Many plants suffer very heavy preand/or post-dispersal seed predation by animals. A few exemplary studies (2,18,38,52,74,87, 106, 110, 111, 124, 140, 161, 162, 171, 181, 187, 196, 197, 203, 205, 207, 217, 220, 227) and a variety of shorter reports scattered through the agricultural, botanical, and zoological literature suggest a large and important, yet unexploited, field of study. It is clear that the pattern of seed predation is highly structured and that it co evolved at, the chemical, spatial, and temporal level. It involves all levels of animal-plant interaction from the internal energy budget of indi­ viduals to the entire community (203 ) . Owing to parental and sibling com­ petition, successful development of a seedling may depend on the seed's dis­ persal (101). Equally important, thc seed must escape from the predators at the seed crop and in the parent plant's habitat before and after dispersal (111 ) . The game is played by mobile predators in search of sessile prey; escape is through a single dispersal move, seed chemistry, parental morphol­ ogy and behavior, and evolutionary change. The processes and patterns are:ideal candidates for ecological and evolu­ tionary analyses (83, 111 ) of the typcs traditionally conducted by zoologi­ cally oriented biologists (37, 51,73, 141, 155,156, 182) or applied to plants in general ( 40,116,208,209,241) . Seed predators and leaf eaters are often lumped together in ecological discussions, but they differ in ways especially important in the coevolution of seed predators and seeds: (a) Like leaves, seeds are highly subdivided and small, but unlike leaves and other vegetative parts, they have very high nutrient values per unit volume. ( b) While obvious in large quantities, they may be extremely inconspicuous once dispersed. ( c) Being comparatively dormant, seeds have low self-repair abilities but can contain high concentra­ tions of toxic compounds (secondary substances) with less sophisticated ad­ aptations against self-intoxication than can leaves and meristematic tissues. ( d ) Seed production is not continually required for direct survival of the parent plant; seed timing, quantity, and quality can therefore be manipu­ lated more freely by natural selection than can these traits of leaves and other vegetative parts. (e ) Seeds are not directly replaced· when killed or germinated; once removed, seeds may be absent far longer than edible vege­ tative parts and thus a common plant species may be rare in time,toa seed

Seed Predation by Animals

discon­ tinuities on many scales in time and space. The patterns of these discontinuities produce an environmental patchwork which exerts powerful influences on the distri­ butions of organisms, their interactions, and their adaptations. Consideration of this environmental patch structure is critical to both the theory and management of populations. Despite the obvious heterogeneity of natural sys­ te.ms, most of the models that form the theoretical fabric of population biology and ecology (and that are increasingly conditioning our perception of reality) tell mathe­ matical stories of populations

http://max2.ese.u-psud.fr/epc/conservation/PDFs/HIPE/Wiens1976.pdf

Population Responses to Patchy Environments

Recruitment limitation of tree population dynamics is poorly understood, because fecundity and dispersal are difficult to characterize in closed stands. We present an approach that estimates seed production and dispersal under closed canopies and four limitations on recruitment: tree density and location, fecundity, seed dispersal, and estab- lishment. Consistent estimates are obtained for 14 canopy species using 5 yr of census data from 100 seed traps and several thousand mapped trees and seedlings from five southern Appalachian forest stands that span gradients in elevation and moisture. Fecundity (seed production per square centimeter of basal area) ranged over four orders of magnitude, from 10 0 cm 2 basal area/yr (Carya, Cornus, Nyssa, Quercus )t o.10 3 cm 2 /yr (Betula). Mean dispersal distance ranged from , 5m( Cornus, Nyssa )t o.20 m (Acer, Betula, Liriodendron, Tsuga) and was positively correlated with fecundity. Species also differ in the degree of seed clumping at fine (1 m 2 ) spatial scales. Dispersal patterns can be classed in two groups based on dispersal vector: wind-dispersed taxa with high fecundities, long-distance dis- persal, and low clumping vs. animal-dispersal taxa with low fecundities, short-distance dispersal, and a high degree of clumping. ''Colonization'' limitations caused by sizes and locations of parent trees, fecundity, and dispersal were quantified as the fraction of sites receiving seed relative to that expected under null models that assume dispersal is nonlocal (i.e., long-distance) and not clumped (i.e., Poisson). Difference among species in coloni- zation levels ranged from those capable of saturating the forest floor with seed in most stands (Acer, Betula, Liriodendron) to ones that leave much of the forest floor without seed, despite presence of adults (Carya, Cornus, Nyssa, Oxydendrum ). Seedling establish- ment is one of the strongest filters on recruitment in our study area. Taken together, our results indicate (1) that fecundity and dispersal can be resolved, even under a closed canopy, and (2) that recruitment of many species is limited by the density and location of source, dispersal patterns, or both.

/pdf/stages-and-spatial-scales-of-recruitment-limitation-in-3q0ieobp1o.pdf

Stages and spatial scales of recruitment limitation in southern appalachian forests

Recruitment limitation of tree population dynamics is poorly understood, because fecundity and dispersal are difficult to characterize in closed stands. We present an approach that estimates seed production and dispersal under closed canopies and four limitations on recruitment: tree density and location, fecundity, seed dispersal, and estab- lishment. Consistent estimates are obtained for 14 canopy species using 5 yr of census data from 100 seed traps and several thousand mapped trees and seedlings from five southern Appalachian forest stands that span gradients in elevation and moisture. Fecundity (seed production per square centimeter of basal area) ranged over four orders of magnitude, from 10° cm2 basal area/yr (Carya, Cornus, Nyssa, Quercus) to >10 3 cmVyr (Betula). Mean dispersal distance ranged from 20 m (Acer, Betula, Liriodendron, Tsuga) and was positively correlated with fecundity. Species also differ in the degree of seed clumping at fine (1 m2) spatial scales. Dispersal patterns can be classed in two groups based on dispersal vector: wind-dispersed taxa with high fecundities, long-distance dis- persal, and low clumping vs. animal-dispersal taxa with low fecundities, short-distance dispersal, and a high degree of clumping. "Colonization" limitations caused by sizes and locations of parent trees, fecundity, and dispersal were quantified as the fraction of sites receiving seed relative to that expected under null models that assume dispersal is nonlocal (i.e., long-distance) and not clumped (i.e., Poisson). Difference among species in coloni- zation levels ranged from those capable of saturating the forest floor with seed in most stands (Acer, Betula, Liriodendron) to ones that leave much of the forest floor without seed, despite presence of adults (Carya, Cornus, Nyssa, Oxydendrum). Seedling establish- ment is one of the strongest filters on recruitment in our study area. Taken together, our results indicate (1) that fecundity and dispersal can be resolved, even under a closed canopy, and (2) that recruitment of many species is limited by the density and location of source, dispersal patterns, or both.

Stages and spatial scales of recruitment limitation in

Camera Traps in Animal Ecology

Typical seed caches made in the fall of the year by white—footed mice (Peromyscus leucopus) and red—backed voles (Clethrionomys gapperi) contained 20—30 pine seeds, buried beneath the litter of pine needles and in contact with mineral soil. Many caches were revisited by the mice and their contents destroyed before winter. Most caches not decimated in the fall were invaded the following spring. However, some of them escaped complete destruction and their unconsumed seeds germinated and produced seedlings. Although mice make their caches in micro—environments highly favorable to germination, their subsequent feedings on the stored seeds, both before snowfall and before and during the spring germination period, greatly deplete the number of potential germinants. Therefore, the contribution to forest regeneration of seeds stored by these small forest mammals is probably of minor significance except when bumper seed crops are produced and large numbers of caches are created. Under these circumstances, stored seeds may be of sufficient quantity to exceed food requirements of the mice. If pirating by other animals is low, then a surplus of seeds is left to help regenerate the forest.

Herschel G. Abbott

Papers

Ecology of Eastern White Pine Seed Caches made by Small Forest Mammals

Tree Seed Preferences of Mice and Voles in the Northeast

A Photoelectric 35-Mm Camera Device for Recording Animal Behavior

Forest Nursery Practices in the United States

Direct seeding of Paper Birch in strop clearcutting.