at least one set of buried remains. At no time prior to submitting the locations of the potential burial sites from the HSI were the actual locations of the sites released or shared with anyone from the analysis team. The two HSI sensors onboard the aircraft span the range of 408-2524nm. A range of indicators that exploit the narrow spectral and spatial resolutions of the two complimentary HSI sensors onboard the aircraft were calculated. Based on the co-occurrence of anomalous pixels within the expected range of the indicators three potential areas conforming to our underlying assumptions of the expected spectral responses (and spatial area) were determined. After submission of the predicted burial locations it was revealed that two of the targets were located within GPS error (10m) of the true burial locations. Furthermore, due to the history of the TPOF site for burial work, investigation of the third target is being considered in the near future. The results clearly demonstrate promise for hyperspectral imaging to aid in the detection of buried remains, however further work is required before these results can justifiably be used in routine scenarios.

Baseline assessment for environmental services payments from satellite imagery: a case study from Costa Rica and Mexico.

In this study we evaluate the accuracy of four global and regional forest cover assessments (MODIS, IGBP, GLC2000, PROARCA) as tools for baseline estimation. We conduct this research at the national scale for Costa Rica and for two tropical dry forest study sites in Costa Rica (Santa Rosa) and Mexico (Chamela-Cuixmala). We found that at the national level, the total forest cover accuracy of the four land cover maps was inflated due to an overestimation of forest in areas with an evergreen canopy. However, the four maps greatly underestimated the extent of the deciduous forest (dry forest); an ecosystem that faces high deforestation pressure and poses complications to the mapping of its extent from remotely sensed data. For the tropical dry forest sites, all maps have low forest cover accuracies (mean for Santa Rosa: 27%; mean for Chamela-Cuixmala: 56%). This has implications for policy implementation.

Calibration and assessment of seasonal changes in leaf area index of a tropical dry forest in different stages of succession.

A simple measure of the amount of foliage present in a forest is leaf area index (LAI; the amount of foliage per unit ground surface area), which can be determined by optical estimation (gap fraction method) with an instrument such as the Li-Cor LAI-2000 Plant Canopy Analyzer. However, optical instruments such as the LAI-2000 cannot directly differentiate between foliage and woody components of the canopy. Studies investigating LAI and its calibration (extracting foliar LAI from optical estimates) in tropical forests are rare. We calibrated optical estimates of LAI from the LAI-2000 with leaf litter data for a tropical dry forest. We also developed a robust method for determining LAI from leaf litter data in a tropical dry forest environment. We found that, depending on the successional stage of the canopy and the season, the LAI-2000 may underestimate LAI by 17% to over 40%. In the dry season, the instrument overestimated LAI by the contribution of the woody area index. Examination of the seasonal variation in LAI for three successional stages in a tropical dry forest indicated differences in timing of leaf fall according to successional stage and functional group (i.e., lianas and trees). We conclude that when calculating LAI from optical estimates, it is necessary to account for the differences between values obtained from optical and semi-direct techniques. In addition, to calculate LAI from litter collected in traps, specific leaf area must be calculated for each species rather than from a mean value for multiple species.