Does Snowmelt Date Affect the Flowering Time of Delphinium Nuttallianum

Price and Waser's (1998) observations on the global warming experiment at the RMBL (Harte et al. 1995, Harte and Shaw 1995) suggested that if winter precipitation patterns (in the form of snow)

don't change, flowering date will shift with global warming by shifting the snowmelt date. If precipitation changes, these predictions could fail. Current models of climate change, however, have no clear predictions about whether precipitation will increase or decrease at a regional level. The most likely scenario is that more snow will fall in the winter, snow will melt earlier, and flowering time will be earlier (Harte, pers. comm. 2000). For example recent work in the Sierra Nevada, California, shows that snow accumulation from 1937 to 1997 in sites above 2,400 m elevation is higher yet still melts earlier (Johnson 1998). At the RMBL, there has been a trend toward higher winter snow accumulation since 1973, and no change in date of snowmelt, although more years are needed to determine if this trend is significant (Inouye et al. 2000). Experimental warming led to both an earlier snowmelt date (Harte et al. 1995, Price and Waser 1998) and corresponding changes in the mean date of flowering (Price and Waser 1998). There was a significant and positive relationship between the mean timing of plant reproduction and snowmelt in eight out of ten species measured: Claytonia lance-olata, Erythronium grandiflorum, Mertensia brevistyla (early-flowering species); Delphinium nuttallianum, Lathyrus leucanthus, Potentilla pul-cherrima (mid-flowering species); Eriogonum subalpinum, and Campanula rotundifolia (late-flowering species). There was no significant positive relationship between the mean time of reproduction and snowmelt date for two of the late flowering species: Ipomopsis aggre-gata (no relationship) and Seriphidium vaseyanum (significant negative relationship). Four species, among them D. nuttallianum, flowered significantly earlier in the heated plots. Early flowering species tended to respond more strongly to experimental warming, suggesting that snowmelt is the primary cue for timing of reproduction in these species. In contrast, species response was less clear when durations of flowering and fruiting and fruit set between treatments were compared (Price and Waser 1998). However, six species showed significant decreases in duration of flowering with later snowmelt, and eight out of nine species showed higher frequencies of fruit maturation (not individually significant, but significant by a binomial test) in the heated plots.

The evidence that snowmelt triggers an early flowering time was only indirect, since the warming experiment changed not only snowmelt time but many other parameters (e.g., soil moisture, soil temperature). A snow removal experiment allows a more direct test of the effect of snowmelt time on D. nuttallianum flowering time. Removing snow experimentally should affect the phenology of D. nuttallianum. When snowcover is decreased in this way, snowmelt date should occur earlier and reproduction of D. nuttallianum should start earlier in the season. Snowmelt date is associated with other environmental factors, including soil temperature, water availability, nutrient availability, exposure to cold, and length of the growing season. Indeed, manipulating snow levels to speed snowmelt increased both spring soil temperature and spring soil moisture.

0 0

Post a comment