Exceedingly high or low concentrations of essential elements and heavy metals can have devastating effects on homeostasis and the health of living organisms. Baseline ranges for many elements have been widely assessed and studied for domesticated species, but such a baseline is lacking for squirrels, especially in remote areas such as Alaska (United States). North American red squirrels (Tamiasciurus hudsonicus) are mesopredators that interact with a vast array of ecosystem components (fungi, seeds and vegetation, predators, and species that co-evolved with squirrels), while being able to thrive in semi-urbanized landscapes of the Anthropocene. As such, they can serve as a sentinel species for ecosystem health. To assess baseline heavy metal exposure and essential element concentrations in Interior Alaskan red squirrels, we collected 158 squirrels through citizen crowdsourcing and extracted livers from each squirrel during necropsies. We grouped the 158 squirrels into 11 composite samples (based on the location of the deceased animal) and analyzed liver tissues for 71 elements using Inductively Coupled Plasma-Sector Field Mass Spectrometry (ICP-MS). Our laboratory results mostly fall within the range reported in the few available literature records for liver tissue analyzed within the Sciuridae family, but tend to be found in the highest quarter of comparable literature records for manganese (Mn), cadmium (Cd), mercury (Hg), nickel (Ni), and lead (Pb). Of the total 71 elements measured, we analyzed 21 using the TreeNet Gradient Boosting Machine Learning algorithm in the Salford Predictive Modeler (SPM) and present three elements (As, Se, and Cd) as case studies. These 21 elements were chosen based on our judgment of their relevance for the health of red squirrels in the study area. We used 224 environmental predictors and ‘scored’ the predictions in SPM, linking the aspatial TreeNet models to the corresponding spatial location. We then generate a predicted geo-referenced element concentration value (data point) for each pixel for each assessed element. Hence, the exploratory spatial predictions enabled us to increase the original composite sample size of 11 sites to a much larger Alaskan landscape (approximately 656,000 predicted sites), leading us to the territory of Big Data approaches, representing a core novelty aspect of this study. Some of the major heavy metal prediction hotspots were found to be near active and inactive military sites (e.g., Murphy Dome and Salcha sites in Interior Alaska for Se, Cd, and As), raising potential concerns for heavy metal contaminations of nearby landscapes. This approach can be further utilized as a planning tool for future on-the-ground sampling of squirrels or other components of the landscape.

This dataset describes public data used for a global orca whale distribution model and conservation assessment using GIS and Machine Learning/AI; details are pending peer review.

In times of human-induced and accelerating climate change and the sixth mass extinction, it remains unknown where species will persist and thrive in the future. This study explores a holistic assemblage-based conservation management assessment using all squirrel species currently occurring in Alaska (six) and the adjacent 600 km buffer region (an additional six) as an exemplary set of study species. We conducted this assessment by predicting the current and future distributions (based on three different Global Climate Models) of all 12 squirrel species, both individually and as an assemblage, and analyzing how their distributions change over time. We then further investigated the actual changes between the current and future predicted distribution and utilized the RAD (Resist, Adapt, Direct) decision framework to develop conservation management actions in response to expected climate change trajectories in Alaska. We found that the current predicted distribution and associated niche collapses for several species by the year 2100, and drastically shifts for most. If the assemblage-wide biodiversity loss is to be minimized throughout the upcoming approximately 75 years, considerable conservation and habitat management need to be planned and successfully implemented. Without meaningful action, at least two species (out of 12; 17%), including an Alaska endemic species, are predicted to suffer drastic population declines and even become extinct.

Context. Squirrel species in Alaska generally lack thorough conservation management plans, while all species are actively hunted with no bag limits, closed seasons, or any other restrictions, and the current ‘management’ is based on ambiguous hand-drawn distribution maps. This indicates a laissez-faire approach to Alaskan squirrel conservation management. Aims. Here, we attempt to improve this current situation by assessing the effectiveness of ensemble machine-learning prediction models as proposed add-ons to the traditional components of conservation management plans toward a more state-of-the-art approach to species conservation. Methods. We combined the Machine Learning algorithms TreeNet, CART, Random Forest, and Maxent with over 200 environmental and socio-economic predictors for the ensemble Super Species Distribution Models. These ensemble models were carried out for all squirrel species individually occurring in Alaska and a 600 km buffer area and two assemblage models combined: a) all species currently occurring only in Alaska and b) all species occurring in Alaska and the 600km buffer area. Key results. Most predicted distribution hotspots for squirrels in Alaska and the 600 km buffer area were observed near road and river systems (close to human activities) and the last glacial maximum refugia. Conclusions & Implications. Applying a machine learning ensemble distribution modeling framework to conservation management plans can add valuable science-based insights to better understand the landscape and species to be managed. Such insights include more accurate guidance on, e.g., habitat protection, hunting regulations, or any collaborative management initiatives. This can also be highly valuable for lands not directly managed by conventional agencies, e.g., land managed by the military or Native communities throughout the Pacific Rim.

Plants with carnivorous traits are of great importance. However, it is globally noticed that without effective conservation, a declining number of suitable habitats, and the wide destruction of currently realized niches, their future does not appear bright. In a region that had all possibilities to “get things right” upon its statehood being granted in 1959, Alaska has, up to the present moment, consistently fallen short in stabilizing an effective conservation system for this group of plant traits and its habitats. Coupled with many other socio-ecological issues that persist from the colonial and industrial past and worsen over time, as well as the rise of new issues such as global climate change and “the great acceleration,” Alaska presents virtually all the typical indicators of a failed socio-ecological state (as a term used by The World Bank). For a wider assessment, here we compiled all Open Access data in the public realm related to eight carnivorous plant species in Alaska and overlaid them for an assessment with nine predictor layers. Beyond raw data, we developed ensemble models serving the purpose of indicating generalized hotspots and coldspots of plants with carnivorous traits, also using citizen-science occurrence data. Further, we analyzed the predicted occurrence with the underlaying land ownership/use types as well as mining claims as an example of a leading industrial activity. Alaska actually hosts the majority of the US National Park System, but we find the majority of carnivorous plants located outside, and a higher predicted occurrence within officially designated mining areas than outside. In our assessment, we see no relevant policy, vision, efficient action, or principles of strategic conservation management applied to plants and their traits, specifically carnivorous plants in Alaska and its leadership. Judged by the major socio-ecological metrics, it confirms evidence that Alaska does not present basic performance metrics of good natural resource management, and thus it would meet the definition of a failed state.

Describing where distribution hotspots and coldspots are located with certainty is crucial for any science-based species management and governance. Thus, here we created the world’s first Super Species Distribution Models (SDMs) including all primate species and the best-available predictor set. These Super SDMs are conducted using modern Machine Learning ensembles like Maxent, TreeNet, RandomForest, CART, CART Boosting and Bagging, and MARS with the utilization of cloud supercomputers (as an add-on option for more powerful models). For the global cold/ hotspot models, we obtained global distribution data from www.GBIF.org (approx. 420,000 raw occurrence records) and utilized the world’s largest environmental predictor set of 201 layers. For this analysis, all occurrences have been merged into one multi-species (400+ species) pixel-based analysis. We quantified the global primate hotspots for Central and Northern South America, West Africa, East Africa, Southeast Asia, Central Asia, and Southern Africa. The global primate coldspots are Antarctica, the Arctic, most temperate regions, and Oceania past the Wallace line. We additionally described all these modeled hotspots/coldspots and discussed reasons for a quantified understanding of where the world’s primates occur (or not). This shows us where the focus for most future research and conservation management efforts should be, using state-of-the-art digital data indication tools with reason. Those areas should be considered of the highest conservation priority, ideally following ‘no killing zones’ and sustainable land stewardship approaches if primates are to have a chance of survival.

With increased globalization, man-made climate change, and urbanization, the landscape – embedded within the Anthropocene - becomes increasingly fragmented. With habitats transitioning and getting lost, globally relevant regions considered ‘pristine', such as Alaska, are no exception. Alaska holds 60% of the U.S. National Park system’s area and is of national and international importance, considering the U.S. is one of the wealthiest nations on earth. These characteristics tie into densities and quantities of human features, e.g., roads, houses, mines, wind parks, agriculture, trails, etc., that can be summarized as ‘impervious surfaces.’ Those are physical impacts and actively affecting urban-driven landscape fragmentation. Using the remote sensing data of the National Land Cover Database (NLCD; https://www.mrlc.gov/data/nlcd-2016-land-cover-alaska ), here we attempt to create the first quantification of this physical human impact on the Alaskan landscape and its fragmentation. We quantified these impacts using the well-established landscape metrics tool ‘Fragstats’, implemented as the R package “landscapemetrics” in the desktop software and through the interface of a Linux Cloud-computing environment. This workflow allows for the first time to overcome the computational limitations of the conventional Fragstats software within a reasonably quick timeframe. Thereby, we are able to analyze a land area as large as approx. 1,517,733 km2 (state of Alaska) while maintaining a high assessment resolution of 30 meters. Based on this traditional methodology, we found that Alaska has a reported physical human impact of c. 0.067%. But when assessed, we additionally overlaid other features that were not included in the input data to highlight the overall true human impact (e.g., roads, trails, airports, governance boundaries in game management and park units, mines, etc.). We found that using remote sensing (human impact layers), Alaska’s human impact is considerably underestimated to a meaningless estimate (0.067%). The state is more seriously fragmented and affected by humans than commonly assumed. Very few areas are truly untouched and display a high patch density with corresponding low mean patch sizes throughout the study area. Instead, the true human impact is likely close to 100% throughout Alaska for several metrics. With these newly created insights, we provide the first state-wide landscape data and inference that are likely of considerable importance for land management entities in the state of Alaska, and for the U.S. National Park systems overall, especially in the changing climate. Likewise, the methodological framework presented here shows an Open Access workflow and can be used as a reference to be reproduced virtually anywhere else on the planet to assess more realistic large-scale landscape metrics. It can also be used to assess human impacts on the landscape for more sustainable landscape stewardship and mitigation in policy.

This dataset contains 200+ environmental predictors used for a series of scientific publications, obtained from public sources. Curated and geo-spatially aligned by Moriz Steiner.

These are the GIS data used for modeling Matschie's Tree Kangaroo (Huon Tree Kangaroo) in Papua New Guinea PNG; for details please see metadata. THe manuscript is currently in revision phase.