LarvaTagger: manual and automatic tagging of Drosophila larval behaviour.

MOTIVATION: As more behavioural assays are carried out in large-scale experiments on Drosophila larvae, the definitions of the archetypal actions of a larva are regularly refined. In addition, video recording and tracking technologies constantly evolve. Consequently, automatic tagging tools for Drosophila larval behaviour must be retrained to learn new representations from new data. However, existing tools cannot transfer knowledge from large amounts of previously accumulated data. We introduce LarvaTagger, a piece of software that combines a pre-trained deep neural network, providing a continuous latent representation of larva actions for stereotypical behaviour identification, with a graphical user interface to manually tag the behaviour and train new automatic taggers with the updated ground truth. RESULTS: We reproduced results from an automatic tagger with high accuracy, and we demonstrated that pre-training on large databases accelerates the training of a new tagger, achieving similar prediction accuracy using less data. AVAILABILITY AND IMPLEMENTATION: All the code is free and open source. Docker images are also available. See gitlab.pasteur.fr/nyx/LarvaTagger.jl.

Identifying historical and future potential lake drainage events on the western Arctic coastal plain of Alaska.

Arctic lakes located in permafrost regions are susceptible to catastrophic drainage. In this study, we reconstructed historical lake drainage events on the western Arctic Coastal Plain of Alaska between 1955 and 2017 using USGS topographic maps, historical aerial photography (1955), and Landsat Imagery (ca. 1975, ca. 2000, and annually since 2000). We identified 98 lakes larger than 10 ha that partially (>25% of area) or completely drained during the 62-year period. Decadal-scale lake drainage rates progressively declined from 2.0 lakes/yr (1955-1975), to 1.6 lakes/yr (1975-2000), and to 1.2 lakes/yr (2000-2017) in the ~30,000-km2 study area. Detailed Landsat trend analysis between 2000 and 2017 identified two years, 2004 and 2006, with a cluster (five or more) of lake drainages probably associated with bank overtopping or headward erosion. To identify future potential lake drainages, we combined the historical lake drainage observations with a geospatial dataset describing lake elevation, hydrologic connectivity, and adjacent lake margin topographic gradients developed with a 5-m-resolution digital surface model. We identified ~1900 lakes likely to be prone to drainage in the future. Of the 20 lakes that drained in the most recent study period, 85% were identified in this future lake drainage potential dataset. Our assessment of historical lake drainage magnitude, mechanisms and pathways, and identification of potential future lake drainages provides insights into how arctic lowland landscapes may change and evolve in the coming decades to centuries.

Spatiotemporal remote sensing of ecosystem change and causation across Alaska.

Contemporary climate change in Alaska has resulted in amplified rates of press and pulse disturbances that drive ecosystem change with significant consequences for socio-environmental systems. Despite the vulnerability of Arctic and boreal landscapes to change, little has been done to characterize landscape change and associated drivers across northern high-latitude ecosystems. Here we characterize the historical sensitivity of Alaska's ecosystems to environmental change and anthropogenic disturbances using expert knowledge, remote sensing data, and spatiotemporal analyses and modeling. Time-series analysis of moderate-and high-resolution imagery was used to characterize land- and water-surface dynamics across Alaska. Some 430,000 interpretations of ecological and geomorphological change were made using historical air photos and satellite imagery, and corroborate land-surface greening, browning, and wetness/moisture trend parameters derived from peak-growing season Landsat imagery acquired from 1984 to 2015. The time series of change metrics, together with climatic data and maps of landscape characteristics, were incorporated into a modeling framework for mapping and understanding of drivers of change throughout Alaska. According to our analysis, approximately 13% (~174,000 ± 8700 km2 ) of Alaska has experienced directional change in the last 32 years (±95% confidence intervals). At the ecoregions level, substantial increases in remotely sensed vegetation productivity were most pronounced in western and northern foothills of Alaska, which is explained by vegetation growth associated with increasing air temperatures. Significant browning trends were largely the result of recent wildfires in interior Alaska, but browning trends are also driven by increases in evaporative demand and surface-water gains that have predominately occurred over warming permafrost landscapes. Increased rates of photosynthetic activity are associated with stabilization and recovery processes following wildfire, timber harvesting, insect damage, thermokarst, glacial retreat, and lake infilling and drainage events. Our results fill a critical gap in the understanding of historical and potential future trajectories of change in northern high-latitude regions.

Tundra be dammed: Beaver colonization of the Arctic.

Increasing air temperatures are changing the arctic tundra biome. Permafrost is thawing, snow duration is decreasing, shrub vegetation is proliferating, and boreal wildlife is encroaching. Here we present evidence of the recent range expansion of North American beaver (Castor canadensis) into the Arctic, and consider how this ecosystem engineer might reshape the landscape, biodiversity, and ecosystem processes. We developed a remote sensing approach that maps formation and disappearance of ponds associated with beaver activity. Since 1999, 56 new beaver pond complexes were identified, indicating that beavers are colonizing a predominantly tundra region (18,293 km2 ) of northwest Alaska. It is unclear how improved tundra stream habitat, population rebound following overtrapping for furs, or other factors are contributing to beaver range expansion. We discuss rates and likely routes of tundra beaver colonization, as well as effects on permafrost, stream ice regimes, and freshwater and riparian habitat. Beaver ponds and associated hydrologic changes are thawing permafrost. Pond formation increases winter water temperatures in the pond and downstream, likely creating new and more varied aquatic habitat, but specific biological implications are unknown. Beavers create dynamic wetlands and are agents of disturbance that may enhance ecosystem responses to warming in the Arctic.

Post-fire stabilization of thaw-affected permafrost terrain in northern Alaska.

In 2007, the Anaktuvuk River fire burned more than 1000 km2 of arctic tundra in northern Alaska, ~ 50% of which occurred in an area with ice-rich syngenetic permafrost (Yedoma). By 2014, widespread degradation of ice wedges was apparent in the Yedoma region. In a 50 km2 area, thaw subsidence was detected across 15% of the land area in repeat airborne LiDAR data acquired in 2009 and 2014. Updating observations with a 2021 airborne LiDAR dataset show that additional thaw subsidence was detected in < 1% of the study area, indicating stabilization of the thaw-affected permafrost terrain. Ground temperature measurements between 2010 and 2015 indicated that the number of near-surface soil thawing-degree-days at the burn site were 3 × greater than at an unburned control site, but by 2022 the number was reduced to 1.3 × greater. Mean annual ground temperature of the near-surface permafrost increased by 0.33 °C/yr in the burn site up to 7-years post-fire, but then cooled by 0.15 °C/yr in the subsequent eight years, while temperatures at the control site remained relatively stable. Permafrost cores collected from ice-wedge troughs (n = 41) and polygon centers (n = 8) revealed the presence of a thaw unconformity, that in most cases was overlain by a recovered permafrost layer that averaged 14.2 cm and 18.3 cm, respectively. Taken together, our observations highlight that the initial degradation of ice-rich permafrost following the Anaktuvuk River tundra fire has been followed by a period of thaw cessation, permafrost aggradation, and terrain stabilization.

Current and projected flood exposure for Alaska coastal communities.

Globally, coastal communities experience flood hazards that are projected to worsen from climate change and sea level rise. The 100-year floodplain or record flood are commonly used to identify risk areas for planning purposes. Remote communities often lack measured flood elevations and require innovative approaches to estimate flood elevations. This study employs observation-based methods to estimate the record flood elevation in Alaska communities and compares results to elevation models, infrastructure locations, and sea level rise projections. In 46 analyzed communities, 22% of structures are located within the record floodplain. With sea level rise projections, this estimate increases to 30-37% of structures by 2100 if structures remain in the same location. Flood exposure is highest in western Alaska. Sea level rise projections suggest northern Alaska will see similar flood exposure levels by 2100 as currently experienced in western Alaska. This evaluation of record flood height, category, and history can be incorporated into hazard planning documents, providing more context for coastal flood exposure than previously existed for Alaska. This basic flood exposure method is transferable to other areas with similar mapping challenges. Identifying current and projected hazardous zones is essential to avoid unintentional development in floodplains and improve long-term safety.

Nepafenac: an ophthalmic nonsteroidal antiinflammatory drug for pain after cataract surgery.

OBJECTIVE: To examine the efficacy of nepafenac in the treatment of pain and inflammation in patients after cataract surgery using evidence from controlled clinical studies. DATA SOURCES: Citations in Google Scholar, PubMed, and Web of Science from January 1, 2005, to March 25, 2013, were identified using nepafenac and cataract as search terms. STUDY SELECTION AND DATA EXTRACTION: The literature search was limited to human studies published in English. Three trials that compared nepafenac with other nonsteroidal antiinflammatory drugs (NSAIDs) were included. DATA SYNTHESIS: The pharmacokinetics and pharmacodynamics of nepafenac 0.1% suspension (and its active metabolite, amfenac) were compared with bromfenac 0.09% solution and ketorolac 0.4% solution with respect to aqueous humor concentrations and ability to reduce cyclooxygenase 1 and 2 (COX-1 and COX-2) enzymes. The maximum concentration (C(max)) values of ketorolac and amfenac were statistically similar, while the C(max) of bromfenac was significantly lower than that of amfenac. Ketorolac most effectively inhibited COX-1 enzymes; COX-2 enzymes were most effectively reduced by amfenac. When nepafenac 0.1% suspension was compared with placebo and ketorolac 0.5% solution, nepafenac achieved a higher percentage cure rate than placebo at day 14 (p = 0.0241). Significant differences in cure rates between nepafenac and ketorolac were not observed. Nepafenac 0.1%, bromfenac 0.09%, and ketorolac 0.45% were compared to determine which most effectively reduced prostaglandin E2 (PGE2) following surgery. PGE2 concentrations were significantly lowest in the ketorolac group, followed by the bromfenac and nepafenac groups, respectively. Topical nepafenac 0.1% suspension was approved in 2005. A 0.3% suspension was approved in October 2012. The 0.3% product may have some advantages over its predecessor: it is dosed once rather than thrice daily, which may increase patient adherence and improve outcomes. The price and dosing frequency of the 0.3% product are comparable to those of bromfenac 0.09% solution. CONCLUSIONS: The 2 nepafenac products appear to be equally efficacious, with a slightly increased adverse event rate in patients using the 0.3% versus 0.1% formulation. Head-to-head clinical trials that compare the 0.3% product with the 0.1% product or other commercially available NSAIDs are unavailable.

Expanding beaver pond distribution in Arctic Alaska, 1949 to 2019.

Beavers were not previously recognized as an Arctic species, and their engineering in the tundra is considered negligible. Recent findings suggest that beavers have moved into Arctic tundra regions and are controlling surface water dynamics, which strongly influence permafrost and landscape stability. Here we use 70 years of satellite images and aerial photography to show the scale and magnitude of northwestward beaver expansion in Alaska, indicated by the construction of over 10,000 beaver ponds in the Arctic tundra. The number of beaver ponds doubled in most areas between ~ 2003 and ~ 2017. Earlier stages of beaver engineering are evident in ~ 1980 imagery, and there is no evidence of beaver engineering in ~ 1952 imagery, consistent with observations from Indigenous communities describing the influx of beavers over the period. Rapidly expanding beaver engineering has created a tundra disturbance regime that appears to be thawing permafrost and exacerbating the effects of climate change.

High-throughput automated methods for classical and operant conditioning of Drosophila larvae.

Learning which stimuli (classical conditioning) or which actions (operant conditioning) predict rewards or punishments can improve chances of survival. However, the circuit mechanisms that underlie distinct types of associative learning are still not fully understood. Automated, high-throughput paradigms for studying different types of associative learning, combined with manipulation of specific neurons in freely behaving animals, can help advance this field. The Drosophila melanogaster larva is a tractable model system for studying the circuit basis of behaviour, but many forms of associative learning have not yet been demonstrated in this animal. Here, we developed a high-throughput (i.e. multi-larva) training system that combines real-time behaviour detection of freely moving larvae with targeted opto- and thermogenetic stimulation of tracked animals. Both stimuli are controlled in either open- or closed-loop, and delivered with high temporal and spatial precision. Using this tracker, we show for the first time that Drosophila larvae can perform classical conditioning with no overlap between sensory stimuli (i.e. trace conditioning). We also demonstrate that larvae are capable of operant conditioning by inducing a bend direction preference through optogenetic activation of reward-encoding serotonergic neurons. Our results extend the known associative learning capacities of Drosophila larvae. Our automated training rig will facilitate the study of many different forms of associative learning and the identification of the neural circuits that underpin them.

Understanding the Effects of Optimal Combination of Spectral Bands on Deep Learning Model Predictions: A Case Study Based on Permafrost Tundra Landform Mapping Using High Resolution Multispectral Satellite Imagery.

Deep learning (DL) convolutional neural networks (CNNs) have been rapidly adapted in very high spatial resolution (VHSR) satellite image analysis. DLCNN-based computer visions (CV) applications primarily aim for everyday object detection from standard red, green, blue (RGB) imagery, while earth science remote sensing applications focus on geo object detection and classification from multispectral (MS) imagery. MS imagery includes RGB and narrow spectral channels from near- and/or middle-infrared regions of reflectance spectra. The central objective of this exploratory study is to understand to what degree MS band statistics govern DLCNN model predictions. We scaffold our analysis on a case study that uses Arctic tundra permafrost landform features called ice-wedge polygons (IWPs) as candidate geo objects. We choose Mask RCNN as the DLCNN architecture to detect IWPs from eight-band Worldview-02 VHSR satellite imagery. A systematic experiment was designed to understand the impact on choosing the optimal three-band combination in model prediction. We tasked five cohorts of three-band combinations coupled with statistical measures to gauge the spectral variability of input MS bands. The candidate scenes produced high model detection accuracies for the F1 score, ranging between 0.89 to 0.95, for two different band combinations (coastal blue, blue, green (1,2,3) and green, yellow, red (3,4,5)). The mapping workflow discerned the IWPs by exhibiting low random and systematic error in the order of 0.17-0.19 and 0.20-0.21, respectively, for band combinations (1,2,3). Results suggest that the prediction accuracy of the Mask-RCNN model is significantly influenced by the input MS bands. Overall, our findings accentuate the importance of considering the image statistics of input MS bands and careful selection of optimal bands for DLCNN predictions when DLCNN architectures are restricted to three spectral channels.