We tested the effects of phylogeny, life history and habitat conditions for seedling establishment on seed germination of 69 arid/semi-arid zone species from northwest China. Final germination percentages had a bimodal distribution, whereas, days to first germination were skewed toward short periods. Stepwise multi-way ANOVAs showed that the amount of variance in final germination percentage among species for seeds incubated in light was explained by phylogeny (13.6%), dispersal mode (11.9%), seed mass (3.6%), and habitat (0.2%); in darkness by dispersal mode (17.5%), phylogeny (12.7%), seed mass (5.7%), and habitat (0%). A series of ANOVAs revealed that the variance in final germination percentages among species is largely dependent upon phylogeny and dispersal mode but that it is also influenced by seed mass and habitat. The effects of dispersal mode and seed mass on final germination percentage among species were phylogenetic group-and habitat-specific. Wind-dispersed seeds had higher germination percentages than unassisted and vertebrate-dispersed seeds. Compared to xerophytes, mesophytes tended to have smaller seed mass, higher mean germination percentages and a greater effect of seed mass on final germination fractions. Our results suggest that phylogeny and life history should be taken into account in studies on the role of natural selection in regulation of seed germination.

Effects of cold-dry storage on dormancy break and viability were determined for seeds of the five subalpine woody species Philadelphus incanus, Berberis vernae, Berberis dubia, Betula utilis, and Picea purpurea collected along an altitudinal gradient on the eastern part of the Qinghai-Tibet Plateau in China. Germination tests were conducted at 20/5°C for seeds stored dry at ambient room temperature for 4 weeks and then at 3-4°C for 0, 6, 12, and 24 weeks. Dormancy break during dry storage, i.e., afterripening, was indicated by an increase in germination percentages and rates. Duration of cold-dry storage and altitude of seed collection had significant effects on germination. With an increase in duration of storage, germination percentages and rates of P. incanus and B. vernae increased with a decrease in altitude of seed collection, while they increased with an increase in altitude for seeds of B. utilis and P. purpurea. Seeds of B. dubia did not exhibit changes in germination percentages and rates with altitude because a high number of seeds remained dormant during storage. Seed viability after 24 weeks of storage ranged from high (88, 93.3, 92.7%) for B. utilis to low (15% for high altitude) for P. incanus. The potential for dormancy break to occur during cold-dry storage should be considered when studies on basic seed dormancy are conducted or when seeds from various locations are stored prior to propagating plants from them.