The Evolution of Self‐pollination in Granite Outcrop Species of Arenaria (Caryophyllaceae). I. Morphological Correlates

Darwin (1876, 1877) demonstrated ly indistinguishable from its outcrossing that outcrossed progeny ofplants are usu-relative, E. sinclairii (Baker, 1967).Aually more vigorous than those produced togamous populations of Lycopersicon by self-fertilization.This observation led esculentum and L. pimpinellifolium difhim to interpret many features of flow-fer from their outcrossing ancestors by ering plants as adaptations for outcross-very slight rearrangements in structure ing.Nevertheless, it is now well-known and position ofanthers and stigmas (Rick, that many plant species have evolved 1950; Rick et al., 1978).In other taxa, mechanisms that promote selfing, Steb-major morphological reorganization may bins (1970) observed that "the evolu-occur in association with a shift in the tionary pathway from obligate outcross-breeding system.Ornduff (1969) has listing based upon self-incompatibility to ed a number of features characterizing a predominant self-fertilization has prob-general trend from predominant outably been followed by more different lines crossing to predominant selfing, Obvious of evolution in flowering plants than has adaptations of outcrossing plants for polany other."There is indeed a long list of lination by animal vectors frequently are such taxa (see reviews by East, 1940; reduced or lost entirely in their autoga- Fryxell, 1957;Stebbins, 1970; Wyatt, mous derivatives.A particularly dra-1983).matic example is Rollins's (1963) and In other taxa, self-compatible but pre-Lloyd's (1965) studies of Leavenworthia.dominantly outcrossing plants may Selfing has arisen many times in indechange such that self-pollination is as-pendent evolutionary lines to produce a sured.This usually involves a breakdown number of new autogamous species.The in dichogamy or other devices that pro-morphological changes accompanying the mote outcrossing.In Clarkia xantiana.evolution of autogamy in each case are for example, stigma maturation regulates strikingly similar.outcrossing rates and "appears to have aThe state of the art with respect to the simple genetic basis" (Moore and Lewis, evolution of autogamy was succinctly 1965).Determination ofoutcrossing rates summarized by Jain (1976), who stated in Lycopersicon also is largely through that "we have more speculation than rigmorphological characters with a simple orous thought, and more theory than real genetic basis (Rick, 1950), although poly-data" regarding the questions of when, genic control of certain features is pos-how, and why it evolved.A general probtulated for some species (Rick et al., lem with most studies of the evolution 1978).In some taxa, there appears to be ofautogamy is that comparisons are made a large environmental component to between "closely related" congeneric breeding system variation (Jain, 1976).species that differ with respect to their In some cases only minor morpholog-rates of outcrossing.This approach imical changes are associated with a shift plicitly assumes that all morphological, from predominant outcrossing to pre-genetic, and ecological differences can be dominant selfing (Stebbins, 1970).The attributed directly to differences in outselfing Eupatorium microstemon is near-crossing rates.