Copp Gordon H, Vilizzi Lorenzo, Mumford John, Fenwick Gemma V, Godard Michael J, Gozlan Rodolphe E
Salmon and Freshwater Fisheries Team, Cefas, Lowestoft, NR33 0HT, UK.
Risk Anal. 2009 Mar;29(3):457-67. doi: 10.1111/j.1539-6924.2008.01159.x. Epub 2008 Dec 12.
Adapted from the weed risk assessment (WRA) of Pheloung, Williams, and Halloy, the fish invasiveness scoring kit (FISK) was proposed as a screening tool for freshwater fishes. This article describes improvements to FISK, in particular the incorporation of confidence (certainty/uncertainty) ranking of the assessors' responses, and reports on the calibration of the score system, specifically: determination of most appropriate score thresholds for classifying nonnative species into low-, medium-, and high-risk categories, assessment of the patterns of assessors' confidences in their responses in the FISK assessments. Using receiver operating characteristic (ROC) curves, FISK was demonstrated to distinguish accurately (and with statistical confidence) between potentially invasive and noninvasive species of nonnative fishes, with the statistically appropriate threshold score for high-risk species scores being >/=19. Within the group of species classed as high risk using this new threshold, a "higher risk" category could be visually identified, at present consisting of two species (topmouth gudgeon Pseudorasbora parva and gibel carp Carassius gibelio). FISK represents a useful and viable tool to aid decision- and policymakers in assessing and classifying freshwater fishes according to their potential invasiveness.
鱼类入侵性评分工具(FISK)是在借鉴费隆、威廉姆斯和哈洛伊的杂草风险评估(WRA)基础上提出的,作为一种淡水鱼类筛选工具。本文介绍了FISK的改进之处,特别是纳入了评估者回答的置信度(确定性/不确定性)排名,并报告了评分系统的校准情况,具体包括:确定将外来物种划分为低、中、高风险类别的最合适分数阈值,评估评估者在FISK评估中对其回答的置信模式。通过使用受试者工作特征(ROC)曲线,证明FISK能够准确(且具有统计学置信度)区分外来鱼类的潜在入侵物种和非入侵物种,高风险物种得分的统计学合适阈值分数为≥19。在使用这个新阈值归类为高风险的物种组中,可以直观地识别出一个“更高风险”类别,目前该类别由两个物种组成(麦穗鱼Pseudorasbora parva和银鲫Carassius gibelio)。FISK是一个有用且可行的工具,可帮助决策者和政策制定者根据淡水鱼类的潜在入侵性对其进行评估和分类。