Multivariate Analysis of Selection in Experimental Populations Derived from Hybridization of Two Ecotypes of the Annual Plant Diodia teres W. (Rubiaceae)

Abstract

Morphologically variable F 2 genotypes derived from hybridization ofcoastal and inland ecotypes of the annual plant Diodia teres were used to identify selection on morphological traits in the natural habitat of each ecotype. These ecotypes occur in very different habitats, and have evolved pronounced morphological differentiation. Selection analysis can suggest whether present patterns of selection can explain morphological differences between ecotypes. F 2 genotypes were characterized morphologically, donally replicated, and transplanted into the habitat of each ecotype. Selection was measured on six morphological traits. Directional and stabilizing selection occurred on many traits; direction and strength of selection varied sharply at different stages of growth, as revealed by a path-analysis approach that divided selection into a set of independent components. Directional selection favored traits of the native population at the coastal habitat, but less so at the inland habitat. Selection was of sufficient strength to create the observed morphological differences between ecotypes in 25-100 generations, given constant selection and sufficient genetic variation. In effects on fitness, most traits were neither independent nor consistently interactive with other traits. Rather, many traits entered into strong but evanescent interactions affecting particular components of fitness. Observed interactions did not support the hypothesis that the morphology of each ecotype was functionally integrated to a high degree.

Inheritance and Natural Selection on Functional Traits

“…Selection coefficients also differed in virtually all comparisons among habitats or treatments for at least some traits (but Winn 1999). Most interestingly, in virtually all ecotypic studies, selection at a site generally favored trait values of native populations, but again for only a subset of the measured traits (Scheiner 1989;Jordan 1991;Bennington andMcGraw 1995, 1996;Dudley 1996a;Dudley and Schmitt 1996;Nagy 1997;Donohue et al 2000). In other words, current patterns of selection might account for some but not all of the phenotypic differentiation among populations.…”

“…A better understanding of cause may come from studies that manipulate the environmental factors that are thought to engender selection on traits (Wade and Kalisz 1990). Likewise, studies that manipulate trait variation, especially those that manipulate single traits independently of variation in correlated characters or disrupt character correlations among traits, may be better suited for identifying direct selection of individual traits on fitness (Wade and Kalisz 1990;Jordan 1991;Andersson and Shaw 1994;Nagy 1997;Arntz et al 2000aArntz et al , 2000b. Last, the full form of the relationship between trait and fitness may be more accurately described in studies that extend the range of phenotypic variation beyond the natural range (Schluter 1988;Wade and Kalisz 1990;Jordan 1991;Bennington andMcGraw 1995, 1996;Dudley 1996a;Dudley and Schmitt 1996;Nagy 1997;Roy et al 1999;Donohue et al 2000;Dorn et al 2000;Juenger and Bergelson 2000;Stanton et al 2000;Arntz and Delph 2001).…”

“…The most general question addressed by these studies was whether selection differs among the habitats occupied by different ecotypes (Biere 1991;van Tienderen and van der Toorn 1991;Andersson and Shaw 1994). More focused studies asked whether the pattern of selection in each habitat favors the trait values of the native ecotype and therefore might account for ecotypic differentiation in these traits (Scheiner 1989;Jordan 1991;Bennington andMcGraw 1995, 1996;Nagy 1997). The most directed studies began with hypotheses concerning the adaptive significance of specific traits in relation to specific habitat differences (Dudley and Schmitt 1996;Donohue et al 2000).…”

“…If this is true, selection in plants should lead to the evolution of coordinated suites of traits, and adaptive differentiation among populations may often consist of divergence in suites of traits. Studies focused on microevolution of interacting sets of functional traits will be required to understand their functional interdependence, to partition direct and indirect pathways of influence among these traits, and to quantify the strength and pattern of selection on suites of traits (i.e., correlational selection) (Lechowicz and Blais 1988;Farris and Lechowicz 1990;Jordan 1991;Bennington and McGraw 1995;Dudley 1996a;Kelly andLevin 1997: Ackerly et al 2000;Arntz et al 2000b;Donohue et al 2000;Juenger and Bergelson 2000;Scheiner et al 2000;Arntz and Delph 2001).…”

Contribution of photosynthetic rate to growth and reproduction in Amaranthus hybridus

“…However, when photosynthesis is considered together with allocation patterns, growth can be more accurately predicted (Dijkstra and Lambers 1989;van den Boogaard et al 1996). A similar increase in explanatory power is achieved by integrating the effects of phenology and morphology to explain variation in reproductive success (Mitchell-Olds and Bergelson 1990;Jordan 1991; van Tienderen and van der Toorn 1991; Kelly 1992; Bennington and McGraw 1995). Presumably the incorporation of physiological contributions would further increase this explanatory power (Lechowicz and Blais 1988;Farris and Lechowicz 1990;Dudley 1996); however, the direct and indirect eects of photosynthetic rate on variation in growth and reproduction remain poorly understood.…”