Seed dynamics of Abies balsamea and Acer saccharum in a deciduous forest of northeastern North America

We studied the seed dynamics of Abies balsamea and Acer saccharum in a sugar maple-yellow birch forest of Quebec, Canada. Seed rain was censused every week from spring to autumn in 1988 and 1989. The soil seed bank was sampled four times during the 1988 growing season. Abies seed rain varied significantly between 1988 and 1989. A total of 92.5 +/- 7.0 seeds m-2 (mean +/- 1 SE) were produced in 1988; 1989 seed production was null. Seed viability in 1988 averaged 31.5%. Dispersal during the winter period accounted for ca. 22% of the annual seed crop. The spatial dispersion pattern of the seed rain was contagious for the autumn period, but was not significantly different from random for the winter period. There was only a weak relationship in the spatial distribution of the seed rain between the autumn and winter periods. Abies did not maintain a persistent seed bank on the site; viable seeds were present on the soil only after seed dispersal had started. Postdispersal seed mortality was relatively high, ca. 70%. Acer seed rain started soon after the flowering period, but consisted at that time only of aborted, underdeveloped samaras. Viable seeds were disseminated from September through early winter. Dispersal during the autumn accounted for ca. 98% of the annual seed crop. There were significant differences between seed rain abundance of 1988 and that of 1989 (118.1 +/- 9.5 samaras m-2 and 158.9 +/- 16.3 samaras m-2 in 1988 and in 1989, respectively). Seed viability also varied significantly between the 2 years, i.e., 19% in 1988 and 5% in 1989. For both 1988 and 1989, samaras were contagiously dispersed over the site. There was a significant positive relationship between the spatial distribution of the seed rain in 1988 and that in 1989. Seed rain abundance was high mainly in the proximity of mature Acer trees. As for Abies, Acer did not maintain a persistent seed bank in the soil; viable seeds were present only in the autumn seed bank sampling, after seed fall had started. Postdispersal seed mortality was relatively low, i.e., 20%. Abies and Acer are quite similar in their regenerative traits, and these contrast sharply with those of Betula alleghaniensis, a regular member of this forest community. Differences in regenerative traits may contribute to the coexistence of these tree species considering the gap regime of the system studied.