To better understand the colonization of bare riparian sites, I examined seedling distribution patterns and seed dispersal processes of 11 common riparian species. The study site was two in-channel sedimentation basins on an intermittent stream flowing into the Tijuana River Valley, southern California. The 11 species relied on one of two vectors to disperse seeds into the basins, either water or wind. Seven species dispersed to the basins via winter stream flows and, of these, three grew mainly at the basin periphery and four grew mainly in the bed. Seed buoyancy tests and tests of dispersal using seed mimics showed that, for these species, the main process producing seedling distributions was that of sorting of the seeds by the stream flows based on seed buoyancy. Most past work on seed dispersal in riparian habitats has been conducted in perennial systems where non-buoyant seeds usually remain in deep water and never germinate, and this has led to the notion that non-buoyant seeds are not effectively dispersed by water. This study shows that non-buoyant seeds are effectively dispersed by flows in intermittent systems because the seeds were able to germinate and grow once the bed had dried. Four species dispersed to the basins via wind, and their seedlings grew in concentric bands around the basins. Monitoring of seed production and tracking of the basin's water level showed that, for these species, seedling distributions were the result of staggered periods of seed dispersal and gradual decline of the water level. Only three of these 11 species were abundant in the five- and 22-year old stands examined, i.e., the wind-dispersed Baccharis salicifolia (Ruiz & Pav.) Pers., Salix lasiolepis Benth., and S. gooddingii C.R. Ball. A banding pattern established at the time of recruitment, with S. lasiolepis higher on the bank than S. gooddingii, was retained in the older forests, making this one of very few examples that show a direct link between seed dispersal processes, seedling distributions, and adult distributions. This study provides insight into the early development of riparian woodlands by identifying and explaining distribution patterns in water- and wind-dispersed colonizers.

Baker cypress (Hesperocyparis bakeri [Jeps.] Bartel) is one of 10 species of cypress found in western North America. It is restricted to a small number of highly disjunct, isolated populations, making it particularly vulnerable to the influences of genetic drift, inbreeding, and reduced gene flow. Baker cypress is fire adapted and the serotinous cones require heat to open and release seeds. Altered fire regimes have negatively impacted the health and vigor of some populations and lower levels of genetic diversity could make this species more susceptible to the impacts of predicted future climate change. Previously, no information on genetic diversity and population structure of Baker cypress was available. We used 12 polymorphic allozyme loci to assess genetic diversity and population structure for eight of the 11 known populations of Baker cypress. Overall mean observed heterozygosity (H_o) was 0.178 and expected heterozygosity (H_e) was 0.204, values higher than for other cypress species and other fire adapted conifers. Although genetic diversity was relatively high, many populations had a deficiency of heterozygotes (fixation index > 0), most likely due to inbreeding and possibly a Wahlund effect. Population differentiation among seven of the eight populations (northernmost population excluded) was 9%, considerably lower than for other conifers with disjunct populations. Our results indicate that the current population structure of the species is likely a fairly recent reduction from a formerly widespread distribution with differentiation among populations resulting from genetic drift. Implications of genetic diversity and population structure for potential restoration work are discussed.

Proposed causes of the latitudinal-propagule size gradient invoke differences among biome structures or seed dispersal syndromes. Latitudinal seed-size gradients so far have been predominantly investigated using entire floras, prompting the question of whether such trends exist at smaller scales. Here we consider effects of latitude and elevation on fruit size between and within species in Arctostaphylos L. (Ericaceae), a zoochorous, primarily Californian chaparral shrub genus. We measured fruit size, strongly correlated with seed size in Arctostaphylos, in three species of this genus from multiple localities in the Sierra Nevada foothills, and conducted a character analysis of fruit size across the genus using the standard California flora plant manual (Baldwin et al. 2012). Across the genus we found a weak negative correlation between fruit size and latitude (0.0026 mm diameter/km) and a weak positive correlation between fruit size and elevation (1.3 mm diameter/km). AIC indicates that these trends are not explained by autocorrelation between fruit size and other variables such as maximum plant size. By contrast, intraspecific field data revealed a positive relationship between fruit size and both elevation and latitude. Propagule size gradients within Arctostaphylos oppose those reported for angiosperms globally. This contrast may result from uniformity of fruit structure and animal dispersal, disturbance ecology of chaparral, or local precipitation gradients characteristic of Mediterranean-type climates. Studies of propagule size gradients within taxa can uncover ecological mechanisms behind this trend that remain obscure at global scales.