Factor Analysis of Fish Distribution Patterns in Western and Central Oklahoma
Factor Analysis of Fish Distribution Patterns in Western and Central Oklahoma
复制标题
俄克拉荷马州西部和中部鱼类分布模式的因素分析
DOI：
10.1093/sysbio/23.2.202
复制
发表时间：
1974
影响因子：
6.5
通讯作者：
R. Black
中科院分区：
文献类型：
--
作者：
M. M. Stevenson;G. Schnell;R. Black
史蒂文森，M.M.，G.D.施奈尔和R.布莱克(动物学和斯托瓦尔科学与历史博物馆，教育学院，教育特别项目，俄克拉荷马大学，诺曼，俄克拉荷马73069)。俄克拉荷马州西部和中部鱼类分布格局的因子分析。系统祖尔。23：202-218-利用因子分析技术分析俄克拉荷马州西部和中部鱼类的分布格局，以确定广义格局及其与某些环境变量的相关性。阿肯色州和红河的主要支流的27个排水单元根据53个物种的存在与否和13个环境变量的值进行了评分。从变量的相关矩阵中提取的主轴倾斜旋转，得到6个因素组。这六个协会被认为具有生物学意义，并提供了对影响俄克拉荷马州鱼类动物群的限制因素的洞察。讨论了历史和环境因素对当前分布群的相对影响，并对未来的区系变化进行了预测。[动物地理学；鱼类分布；俄克拉荷马州；因子分析。]传统上，动物地理学研究往往涉及到基于预设的解释性假设对“区域”进行分类或物种分布模式分组。虽然这种分析的结论可能是正确的，但很难有效地评价结果，因为所提出的推理往往似乎包含相当程度的循环。然而，在Fisher(1968)和Smith and Fisher(1970)中，这种创建(‘最合理分组模式’的先验方法被抛弃了。例如，Smith和Fisher(1970)利用因子分析，客观地将堪萨斯州鱼类种类的分布数据(和环境数据)整理成几个概括性的模式，然后才尝试对影响每个群体的共同偶然因素进行解释。我们使用了类似的方法来分析和解释俄克拉荷马州西部和中部(例如，970子午线以西)鱼类分布的信息。这样的方法应该能让我们深入了解限制因素，无论是单独的还是现在的1个地址：杜兰大学，系统和环境生物学实验室，环境科学中心，1号路线，46-B信箱，路易斯安那州贝尔查斯，70037。这些因素结合在一起，影响了过去的鱼类分布，并至少在一定程度上预测了未来环境变化可能产生的影响。之所以选择这个相当有分界的区域，主要是因为河流类型总体上很简单，而且可用的集合相对完整。我们觉得，在相对简单的环境中，模式更有可能是可识别和/或可解释的。主要系统通常是二级和三级流(根据Horton，1945)，具有宽阔的泛滥平原、低起伏、砂质海底和相对较高的溶解固体含量。然而，该区域也包含了这一概括的几个主要例外地区(例如威奇托山)。通过我们的分析，发现了物种分布的某些广义模式，回顾起来，这些模式是建议的，而且往往是根据所包括的一些物种的已知生态需求而预测的。作为一个工作假设，我们假设一个或多个测试的环境变量将与每一种可识别的分布模式相关并有助于解释每一种可识别的分布模式。联席-
Stevenson, M. M., G. D. Schnell, and R. Black (Department of Zoology and Stovall Museum of Science and History, and College of Education, Education Special Projects [Black], University of Oklahoma, Norman, Oklahoma 73069) 1974. Factor analysis of fish distribution patterns in western and central Oklahoma. Syst. Zool. 23:202-218.-The distribution patterns of western and central Oklahoma fishes were analyzed, utilizing factor analytical techniques, to determine generalized patterns and their correlation with certain environmental variables. Twenty-seven drainage units of the major Arkansas and Red river tributaries were scored for the presence or absence of each of 53 species and values for 13 environmental variables. Six factor groups were obtained through oblique rotation of principal axes that were factored from the correlation matrix of the variables. These six associations were judged biologically meaningful and provide insight into the limiting factors affecting the Oklahoma fish fauna. The relative effects of historical and environmental components on current distribution groupings are discussed and some predictions are made conceming future faunal changes. [Zoogeography; fish distribution; Oklahoma; factor analysis.] Traditionally, zoogeographic studies often have involved the classification of "regions" or grouping of species distribution patterns on the basis of a presupposed explanatory hypothesis. While the conclusions of such an analysis may be correct, it is difficult to effectively evaluate results since the reasoning presented often appears to contain a considerable degree of circularity. However, in Fisher (1968) and Smith and Fisher (1970), this a priori method of creating the ('most plausible grouping patterns" was abandoned. For example, Smith and Fisher (1970) objectively arranged distribution data on Kansas fish species (and environmental data) into a few generalized patterns using factor analysis, and only then were interpretations attempted of the common casual factors affecting each group. We have utilized a similar method to analyze and interpret information on fish distributions in western and central Oklahoma (e.g., west of the 970 meridian). Such an approach should give some insight into the limiting factors which, singularly or in 1 Present address: Tulane University, Systematic and Environmental Biology Laboratory, Environmental Science Center, Route 1, Box 46-B, Belle Chasse, Louisiana 70037. combination, have affected fish distributions in the past, and at least some degree of predictive ability concerning the possible effects of environmental modifications in the future. This rather delimited region was chosen mainly because of the general simplicity of stream type and the relatively complete collections available. We felt it more likely that patterns might be discernable and/or explainable in a relatively uncomplicated environment. The major systems are most often second and third order streams (in the sense of Horton, 1945), with wide flood plains, low relief, sandy bottoms, and a relatively high content of dissolved solids. However, the region also contains a few areas that are major exceptions to this generalization (e.g., the Wichita Mountains). With our analysis, certain generalized patterns of species distributions were discovered which in retrospect are suggested and often predictable on the basis of known ecological requirements of some of the included species. As a working hypothesis, we assumed that one or more of the tested environmental variables would be correlated with and help explain each of the discernable distributional patterns. The co-