How Does Timing of Flowering Affect Reproduction in Natural Populations

Shifting the date of snowmelt experimentally with a warming experiment is one approach for investigating the possible effect of climate change. In order to circumvent the limitations of the experimental approach and to have a better understanding of the fitness consequences of flowering at different dates, I complemented the experimental observations in 1997 with observations of two populations that naturally flower at different times.

Flowering time has been a topic of interest for many years. Questions related to phenology are diverse, ranging from descriptions of the temporal pattern of flower presentation to attempts to understand the selective forces that determine the shape of the flowering curves (Thomson 1980) and their timing (e.g., Waser 1978, 1979, O'Neil 1997 at the population level; Poole and Rathcke 1979, Rathcke 1988 at the community level).Within a region, not all populations flower at the same time (Bliss 1956, Billings and Bliss 1959, Jackson 1966, Galen and Stanton 1991). Campbell (1987) showed that, among six populations of Veronica cusickii in the Olympic Mountains, Washington, that differ in the duration of snow cover and concordantly in flowering time, only three were pollen limited. Campbell and Halama's (1993) experimental work showed that seed production in Ipomopsis aggregata was a complex result of the interaction of pollen and resource levels; in addition, their results showed that the effect of pollen and resources on seed production changes through the season. An increase of resources lengthens flower production, producing significantly more flowers only in the late season, whereas pollen addition increased significantly seed set per flower only in last first 2 weeks.

The flowering phenology of D. nuttallianum has already been studied in the meadows of the RMBL, and those studies have suggested that competition for pollination is a selective force acting to maintain divergent flowering times of D. nuttallianum and sympatric Ipomopsis aggregata (Waser 1978). My preliminary work in 1997 showed that the pollinator guild changes over the season. The early population flowered when no I. aggregatta were in flower and few other D. nuttallianum populations were in flower. Bumblebees and hummingbirds pollinated the early population. In contrast, the late population flowered when most D. nuttallianum populations were in flower and I. aggregata was in flower as well. Solitary bees and bumblebees pollinated the late population.

Many studies in the literature have related time of flowering and seed set per flower or per plant. Few of these studies have looked at the quality of the seeds produced through the season. To my knowledge only Galen and Stanton (1991, 1993) and Lacey and Pace (1983) have done so. They used, respectively, seed weight and germination rates, and germination, survivorship, and growth to assess the quality of the offspring from individuals flowering at different times. Another indirect way to estimate quality of the seeds is to measure outcrossing rates. There are few papers that study the relationship between time of blooming and outcrossing rates and among them none does so experimentally. Stephenson (1982) conducted a nonexperimental study of Catalpa speciosa and concluded that plants that flower later in the season have higher outcrossing rates. Only outcrossed flowers form fruits in this species, so fruit production became his estimate for outcrossing. Zimmerman (1987), working with Polemonium foliosissimum, also concluded that late plants have higher outcrossing potential. He based his conclusion on pollinators' movements through the season. For D. nuttal-lianum, experimental crossings indicated that self versus outcrossed seeds differed in fitness. Outcrossed individuals outperformed the self-pollinated progeny (outbred progeny, grew larger, survived longer, and reached sexual maturity faster) (Waser and Price 1994).

0 0

Post a comment