General Electric (Finland)

Abstract A large array of species distribution model ( SDM ) approaches has been developed for explaining and predicting the occurrences of individual species or species assemblages. Given the wealth of existing models, it is unclear which models perform best for interpolation or extrapolation of existing data sets, particularly when one is concerned with species assemblages. We compared the predictive performance of 33 variants of 15 widely applied and recently emerged SDM s in the context of multispecies data, including both joint SDM s that model multiple species together, and stacked SDM s that model each species individually combining the predictions afterward. We offer a comprehensive evaluation of these SDM approaches by examining their performance in predicting withheld empirical validation data of different sizes representing five different taxonomic groups, and for prediction tasks related to both interpolation and extrapolation. We measure predictive performance by 12 measures of accuracy, discrimination power, calibration, and precision of predictions, for the biological levels of species occurrence, species richness, and community composition. Our results show large variation among the models in their predictive performance, especially for communities comprising many species that are rare. The results do not reveal any major trade‐offs among measures of model performance; the same models performed generally well in terms of accuracy, discrimination, and calibration, and for the biological levels of individual species, species richness, and community composition. In contrast, the models that gave the most precise predictions were not well calibrated, suggesting that poorly performing models can make overconfident predictions. However, none of the models performed well for all prediction tasks. As a general strategy, we therefore propose that researchers fit a small set of models showing complementary performance, and then apply a cross‐validation procedure involving separate data to establish which of these models performs best for the goal of the study.

Social media data is increasingly used as a proxy for human activity in different environments, including protected areas, where collecting visitor information is often laborious and expensive, but important for management and marketing. Here, we compared data from Instagram, Twitter and Flickr, and assessed systematically how park popularity and temporal visitor counts derived from social media data perform against high-precision visitor statistics in 56 national parks in Finland and South Africa in 2014. We show that social media activity is highly associated with park popularity, and social media-based monthly visitation patterns match relatively well with the official visitor counts. However, there were considerable differences between platforms as Instagram clearly outperformed Twitter and Flickr. Furthermore, we show that social media data tend to perform better in more visited parks, and should always be used with caution. Based on stakeholder discussions we identified potential reasons why social media data and visitor statistics might not match: the geography and profile of the park, the visitor profile, and sudden events. Overall the results are encouraging in broader terms: Over 60% of the national parks globally have Twitter or Instagram activity, which could potentially inform global nature conservation.

Abstract Can social media data be used as an alternative to traditional surveys to understand tourists’ preferences for nature‐based experiences in protected areas? We explored this by comparing preferences for biodiversity obtained from a traditional survey conducted in Kruger National Park, South Africa, with observed preferences assessed from over 13,600 pictures shared on Instagram and Flickr by tourists visiting the park in the same period. We found no significant difference between the preferences of tourists as stated in the surveys and the preferences revealed by social media content. Overall, large‐bodied mammals were found to be the favorite group, both in the survey and on social media platforms. However, Flickr was found to better match tourists’ preference for less‐charismatic biodiversity. Our findings suggest that social media content can be used as a cost‐efficient way to explore, and for more continuous monitoring of, preferences for biodiversity and human activities in protected areas.

Conservation of species and ecosystems is increasingly difficult because anthropogenic impacts are pervasive and accelerating. Under this rapid global change, maximizing conservation success requires a paradigm shift from maintaining ecosystems in idealized past states toward facilitating their adaptive and functional capacities, even as species ebb and flow individually. Developing effective strategies under this new paradigm will require deeper understanding of the long-term dynamics that govern ecosystem persistence and reconciliation of conflicts among approaches to conserving historical versus novel ecosystems. Integrating emerging information from conservation biology, paleobiology, and the Earth sciences is an important step forward on the path to success. Maintaining nature in all its aspects will also entail immediately addressing the overarching threats of growing human population, overconsumption, pollution, and climate change.

BACKGROUND: Dexmedetomidine, a selective alpha(2)-adrenoceptor agonist, induces a unique, sleep-like state of sedation. The objective of the present work was to study human electroencephalogram (EEG) sleep spindles during dexmedetomidine sedation and compare them with spindles during normal physiological sleep, to test the hypothesis that dexmedetomidine exerts its effects via normal sleep-promoting pathways. METHODS: EEG was continuously recorded from a bipolar frontopolar-laterofrontal derivation with Entropy Module (GE Healthcare) during light and deep dexmedetomidine sedation (target-controlled infusions set at 0.5 and 3.2 ng/ml) in 11 healthy subjects, and during physiological sleep in 10 healthy control subjects. Sleep spindles were visually scored and quantitatively analyzed for density, duration, amplitude (band-pass filtering) and frequency content (matching pursuit approach), and compared between the two groups. RESULTS: In visual analysis, EEG activity during dexmedetomidine sedation was similar to physiological stage 2 (S2) sleep with slight to moderate amount of slow-wave activity and abundant sleep spindle activity. In quantitative EEG analyses, sleep spindles were similar during dexmedetomidine sedation and normal sleep. No statistically significant differences were found in spindle density, amplitude or frequency content, but the spindles during dexmedetomidine sedation had longer duration (mean 1.11 s, SD 0.14 s) than spindles in normal sleep (mean 0.88 s, SD 0.14 s; P=0.0014). CONCLUSIONS: Analysis of sleep spindles shows that dexmedetomidine produces a state closely resembling physiological S2 sleep in humans, which gives further support to earlier experimental evidence for activation of normal non-rapid eye movement sleep-promoting pathways by this sedative agent.

Data collected under the auspices of the BIFROST GPS project yield a geographically dense suite of estimates of present‐day, three‐dimensional (3‐D) crustal deformation rates in Fennoscandia [ Johansson et al. , 2002 ]. A preliminary forward analysis of these estimates [ Milne et al. , 2001 ] has indicated that models of ongoing glacial isostatic adjustment (GIA) in response to the final deglaciation event of the current ice age are able to provide an excellent fit to the observed 3‐D velocity field. In this study we revisit our previous GIA analysis by considering a more extensive suite of forward calculations and by performing the first formal joint inversion of the BIFROST rate estimates. To establish insight into the physics of the GIA response in the region, we begin by decomposing a forward prediction into the three contributions associated with the ice, ocean, and rotational forcings. From this analysis we demonstrate that recent advances in postglacial sea level theory, in particular the inclusion of rotational effects and improvements in the treatment of the ocean load in the vicinity of an evolving continental margin, involve peak signals that are larger than the observational uncertainties in the BIFROST network. The forward analysis is completed by presenting predictions for a pair of Fennoscandian ice histories and an extensive suite of viscoelastic Earth models. The former indicates that the BIFROST data set provides a powerful discriminant of such histories. The latter yields bounds on the (assumed constant) upper and lower mantle viscosity (ν UM , ν LM ); specifically, we derive a 95% confidence interval of 5 × 10 20 ≤ ν UM ≤ 10 21 Pa s and 5 × 10 21 ≤ ν LM ≤ 5 × 10 22 Pa s, with some preference for (elastic) lithospheric thickness in excess of 100 km. The main goal of the (Bayesian) inverse analysis is to estimate the radial resolving power of the BIFROST GPS data as a function of depth in the mantle. Assuming a reasonably accurate ice history, we demonstrate that this resolving power varies from ∼200 km near the base of the upper mantle to ∼700 km in the top portion of the lower mantle. We conclude that the BIFROST data are able to resolve structure on radial length scale significantly smaller than a single upper mantle layer. However, these data provide little constraint on viscosity in the bottom half of the mantle. Finally, elements of both the forward and inverse analyses indicate that radial and horizontal velocity estimates provide distinct constraints on mantle viscosity.

Recent studies from mountainous areas of small spatial extent (<2500 km(2) ) suggest that fine-grained thermal variability over tens or hundreds of metres exceeds much of the climate warming expected for the coming decades. Such variability in temperature provides buffering to mitigate climate-change impacts. Is this local spatial buffering restricted to topographically complex terrains? To answer this, we here study fine-grained thermal variability across a 2500-km wide latitudinal gradient in Northern Europe encompassing a large array of topographic complexities. We first combined plant community data, Ellenberg temperature indicator values, locally measured temperatures (LmT) and globally interpolated temperatures (GiT) in a modelling framework to infer biologically relevant temperature conditions from plant assemblages within <1000-m(2) units (community-inferred temperatures: CiT). We then assessed: (1) CiT range (thermal variability) within 1-km(2) units; (2) the relationship between CiT range and topographically and geographically derived predictors at 1-km resolution; and (3) whether spatial turnover in CiT is greater than spatial turnover in GiT within 100-km(2) units. Ellenberg temperature indicator values in combination with plant assemblages explained 46-72% of variation in LmT and 92-96% of variation in GiT during the growing season (June, July, August). Growing-season CiT range within 1-km(2) units peaked at 60-65°N and increased with terrain roughness, averaging 1.97 °C (SD = 0.84 °C) and 2.68 °C (SD = 1.26 °C) within the flattest and roughest units respectively. Complex interactions between topography-related variables and latitude explained 35% of variation in growing-season CiT range when accounting for sampling effort and residual spatial autocorrelation. Spatial turnover in growing-season CiT within 100-km(2) units was, on average, 1.8 times greater (0.32 °C km(-1) ) than spatial turnover in growing-season GiT (0.18 °C km(-1) ). We conclude that thermal variability within 1-km(2) units strongly increases local spatial buffering of future climate warming across Northern Europe, even in the flattest terrains.

Abstract The regional variability in tundra and boreal carbon dioxide (CO 2 ) fluxes can be high, complicating efforts to quantify sink‐source patterns across the entire region. Statistical models are increasingly used to predict (i.e., upscale) CO 2 fluxes across large spatial domains, but the reliability of different modeling techniques, each with different specifications and assumptions, has not been assessed in detail. Here, we compile eddy covariance and chamber measurements of annual and growing season CO 2 fluxes of gross primary productivity (GPP), ecosystem respiration (ER), and net ecosystem exchange (NEE) during 1990–2015 from 148 terrestrial high‐latitude (i.e., tundra and boreal) sites to analyze the spatial patterns and drivers of CO 2 fluxes and test the accuracy and uncertainty of different statistical models. CO 2 fluxes were upscaled at relatively high spatial resolution (1 km 2 ) across the high‐latitude region using five commonly used statistical models and their ensemble, that is, the median of all five models, using climatic, vegetation, and soil predictors. We found the performance of machine learning and ensemble predictions to outperform traditional regression methods. We also found the predictive performance of NEE‐focused models to be low, relative to models predicting GPP and ER. Our data compilation and ensemble predictions showed that CO 2 sink strength was larger in the boreal biome (observed and predicted average annual NEE −46 and −29 g C m −2 yr −1 , respectively) compared to tundra (average annual NEE +10 and −2 g C m −2 yr −1 ). This pattern was associated with large spatial variability, reflecting local heterogeneity in soil organic carbon stocks, climate, and vegetation productivity. The terrestrial ecosystem CO 2 budget, estimated using the annual NEE ensemble prediction, suggests the high‐latitude region was on average an annual CO 2 sink during 1990–2015, although uncertainty remains high.

Ice Age megafauna have long been known to be associated with global cooling during the Pleistocene, and their adaptations to cold environments, such as large body size, long hair, and snow-sweeping structures, are best exemplified by the woolly mammoths and woolly rhinos. These traits were assumed to have evolved as a response to the ice sheet expansion. We report a new Pliocene mammal assemblage from a high-altitude basin in the western Himalayas, including a primitive woolly rhino. These new Tibetan fossils suggest that some megaherbivores first evolved in Tibet before the beginning of the Ice Age. The cold winters in high Tibet served as a habituation ground for the megaherbivores, which became preadapted for the Ice Age, successfully expanding to the Eurasian mammoth steppe.

The aim of X-ray tomography is to reconstruct an unknown physical body from a collection of projection images. When the projection images are only available from a limited angle of view, the reconstruction problem is a severely ill-posed inverse problem. Statistical inversion allows stable solution of the limited-angle tomography problem by complementing the measurement data by a priori information. In this work, the unknown attenuation distribution inside the body is represented as a wavelet expansion, and a Besov space prior distribution together with positivity constraint is used. The wavelet expansion is thresholded before reconstruction to reduce the dimension of the computational problem. Feasibility of the method is demonstrated by numerical examples using in vitro data from mammography and dental radiology.

Abstract Aim To quantify whether species distribution models ( SDMs ) can reliably forecast species distributions under observed climate change. In particular, to test whether the predictive ability of SDMs depends on species traits or the inclusion of land cover and soil type, and whether distributional changes at expanding range margins can be predicted accurately. Location F inland Methods Using 10‐km resolution butterfly atlas data from two periods, 1992–99 ( t 1 ) and 2002–09 (t 2 ), with a significant between‐period temperature increase, we modelled the effects of climatic warming on butterfly distributions with boosted regression trees ( BRT s) and generalized additive models ( GAM s). We evaluated model performance by using the split‐sample approach with data from t 1 (‘non‐independent validation’), and then compared model projections based on data from t 1 with species' observed distributions in t 2 (‘independent validation’). We compared climate‐only SDM s to SDM s including land cover, soil type, or both. Finally, we related model performance to species traits and compared observed and predicted distributional shifts at northern range margins. Results SDM s showed fair to good model fits when modelling butterfly distributions under climate change. Model performance was lower with independent compared with non‐independent validation and improved when land cover and soil type variables were included, compared with climate‐only models. SDM s performed less well for highly mobile species and for species with long flight seasons and large ranges. When forecasting changes at northern range margins, correlations between observed and predicted range shifts were predominantly low. Main conclusions SDM s accurately describe current distributions of most species, yet their performance varies with species traits and the inclusion of land cover and soil type variables. Moreover, their ability to predict range shifts under climate change is limited, especially at the expanding edge. More tests with independent validations are needed to fully understand the predictive potential of SDMs across taxa and biomes.

Abstract Aim Although species distribution models (SDMs) traditionally link species occurrences to free‐air temperature data at coarse spatio‐temporal resolution, the distribution of organisms might instead be driven by temperatures more proximal to their habitats. Several solutions are currently available, such as downscaled or interpolated coarse‐grained free‐air temperatures, satellite‐measured land surface temperatures (LST) or in‐situ‐measured soil temperatures. A comprehensive comparison of temperature data sources and their performance in SDMs is, however, currently lacking. Location Northern Scandinavia. Time period 1970–2017. Major taxa studied Higher plants. Methods We evaluated different sources of temperature data (WorldClim, CHELSA, MODIS, E‐OBS, topoclimate and soil temperature from miniature data loggers), differing in spatial resolution (from 1″ to 0.1°), measurement focus (free‐air, ground‐surface or soil temperature) and temporal extent (year‐long versus long‐term averages), and used them to fit SDMs for 50 plant species with different growth forms in a high‐latitudinal mountain region. Results Differences between these temperature data sources originating from measurement focus and temporal extent overshadow the effects of temporal climatic differences and spatio‐temporal resolution, with elevational lapse rates ranging from −0.6°C per 100 m for long‐term free‐air temperature data to −0.2°C per 100 m for in‐situ soil temperatures. Most importantly, we found that the performance of the temperature data in SDMs depended on the growth forms of species. The use of in‐situ soil temperatures improved the explanatory power of our SDMs ( R 2 on average +16%), especially for forbs and graminoids ( R 2 +24 and +21% on average, respectively) compared with the other data sources. Main conclusions We suggest that future studies using SDMs should use the temperature dataset that best reflects the ecology of the species, rather than automatically using coarse‐grained data from WorldClim or CHELSA.

Estimation of non-discrete physical quantities from indirect linear measurements is considered. Bayesian solution of such an inverse problem involves discretizing the problem and expressing available a priori information in the form of a prior distribution in a finite-dimensional space. Since a priori information is independent of the measurement, the discretization of the unknown quantity can be arbitrarily fine regardless of the number of measurements. The main result is that Bayesian conditional mean estimates for total variation prior distribution are not edge-preserving with very fine discretizations of the model space. Theoretical findings are illustrated by a numerical example with computer simulated data.

des tablissements d'enseignement et de recherche franais ou trangers, des laboratoires publics ou privs.

Article impact statement: Machine learning can be used to automatically monitor and assess illegal wildlife trade on social media platforms.

BACKGROUND: Mean turnover times and the time spent in the operating room (OR) can be reduced by concurrent induction of anesthesia. Previous studies of anesthesia induction outside the OR have concentrated either on anesthesia-controlled time or turnover time. The goal of this study was to investigate the impact of an induction room model on the whole surgical process, its phases and delays between the phases, and the number of cases performed during the 7-h working day. METHODS: A prospective analysis of OR times was conducted for 5 weeks with the traditional induction-in-the-OR model followed by 4 weeks with a new model: A team of two nurses and one anesthesiologist was added to one OR to perform parallel anesthesia induction in a separate induction room. The durations of phases of surgical process, number of completed cases between 7:45 am and 3:00 pm, and daily raw utilization of the OR were assessed. Results were compared to those measured before the intervention. RESULTS: The mean nonoperative time was reduced by 45.6%, whereas surgery time remained unchanged. The time savings contributed to the concurrent anesthesia induction and the cut down in delays between the phases. The new model allowed one additional case to be performed during the 7-h working day. CONCLUSIONS: Anesthesia induction outside the OR can increase the number of surgical cases performed during a regular workday.

Policies and research increasingly focus on the protection of ecosystem services (ESs) through priority-area conservation. Priority areas for ESs should be identified based on ES capacity and ES demand and account for the connections between areas of ES capacity and demand (flow) resulting in areas of unique demand-supply connections (flow zones). We tested ways to account for ES demand and flow zones to identify priority areas in the European Union. We mapped the capacity and demand of a global (carbon sequestration), a regional (flood regulation), and 3 local ESs (air quality, pollination, and urban leisure). We used Zonation software to identify priority areas for ESs based on 6 tests: with and without accounting for ES demand and 4 tests that accounted for the effect of ES flow zone. There was only 37.1% overlap between the 25% of priority areas that encompassed the most ESs with and without accounting for ES demand. The level of ESs maintained in the priority areas increased from 23.2% to 57.9% after accounting for ES demand, especially for ESs with a small flow zone. Accounting for flow zone had a small effect on the location of priority areas and level of ESs maintained but resulted in fewer flow zones without ES maintained relative to ignoring flow zones. Accounting for demand and flow zones enhanced representation and distribution of ESs with local to regional flow zones without large trade-offs relative to the global ES. We found that ignoring ES demand led to the identification of priority areas in remote regions where benefits from ES capacity to society were small. Incorporating ESs in conservation planning should therefore always account for ES demand to identify an effective priority network for ESs.

Shifts in precipitation regimes are an inherent component of climate change, but in low-energy systems are often assumed to be less important than changes in temperature. Because soil moisture is the hydrological variable most proximally linked to plant performance during the growing season in arctic-alpine habitats, it may offer the most useful perspective on the influence of changes in precipitation on vegetation. Here we quantify the influence of soil moisture for multiple vegetation properties at fine spatial scales, to determine the potential importance of soil moisture under changing climatic conditions. A fine-scale data set, comprising vascular species cover and field-quantified ecologically relevant environmental parameters, was analysed to determine the influence of soil moisture relative to other key abiotic predictors. Soil moisture was strongly related to community composition, species richness and the occurrence patterns of individual species, having a similar or greater influence than soil temperature, pH and solar radiation. Soil moisture varied considerably over short distances, and this fine-scale heterogeneity may contribute to offsetting the ecological impacts of changes in precipitation for species not limited to extreme soil moisture conditions. In conclusion, soil moisture is a key driver of vegetation properties, both at the species and community level, even in this low-energy system. Soil moisture conditions represent an important mechanism through which changing climatic conditions impact vegetation, and advancing our predictive capability will therefore require a better understanding of how soil moisture mediates the effects of climate change on biota.

BACKGROUND: This multicenter study evaluated the effect of a new depth of anesthesia-monitoring device based on time-frequency-balanced spectral entropy of electroencephalogram monitoring (GE Healthcare Finland, Helsinki, Finland) on consumption of anesthetic drugs and recovery times after anesthesia. METHODS: The study was a prospective, randomized, single-blind study performed in six hospitals in Finland, Sweden, and Norway. After institutional review board approval and written informed consent from each patient, the patients were randomly allocated to anesthesia with entropy values either shown (entropy group) or not shown (control group). Anesthesia was maintained with propofol, nitrous oxide, and alfentanil. In the entropy group, propofol was given to keep the state entropy value between 45 and 65, and alfentanil was given to keep the state entropy-response entropy difference below 10 units and heart rate and blood pressure within +/-20% of the baseline values. The control group patients were anesthetized to keep heart rate and blood pressure within +/-20% of the baseline values. Statistical methods included Mann-Whitney U test and unpaired t tests. RESULTS: A total of 368 patients were studied. In the entropy group, entropy values were higher during the whole operation and especially during the last 15 min (P < 0.001). Consequently, propofol consumption was smaller in the entropy group during the whole anesthesia period (P < 0.001) and especially during the last 15 min (P < 0.001). This shortened the time delay in the early recovery parameters in the entropy group. CONCLUSION: Entropy monitoring assisted titration of propofol, especially during the last part of the procedures, as indicated by higher entropy values, decreased consumption of propofol, and shorter recovery times in the entropy group.

BACKGROUND: No validated monitoring method is available for evaluating the nociception/antinociception balance. We assessed the surgical stress index (SSI), computed from finger photoplethysmographic waveform amplitudes and pulse-to-pulse intervals, in patients undergoing shoulder surgery under general anesthesia (GA) and interscalene plexus block and in patients with GA only. METHODS: In this prospective, randomized study in 26 patients, increased blood pressure (BP) or heart rate, movement, and coughing were considered to be signs of intraoperative nociception and were treated with alfentanil. GA was maintained with desflurane aiming at a State Entropy level of 50. Photoplethysmographic waveforms were collected from the contra-lateral arm to the surgery and SSI values from 0 (no surgical stress) to 100 (maximal surgical stress) were calculated off-line. RESULTS: Two minutes after skin incision, SSI had not increased in the plexus group and was lower in the plexus group (38 +/- 13) compared with the controls (58 +/- 13, P<0.005). Among the controls, 1 min before alfentanil administration, the SSI value was higher than during periods of adequate antinociception, 59 +/- 11 vs. 39 +/- 12 (P<0.01). The total cumulative need for alfentanil was higher in controls (2.7 +/- 1.2 mg) compared with the plexus group (1.6 +/- 0.5 mg; P=0.008). Tetanic stimulation to the ulnar region of the hand increased SSI significantly only among the patients with plexus block not covering the site of the stimulation. CONCLUSION: SSI values were lower in patients with plexus block covering the sites of nociceptive stimuli. In detecting nociceptive stimuli, SSI had better performance than heart rate, BP, or response entropy.