Ohsowski Brian M, Redding Cassidy, Geddes Pamela, Lishawa Shane C
School of Environmental Sustainability, Loyola University Chicago, Chicago, IL, United States.
Department of Biology and Environmental Science Program, Northeastern Illinois University, Chicago, IL, United States.
Front Plant Sci. 2024 Mar 12;15:1348144. doi: 10.3389/fpls.2024.1348144. eCollection 2024.
Two species of clonal [ (native) and (exotic)] hybridize to form the highly invasive, heterotic (high vigor) in North American wetlands leading to increased primary production, litter accumulation, and biodiversity loss. Conservation of has become critical as invasive has overwhelmed wetlands. In the field, taxa identification is difficult due to subtle differences in morphology, and molecular identification is often unfeasible for managers. Furthermore, improved methods to non-destructively estimate biomass is imperative to enhance ecological impact assessments. To address field-based ID limitations, our study developed a predictive model from 14 characters in 7 northern Michigan wetlands to accurately distinguish taxa (n = 33) via linear discriminant analysis (LDA) of molecularly identified specimens. In addition, our study developed a partial least squares regression (PLS) model to predict biomass from field collected measurements (n = 75). Results indicate that two field measurements [, ] can accurately differentiate the three taxa and advanced-generation hybrids. The LDA model had a 100% correct prediction rate of . The selected PLS biomass prediction model () improved upon existing simple linear regression (SLR) height-to-biomass predictions. The rapid field-based identification and biomass assessment tools presented in this study advance targeted management for regional conservation of and ecological restoration of wetlands impacted by invasive taxa.
两种克隆物种（本地种和外来种）杂交形成了北美湿地中极具入侵性、杂种优势明显（活力高）的[物种名称]，导致初级生产力增加、凋落物积累以及生物多样性丧失。随着入侵性[物种名称]泛滥成灾，对其进行保护变得至关重要。在野外，由于形态上的细微差异，[物种名称]的分类鉴定很困难，而分子鉴定对管理人员来说往往不可行。此外，改进非破坏性估计[物种名称]生物量的方法对于加强生态影响评估至关重要。为了解决基于野外的[物种名称]鉴定限制，我们的研究从密歇根州北部7个湿地的14个[物种名称]特征建立了一个预测模型，通过对分子鉴定标本的线性判别分析（LDA）准确区分[物种名称]分类群（n = 33）。此外，我们的研究建立了一个偏最小二乘回归（PLS）模型，根据野外采集的测量数据（n = 75）预测[物种名称]生物量。结果表明，两项野外测量[具体测量项目]可以准确区分三种[物种名称]分类群和高级杂交种。LDA模型对[具体分类群]的预测准确率为100%。所选的PLS生物量预测模型（[模型名称]）比现有的简单线性回归（SLR）高度与生物量预测有所改进。本研究中提出的基于野外的快速[物种名称]鉴定和生物量评估工具推进了针对[物种名称]区域保护和受入侵[物种名称]分类群影响的湿地生态恢复的目标管理。
