Search

The task of modeling the distribution of a large number of tree species under future climate scenarios presents unique challenges. First, the model must be robust enough to handle climate data outside the current range without producing unacceptable instability in the output. In addition, the technique should have automatic search mechanisms built in to select the most appropriate values for input model parameters for each species so that minimal effort is required when these parameters are fine-tuned for individual tree species. We evaluated four statistical models—Regression Tree Analysis (RTA), Bagging Trees (BT), Random Forests (RF), and Multivariate Adaptive Regression Splines (MARS)—for predictive vegetation mapping under current and future climate scenarios according to the Canadian Climate Centre global circulation model. To test, we applied these techniques to four tree species common in the eastern United States: loblolly pine ( Pinus taeda), sugar maple ( Acer saccharum), American beech ( Fagus grandifolia), and white oak ( Quercus alba). When the four techniques were assessed with Kappa and fuzzy Kappa statistics, RF and BT were superior in reproducing current importance value (a measure of basal area in addition to abundance) distributions for the four tree species, as derived from approximately 100,000 USDA Forest Service’s Forest Inventory and Analysis plots. Future estimates of suitable habitat after climate change were visually more reasonable with BT and RF, with slightly better performance by RF as assessed by Kappa statistics, correlation estimates, and spatial distribution of importance values. Although RTA did not perform as well as BT and RF, it provided interpretive models for species whose distributions were captured well by our current set of predictors. MARS was adequate for predicting current distributions but unacceptable for future climate. We consider RTA, BT, and RF modeling approaches, especially when used together to take advantage of their individual strengths, to be robust for predictive mapping and recommend their inclusion in the ecological toolbox. [ABSTRACT FROM AUTHOR]

In dairy calves raised for veal, typical clinical signs of bovine respiratory disease (BRD) are ocular discharge, nasal discharge, ear droop or head tilt, abnormal respiration, cough, and increased rectal temperature. Despite the existence of several clinical scoring systems, there are few studies on the variability of human recognition of individual BRD clinical signs. The objective of this study was therefore to assess the inter-rater agreement of BRD clinical signs in veal calves. We hypothesized that BRD clinical signs were not detected equally between veterinarians, technicians, and producers of the veal industry and that some clinical signs have higher inter-rater agreement than others. During 2017-2018, we prospectively recorded 524 videos of physical examinations of random veal calves from 48 different batches in Québec, Canada. A researcher, not involved in the inter-rater assessment, classified each video as presence/absence of each BRD clinical sign except rectal temperature. For each of the 5 clinical signs, 15 videos with and 15 videos without the clinical signs were randomly selected to avoid kappa paradoxes. Those 30 videos were then presented in a random order to experienced raters of BRD in veal calves: 6 veterinarians, 6 technicians, and 6 producers. The raters assessed the clinical signs using scores based on the Wisconsin and California scoring system with modifications (0 = absent, 1 = mild, 2 = moderate, 3 = severe for nasal discharge, ocular discharge, and ear droop or head tilt; and 0 = absent, 1 = moderate, 2 = severe for abnormal respiration and induced cough). We used median percentage agreement (Pa), median Cohen's kappa (κ), and Gwet's agreement coefficient 1 (AC1) to assess inter-rater agreement. The effect of scale combination was also tested to determine the optimal combination (4-scale 0/1/2/3 vs. 3-scale 0/1/2 vs. 2-scale 0/1,2,3; 0,1/2,3; or 0/1,2). The differences of inter-rater agreement between veterinarians, technicians, and producers were estimated by a Wilcoxon rank-sum test. The 2-scale combination (0,1/2,3 or 0/1,2) had the highest inter-rater agreement for all clinical signs. With this combination, induced cough was the clinical sign with the highest inter-rater agreement (Pa = 0.93; κ = 0.79; AC1 = 0.87) and abnormal respiration was the sign with the lowest inter-rater agreement (Pa = 0.77; κ = 0.20; AC1 = 0.74). According to Pa and AC1 values, the 2-scale inter-rater agreement of the 5 clinical signs was good (value > 0.6). According to κ, only ear droop or head tilt and induced cough had a substantial 2-scale inter-rater agreement (κ > 0.6). In general, the 2-scale inter-rater agreement was better among veterinarians than among technicians and producers, except for the ear droop/head tilt, where agreement was better among producers. We concluded that with severity scores assessed on a scale of 2 (0,1/2,3 or 0/1,2), the inter-rater agreement of BRD clinical signs was variable according to the sign in veal calves. BRD clinical signs were not detected equally between veterinarians, technicians, and producers of the veal industry. Future research could determine if this discrepancy could be improved by standardization training., (© 2021, The Authors. Published by Elsevier Inc. and Fass Inc. on behalf of the American Dairy Science Association®. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).)

It is very common to examine reliability of triage scales using (weighted) kappa statistics. The point is that weighted kappa has grossly underestimated disagreements by one category and put more emphasis on extreme category disagreements; therefore, low prevalence of critically-ill and non-urgent patients has excluded the effect of extreme categories disagreement from calculated kappa coefficient and also contributed to significant overestimation. As a result, weighted kappa coefficient as an estimate of scale reliability is overestimated by the anchoring effect. [ABSTRACT FROM AUTHOR]