Clinal variation for only some phenological traits across a species range.

Phenology is the timing of life cycle events of an organism. Alterations in phenology can have profound effects on individual fitness, population growth, and community dynamics. Recent changes in climate have altered the phenology of many organisms, which may result in selection to shift phenological traits. Understanding the relationship between local climates and population differentiation in phenology will allow us to anticipate responses to novel selective environments caused by global climate change. We evaluated population differentiation in the number of days to germination, first flower, and fruit maturation for 33 populations throughout the range of Campanulastrum americanum (American Bellflower). Germination and fruit maturation had geographical clines with earlier timing in populations from northern latitudes. Northern sites were cooler and drier, suggesting potential adaptive differentiation of the shorter life cycle associated with earlier phenology. Similarly, higher elevations were cooler and had earlier fruit maturation. However, seed germination was later in higher elevation populations. Although there was substantial variation in the day to first flower, ranging 40 days between population means, it was idiosyncratic and not related to latitude, suggesting differentiation in response to selective factors distinct from those on germination and fruit maturation. Thus, germination and fruit maturation in C. americanum may shift in response to selection by rising temperatures. However, such changes are not expected for flowering time, a typical indicator of climate change.

Interrelationships among life-history traits in three California oaks.

Life-history traits interact in important ways. Relatively few studies, however, have explored the relationships between life-history traits in long-lived taxa such as trees. We examined patterns of energy allocation to components of reproduction and growth in three species of California oaks (Quercus spp.) using a combination of annual acorn censuses, dendrometer bands to measure radial increment, and litterfall traps. Our results are generally consistent with the hypothesis that energy invested in reproduction detracts from the amount of energy available for growth in these long-lived taxa; i.e., there are trade-offs between these traits. The relationships between reproduction and growth varied substantially among specific trait combinations and tree species, however, and in some cases were in the direction opposite that expected based on the assumption of trade-offs between them. This latter finding appears to be a consequence of the pattern of resource use across years in these long-lived trees contrasting with the expected partitioning of resource use within years in short-lived taxa. Thus, the existence and magnitude of putative trade-offs varied depending on whether the time scale considered was within or across years. Collectively, our results indicate that negative relationships between fundamental life-history traits can be important at multiple levels of modular organization and that energy invested in reproduction can have measurable consequences in terms of the amount of energy available for future reproduction and both current and future growth.

A role for nonadaptive processes in plant genome size evolution?

Genome sizes vary widely among species, but comprehensive explanations for the emergence of this variation have not been validated. Lynch and Conery (2003) hypothesized that genome expansion is maladaptive, and that lineages with small effective population size (N(e)) evolve larger genomes than those with large N(e) as a consequence of the lowered efficacy of natural selection in small populations. In addition, mating systems likely affect genome size evolution via effects on both N(e) and the spread of transposable elements (TEs). We present a comparative analysis of the effects of N(e) and mating system on genome size evolution in seed plants. The dataset includes 205 species with monoploid genome size estimates (corrected for recent polyploidy) ranging from 2Cx = 0.3 to 65.9 pg. The raw data exhibited a strong positive relationship between outcrossing and genome size, a negative relationship between N(e) and genome size, but no detectable N(e)x outcrossing interaction. In contrast, phylogenetically independent contrast analyses found only a weak relationship between outcrossing and genome size and no relationship between N(e) and genome size. Thus, seed plants do not support the Lynch and Conery mechanism of genome size evolution. Further work is needed to disentangle contrasting effects of mating systems on the efficacy of selection and TE transmission.

Mating system and ploidy influence levels of inbreeding depression in Clarkia (Onagraceae).

Inbreeding depression is the reduction in offspring fitness associated with inbreeding and is thought to be one of the primary forces selecting against the evolution of self-fertilization. Studies suggest that most inbreeding depression is caused by the expression of recessive deleterious alleles in homozygotes whose frequency increases as a result of self-fertilization or mating among relatives. This process leads to the selective elimination of deleterious alleles such that highly selfing species may show remarkably little inbreeding depression. Genome duplication (polyploidy) has also been hypothesized to influence levels of inbreeding depression, with polyploids expected to exhibit less inbreeding depression than diploids. We studied levels of inbreeding depression in allotetraploid and diploid species of Clarkia (Onagraceae) that vary in mating system (each cytotype was represented by an outcrossing and a selfing species). The outcrossing species exhibited more inbreeding depression than the selfing species for most fitness components and for two different measures of cumulative fitness. In contrast, though inbreeding depression was generally lower for the polyploid species than for the diploid species, the difference was statistically significant only for flower number and one of the two measures of cumulative fitness. Further, we detected no significant interaction between mating system and ploidy in determining inbreeding depression. In sum, our results suggest that a taxon's current mating system is more important than ploidy in influencing levels of inbreeding depression in natural populations of these annual plants.

Polyploidy and self-fertilization in flowering plants.

Mating systems directly control the transmission of genes across generations, and understanding the diversity and distribution of mating systems is central to understanding the evolution of any group of organisms. This basic idea has been the motivation for many studies that have explored the relationships between plant mating systems and other biological and/or ecological phenomena, including a variety of floral and environmental characteristics, conspecific and pollinator densities, growth form, parity, and genetic architecture. In addition to these examples, a potentially important but poorly understood association is the relationship between plant mating systems and genome duplication, i.e., polyploidy. It is widely held that polyploid plants self-fertilize more than their diploid relatives, yet a formal analysis of this pattern does not exist. Data from 235 species of flowering plants were used to analyze the association between self-fertilization and ploidy. Phylogenetically independent contrasts and cross-species analyses both lend support to the hypothesis that polyploids self-fertilize more than diploids. Because polyploidy and self-fertilization are so common among angiosperms, these results contribute not only to our understanding of the relationship between mating systems and polyploidy in particular, but more generally, to our understanding of the evolution of flowering plants.