How important is seed predation to recruitment in stable populations of long-lived perennials?

The many evolutionary modifications to seed biology in response to seed predation do not necessarily imply that seed predators have an important impact on population recruitment. This is because competition between individual plants for rare safe sites can cause an 'oversupply' of seeds so far as a population is concerned. The importance of seed losses to population recruitment at any point in time is related to the abundance of safe sites (Fig. 1): it is zero when safe sites are absent, negligible when safe sites are rare, and greatest when safe sites are numerous enough for recruitment to be limited by seed supply. Here I interpret the impact of severe seed losses on population recruitment in four species of long-lived perennials (Eucalyptus baxteri, Leptospermum juniperinum, Casuarina pusilla and L. myrsinoides) by considering these losses in terms of the overall seed dynamics of the populations. I focus on seed supply and seedling survival, as a measure of the current abundance of safe sites, and the maintenance of seed banks, as a measure of the ability of populations to exploit any future changes in safe site abundance. Insect seed predators destroyed about 95% of total seeds in each case. However, these losses do not necessarily have an important impact on population recruitment, because: (i) in most years recruitment appears to be limited by a rarity of safe sites, and not by seed supply (which was as high as 43 germinable seeds/m2/yr); and (ii) the losses did not prevent the establishment of large seed banks (ranging from >30 to >1000 viable seeds/m2) potentially capable of exploiting temporary conditions favourable for recruitment. In contrast to the situation with many annual plants, patterns of recruitment in stable populations of long-lived perennials are often extremely complex, and the significance of seed losses therefore difficult to determine.