Consequences of underexplored variation in biodiversity indices used for land-use prioritization.

For biodiversity protection to play a persuasive role in land-use planning, conservationists must be able to offer objective systems for ranking which natural areas to protect or convert. Representing biodiversity in spatially explicit indices is challenging because it entails numerous judgments regarding what variables to measure, how to measure them, and how to combine them. Surprisingly few studies have explored this variation. Here, we explore how this variation affects which areas are selected for agricultural conversion by a land-use prioritization model designed to reduce the biodiversity losses associated with agricultural expansion in Zambia. We first explore the similarity between model recommendations generated by three recently published composite indices and a commonly used rarity-weighted species richness metric. We then explore four underlying sources of ecological and methodological variation within these and other approaches, including different terrestrial vertebrate taxonomic groups, different species-richness metrics, different mathematical methods for combining layers, and different spatial resolutions of inputs. The results generated using different biodiversity approaches show very low spatial agreement regarding which areas to convert to agriculture. There is little overlap in areas identified for conversion using previously published indices (mean Jaccard similarity, Jw , between 0.3 and 3.7%), different taxonomic groups (5.0% < mean Jw  < 13.5%), or different measures of species richness (15.6% < mean Jw  < 33.7%). Even with shared conservation goals, different methods for combining layers and different input spatial resolutions still produce meaningful, though smaller, differences among areas selected for conversion (40.9% < mean Jw  < 67.5%). The choice of taxonomic group had the largest effect on conservation priorities, followed by the choice of species richness metric, the choice of combination method, and finally the choice of spatial resolution. These disagreements highlight the challenge of objectively representing biodiversity in land-use planning tools, and present a credibility challenge for conservation scientists seeking to inform policy making. Our results suggest an urgent need for a more consistent and transparent framework for designing the biodiversity indices used in land-use planning, which we propose here.

A large-area, spatially continuous assessment of land cover map error and its impact on downstream analyses.

Land cover maps increasingly underlie research into socioeconomic and environmental patterns and processes, including global change. It is known that map errors impact our understanding of these phenomena, but quantifying these impacts is difficult because many areas lack adequate reference data. We used a highly accurate, high-resolution map of South African cropland to assess (1) the magnitude of error in several current generation land cover maps, and (2) how these errors propagate in downstream studies. We first quantified pixel-wise errors in the cropland classes of four widely used land cover maps at resolutions ranging from 1 to 100 km, and then calculated errors in several representative "downstream" (map-based) analyses, including assessments of vegetative carbon stocks, evapotranspiration, crop production, and household food security. We also evaluated maps' spatial accuracy based on how precisely they could be used to locate specific landscape features. We found that cropland maps can have substantial biases and poor accuracy at all resolutions (e.g., at 1 km resolution, up to ∼45% underestimates of cropland (bias) and nearly 50% mean absolute error (MAE, describing accuracy); at 100 km, up to 15% underestimates and nearly 20% MAE). National-scale maps derived from higher-resolution imagery were most accurate, followed by multi-map fusion products. Constraining mapped values to match survey statistics may be effective at minimizing bias (provided the statistics are accurate). Errors in downstream analyses could be substantially amplified or muted, depending on the values ascribed to cropland-adjacent covers (e.g., with forest as adjacent cover, carbon map error was 200%-500% greater than in input cropland maps, but ∼40% less for sparse cover types). The average locational error was 6 km (600%). These findings provide deeper insight into the causes and potential consequences of land cover map error, and suggest several recommendations for land cover map users.

Using changes in agricultural utility to quantify future climate-induced risk to conservation.

Much of the biodiversity-related climate change impacts research has focused on the direct effects to species and ecosystems. Far less attention has been paid to the potential ecological consequences of human efforts to address the effects of climate change, which may equal or exceed the direct effects of climate change on biodiversity. One of the most significant human responses is likely to be mediated through changes in the agricultural utility of land. As farmers adapt their practices to changing climates, they may increase pressure on some areas that are important to conserve (conservation lands) whereas lessening it on others. We quantified how the agricultural utility of South African conservation lands may be altered by climate change. We assumed that the probability of an area being farmed is linked to the economic benefits of doing so, using land productivity values to represent production benefit and topographic ruggedness as a proxy for costs associated with mechanical workability. We computed current and future values of maize and wheat production in key conservation lands using the DSSAT4.5 model and 36 crop-climate response scenarios. Most conservation lands had, and were predicted to continue to have, low agricultural utility because of their location in rugged terrain. However, several areas were predicted to maintain or gain high agricultural utility and may therefore be at risk of near-term or future conversion to cropland. Conversely, some areas were predicted to decrease in agricultural utility and may therefore prove easier to protect from conversion. Our study provides an approximate but readily transferable method for incorporating potential human responses to climate change into conservation planning.

Projected climate impacts to South African maize and wheat production in 2055: a comparison of empirical and mechanistic modeling approaches.

Crop model-specific biases are a key uncertainty affecting our understanding of climate change impacts to agriculture. There is increasing research focus on intermodel variation, but comparisons between mechanistic (MMs) and empirical models (EMs) are rare despite both being used widely in this field. We combined MMs and EMs to project future (2055) changes in the potential distribution (suitability) and productivity of maize and spring wheat in South Africa under 18 downscaled climate scenarios (9 models run under 2 emissions scenarios). EMs projected larger yield losses or smaller gains than MMs. The EMs' median-projected maize and wheat yield changes were -3.6% and 6.2%, respectively, compared to 6.5% and 15.2% for the MM. The EM projected a 10% reduction in the potential maize growing area, where the MM projected a 9% gain. Both models showed increases in the potential spring wheat production region (EM = 48%, MM = 20%), but these results were more equivocal because both models (particularly the EM) substantially overestimated the extent of current suitability. The substantial water-use efficiency gains simulated by the MMs under elevated CO2 accounted for much of the EM-MM difference, but EMs may have more accurately represented crop temperature sensitivities. Our results align with earlier studies showing that EMs may show larger climate change losses than MMs. Crop forecasting efforts should expand to include EM-MM comparisons to provide a fuller picture of crop-climate response uncertainties.

High Resolution, Annual Maps of Field Boundaries for Smallholder-Dominated Croplands at National Scales.

Mapping the characteristics of Africa's smallholder-dominated croplands, including the sizes and numbers of fields, can provide critical insights into food security and a range of other socioeconomic and environmental concerns. However, accurately mapping these systems is difficult because there is 1) a spatial and temporal mismatch between satellite sensors and smallholder fields, and 2) a lack of high-quality labels needed to train and assess machine learning classifiers. We developed an approach designed to address these two problems, and used it to map Ghana's croplands. To overcome the spatio-temporal mismatch, we converted daily, high resolution imagery into two cloud-free composites (the primary growing season and subsequent dry season) covering the 2018 agricultural year, providing a seasonal contrast that helps to improve classification accuracy. To address the problem of label availability, we created a platform that rigorously assesses and minimizes label error, and used it to iteratively train a Random Forests classifier with active learning, which identifies the most informative training sample based on prediction uncertainty. Minimizing label errors improved model F1 scores by up to 25%. Active learning increased F1 scores by an average of 9.1% between first and last training iterations, and 2.3% more than models trained with randomly selected labels. We used the resulting 3.7 m map of cropland probabilities within a segmentation algorithm to delineate crop field boundaries. Using an independent map reference sample (n = 1,207), we found that the cropland probability and field boundary maps had respective overall accuracies of 88 and 86.7%, user's accuracies for the cropland class of 61.2 and 78.9%, and producer's accuracies of 67.3 and 58.2%. An unbiased area estimate calculated from the map reference sample indicates that cropland covers 17.1% (15.4-18.9%) of Ghana. Using the most accurate validation labels to correct for biases in the segmented field boundaries map, we estimated that the average size and total number of field in Ghana are 1.73 ha and 1,662,281, respectively. Our results demonstrate an adaptable and transferable approach for developing annual, country-scale maps of crop field boundaries, with several features that effectively mitigate the errors inherent in remote sensing of smallholder-dominated agriculture.

Nano and Micro Unmanned Aerial Vehicles (UAVs): A New Grand Challenge for Precision Agriculture?

By collecting data at spatial and temporal scales that are inaccessible to satellite and field observation, unmanned aerial vehicles (UAVs) are revolutionizing a number of scientific and management disciplines. UAVs may be particularly valuable for precision agricultural applications, offering strong potential to improve the efficiency of water, nutrient, and disease management. However, some authors have suggested that the UAV industry has overhyped the potential value of this technology for agriculture, given that it is difficult for non-specialists to operate UAVs as well as to process and interpret the resulting data. Here, we analyze the barriers to applying UAVs for precision agriculture, which range from regulatory issues to technical requirements. We then evaluate how new developments in the nano- and micro-UAV (NAV and MAV, respectively) markets may help to overcome these barriers. Among the possible breakthroughs that we identify is the ability of NAV/MAV platforms to directly quantify plant traits using methods (e.g., object-oriented classification) that require less image calibration and interpretation than spectral index-based approaches. We suggest that this potential, when combined with steady improvements in sensor miniaturization, flight precision, and autonomy as well as cloud-based image processing, will make UAVs a tool with much broader adoption by agricultural managers in the near future. If this wider uptake is realized, then UAVs have real potential to improve agriculture's resource-use efficiency. © 2020 by John Wiley & Sons, Inc.

The spatial and temporal domains of modern ecology.

To understand ecological phenomena, it is necessary to observe their behaviour across multiple spatial and temporal scales. Since this need was first highlighted in the 1980s, technology has opened previously inaccessible scales to observation. To help to determine whether there have been corresponding changes in the scales observed by modern ecologists, we analysed the resolution, extent, interval and duration of observations (excluding experiments) in 348 studies that have been published between 2004 and 2014. We found that observational scales were generally narrow, because ecologists still primarily use conventional field techniques. In the spatial domain, most observations had resolutions ≤1 m2 and extents ≤10,000 ha. In the temporal domain, most observations were either unreplicated or infrequently repeated (>1 month interval) and ≤1 year in duration. Compared with studies conducted before 2004, observational durations and resolutions appear largely unchanged, but intervals have become finer and extents larger. We also found a large gulf between the scales at which phenomena are actually observed and the scales those observations ostensibly represent, raising concerns about observational comprehensiveness. Furthermore, most studies did not clearly report scale, suggesting that it remains a minor concern. Ecologists can better understand the scales represented by observations by incorporating autocorrelation measures, while journals can promote attentiveness to scale by implementing scale-reporting standards.

Biodiversity at risk under future cropland expansion and intensification.

Agriculture is the leading driver of biodiversity loss. However, its future impact on biodiversity remains unclear, especially because agricultural intensification is often neglected, and high path-dependency is assumed when forecasting agricultural development-although the past suggests that shock events leading to considerable agricultural change occur frequently. Here, we investigate the possible impacts on biodiversity of pathways of expansion and intensification. Our pathways are not built to reach equivalent production targets, and therefore they should not be directly compared; they instead highlight areas at risk of high biodiversity loss across the entire option space of possible agricultural change. Based on an extensive database of biodiversity responses to agriculture, we find 30% of species richness and 31% of species abundances potentially lost because of agricultural expansion across the Amazon and Afrotropics. Only 21% of high-risk expansion areas in the Afrotropics overlap with protected areas (compared with 43% of the Neotropics). Areas at risk of biodiversity loss from intensification are found in India, Eastern Europe and the Afromontane region (7% species richness, 13% abundance loss). Many high-risk regions are not adequately covered by conservation prioritization schemes, and have low national conservation spending and high agricultural growth. Considering rising agricultural demand, we highlight areas where timely land-use planning may proactively mitigate biodiversity loss.