Temporal changes in seed and bud banks as sources of new propagules during prairie restoration

Background/Question/Methods

Plants maintain community structure through regeneration dynamics, which consists of the seed and bud banks.  Seed banks are generally classified as either transient, where seeds can germinate soon after leaving the parent plant, or persistent, where seeds remain dormant in the soil.  Although the seed bank may play a role in population establishment, systems dominated by perennial species can rely more extensively on vegetative reproduction (buds).  Recent studies have shown that the bud bank is actually more important for regeneration and maintenance of tallgrass prairies than seed banks.  The objective of this study was to quantify seed banks, seedling emergence and bud banks across a restoration chronosequence (sixteen 3-11 year old restorations).  We quantified seed and bud banks across a chronosequence of restored prairie to determine temporal dynamics of new propagules in restored prairie and the contribution of each propagule source in native prairie. At each site, ten soil cores (5.5 x 10cm) were removed to quantify the seed bank and five soil monoliths (25 x 25 x 10cm) were extracted and examined for buds.  Temporal dynamics in seed and bud banks across the restoration chronosequence were evaluated using linear and non-linear regression.

Results/Conclusions

A total of 1,412 seedlings from 59 species and 20 families germinated in the greenhouse.  Bud bank samples consisted of 52 species from 12 families.  Species richness (r² = 0.48, P = 0.009) and diversity (r² = 0.33, P = 0.031) of the seed bank declined with restoration age.  Contrary to our predictions, the transient and persistent seed banks became more dissimilar with restoration age (r² = 0.28, P = 0.041).  Total buds (r² = 0.37, P = 0.012) increased across the restoration chronosequence.  There was no directional change in the bud bank of C₃ and C₄ grasses and forbs across the chronosequence of restored grasslands.  Emerging seedlings declined with restoration age (r² = 0.41, P = 0.013).  This study demonstrates the importance of initial seed input during the early stages of restoration, but over time community structure reflects the composition of the bud bank.  A decline in seedling emergence across the restoration chronosequence may be attributed to a lack of suitable microhabitats for recruitment of new individuals.