The classification of savanna woodland tree species from high-resolution Remotely Piloted Aircraft Systems (RPAS) imagery is a complex and challenging task. Difficulties for both traditional remote sensing algorithms and human observers arise due to low interspecies variability (species difficult to discriminate because they are morphologically similar) and high intraspecies variability (individuals of the same species varying to the extent that they can be misclassified), and the loss of some taxonomic features commonly used for identification when observing trees from above. Deep neural networks are increasingly being used to overcome challenges in image recognition tasks. However, supervised deep learning algorithms require high-quality annotated and labelled training data that must be verified by subject matter experts. While training datasets for trees have been generated and made publicly available, they are mostly acquired in the Northern Hemisphere and lack species-level information. We present a training dataset of tropical Northern Australia savanna woodland tree species that was generated using RPAS and on-ground surveys to confirm species labels. RPAS-derived imagery was annotated, resulting in 2547 polygons representing 36 tree species. A baseline dataset was produced consisting of: (i) seven orthomosaics that were used for in-field labelling; (ii) a tiled dataset at 1024 × 1024 pixel size in Common Objects in Context (COCO) format that can be used for deep learning model training; (iii) and the annotations.

Northern Australia's savannas are among the most fire‐prone biomes on Earth and are dominated by eucalypts (Eucalyptus and Corymbia spp.). It is not clear what processes allow this group to dominate under such extreme fire frequencies and whether a superior ability to compete for nutrients and water might play a role. There is evidence that eucalypts are adapted to frequent fires; juvenile eucalypts escape the fire trap by growing rapidly in height between fires. However, non‐eucalypts are less able to escape the fire trap and tend to have stand structures strongly skewed toward suppressed juveniles. The mechanisms that drive these contrasting fire responses are not well understood. Here, we describe the results of a controlled glasshouse seedling experiment that evaluated the relative importance of nutrient and water availability in determining height growth and biomass growth of two eucalypt and one noneucalypt tree species, common in northern Australian savannas. We demonstrate that growth of eucalypt seedlings is particularly responsive to nutrient addition. Eucalypt seedlings are able to rapidly utilize soil nutrients and accumulate biomass at a much greater rate than noneucalypt seedlings. We suggest that a seasonal spike in nutrient availability creates a nutrient‐rich microsite that allows eucalypt seedlings to rapidly gain height and biomass, increasing their likelihood of establishing successfully and reaching a fire‐resistant size. Our results extend our understanding of how eucalypts dominate northern Australian savannas under extremely high fire frequencies.

Land-use conversion to non-native species plantations not only affects biodiversity but also alters important ecosystem functions including above- and below-ground carbon sequestration, and CO2 release rates from soils via soil respiration. Though the role of soil temperature and moisture on soil respiration is well recognized, little is known about how their effects vary across different land-use types. This study looked at the effects of land-cover change on temporal patterns of soil respiration in a montane forest-grassland-plantation matrix, a highly diverse but climatically sensitive ecosystem in the tropical Western Ghats of India. Among native vegetation types, soil respiration rates were higher in grassland compared to forest patches. Invasion of grassland by an exotic tree species - wattle (Acacia mearnsii) reduced soil respiration rates to levels similar to that of forests. However, conversion of native grasslands to non-native pine (Pinus patula) plantations led to the largest declines in soil respiration rates. In addition, the sensitivity of soil respiration to changes in temperature and moisture differed between different vegetation types. Across all vegetation types, respiration was largely insensitive to changes in soil temperature when moisture levels were low. However, when soil moisture levels were high, respiration increased with temperature in grassland and wattle patches, decreased in the case of pine plantations, and remained largely unchanged in shola forests. Our results suggest that changes in aboveground vegetation type can significantly affect soil C cycling even in the absence of any underlying differences in soil type.