Alternate bearing, predator satiation and seedling recruitment in Quercus robur L.

1 We investigated the relationship between seed production and seedling recruitment in Quercus robur in order to determine whether oak recruitment was seed-limited, herbivore-limited or microsite-limited. Over a 15-year period, Q. robur showed a pattern of alternate bearing, with significant (but not complete) synchrony between individual trees at several sites in south-east England. 2 Seedling recruitment was assessed in three ways: by annual destructive sampling, by monitoring permanent quadrats, and by ageing destructive samples of saplings using basal ring counts. 3 Seedling recruitment was closely correlated with the size of the acorn crop in some habitats but not others. In Sunningdale, where a competition-free seed bed was available every year and where seed predators were relatively scarce, oak seedlings appeared almost every year, and three peak densities of seedling recruitment followed three peaks of acorn production during the period 1985-92. In Silwood Park, where rabbits were abundant, there was little seedling recruitment except within rabbit exclosures and then only in years of peak acorn production. In Windsor Great Park, rabbit densities were lower and seedling recruitment was higher than at Silwood Park, but the association between seedling recruitment and acorn production could not be established with confidence because of imprecision in ageing the saplings. 4 Rates of acorn loss to invertebrate seed predators were broadly consistent with the hypothesis of predator satiation; low acorn crops suffered a higher percentage mortality than large acorn crops. The two principal acorn-feeding insects (the alien cynipid gall wasp Andricus quercuscalicis and the native weevil Curculio glandium) killed between 30% and 90% of the acorn crop in given years and between 0% and 100% of the acorn crop on individual trees. Vertebrate herbivores (principally rabbits and wood mice) destroyed most of the sound acorns in low years, removing experimentally placed acorns within 24 h. 5 The effectiveness of alternate bearing as a defence against short-lived, specialist insect herbivores and long-lived, generalist vertebrate herbivores is discussed. The evidence presently available does not allow us to discriminate between the competing hypotheses that alternate bearing is an evolved, antipredator strategy on the part of the oaks, or a constraint imposed by limited resource availability within the trees. 6 We conclude that the relationship between seed production and seedling recruitment in Quercus robur is variable from site to site: oak recruitment is seed-limited in open sites where vertebrate herbivores are scarce, microsite-limited beneath a dense forest canopy, but herbivore-limited in other sites, especially those where rabbits are abundant.