Soil seed banks, both transient and persistent, are essential in shaping and maintaining the population dynamics of many plant species. Changes in the size and structure of the seed bank can affect the asymptotic behavior of demographic models and, thus, any projections of population viability. Properly accounting for these early stages becomes particularly critical under conditions of rapid environmental change and disruption of natural and historic disturbance patterns. Paradoxically, it appears that a large portion of the plant demographic models available in the literature either assume an absence of persistent seed banks or downplay their importance. We evaluate the validity of such assumptions and examine potential factors contributing to inconsistencies in the reports of seed banks. We are performing systematic reviews of the available literature on plant demographic models and seed bank longevity. We assess the congruency between both types of studies at different taxonomic levels. Further, we are analyzing emerging patterns in the context of phylogenetic relationships, habitat types, and other seed ecology characteristics, including dormancy and predation.
Results/Conclusions
A preliminary exploration of demographic models available in an on-line repository (COMPADRE Plant Matrix Database) indicated only 39 of the 695 species represented were associated with models that included seed bank parameters. Independently, we obtained seed longevity information for 111 of these species from the LEDA Traitbase, which compiles studies of seed banks through a strict protocol and classifies them into long-term persistent (>5 years), short-term persistent (1-5 years) and transient (<1year). Remarkably, we found only 34.2% congruence between the two databases, with 71 species expected to form seed banks (25 short-term and 46 long-term) but that were omitted in the population models. We are currently analyzing associated variables to identify potential factors behind such inconsistencies. We are incorporating information regarding the sampling method used for each entry in the LEDA Traitbase. We are also collecting habitat data which will be combined with information on seed characteristics, such as dormancy type. Finally, we compare this information within a phylogenetic framework. These results would help to delineate areas of conflict and suggest potential solutions. We are expanding the scope of this work by searching in the literature for studies not represented in the two databases.