Kelly John K
Department of Biology, University of Oregon, Eugene, Oregon, 97403.
Evolution. 1999 Apr;53(2):350-357. doi: 10.1111/j.1558-5646.1999.tb03771.x.
The structured linear model (SLM) is generalized to treat selection on multiple, correlated characters. Four different causes of phenotypic correlations are distinguished by the SLM: environmental covariance, identity disequilibrium, pleiotropy, and linkage disequilibrium. Each is characterized by distinct variables because they have different implications for character evolution. Correlations due to identity disequilibrium and linkage disequilibrium depend on both the mating system and the selection regime. As a consequence, they will evolve rapidly under selection. Correlations due to pleiotropy or environmental factors will evolve more slowly and are characterized by parameters that can be estimated from comparisons among relatives. These parameters include several novel "inbreeding covariance components" that emerge from the interaction of inbreeding and genetic dominance. Although data are limited, current estimates suggest that the expression of these components may substantially alter the pattern of multitrait evolution in self-fertilizing populations.
结构化线性模型（SLM）被推广用于处理对多个相关性状的选择。SLM区分了表型相关性的四种不同原因：环境协方差、同型不平衡、多效性和连锁不平衡。每种原因都由不同的变量表征，因为它们对性状进化有不同的影响。由于同型不平衡和连锁不平衡导致的相关性取决于交配系统和选择机制。因此，它们在选择作用下会迅速进化。由多效性或环境因素导致的相关性进化较慢，其特征是可以通过亲属间比较来估计的参数。这些参数包括几个新的“近亲繁殖协方差分量”，它们源自近亲繁殖和遗传显性的相互作用。尽管数据有限，但目前的估计表明，这些分量的表达可能会显著改变自花受精种群中多性状进化的模式。
