Long-distance seed dispersal influences many key aspects of the biology of plants, including spread of invasive species, metapopulation dynamics, and diversity and dynamics in plant communities. However, because long-distance seed dispersal is inherently hard to measure, there are few data sets that characterize the tails of seed dispersal curves. This paper is structured around two lines of argument. First, we argue that long-distance seed dispersal is of critical importance and, hence, that we must collect better data from the tails of seed dispersal curves. To make the case for the importance of long-distance seed dispersal, we review existing data and models of long-distance seed dispersal, focusing on situations in which seeds that travel long distances have a critical impact (colonization of islands, Holocene migrations, response to global change, metapopulation biology). Second, we argue that genetic methods provide a broadly applicable way to monitor long-distance seed dispersal; to place this argument in context, we review genetic estimates of plant migration rates. At present, several promising genetic approaches for estimating long-distance seed dispersal are under active development, including assignment methods, likelihood methods, genealogical methods, and genealogical/demographic methods. We close the paper by discussing important but as yet largely unexplored areas for future research.

It has been argued that nonstandard mechanisms of dispersal are often responsible for long-distance dispersal in plants. For example, plant seeds that appear to be adapted for wind dispersal may occasionally be dispersed long distances by birds, or vice versa. In this paper, we explore whether existing data on dispersal distances, colonization rates, and migration rates support the idea that dispersal processes suggested by the morphology of the dispersal unit are responsible for long distance dispersal. We conclude that the relationship between morphologically defined dispersal syndrome and long-distance dispersal is poor. This relationship is poor because the relationship between the morphology of dispersal units and the multiple processes that move seeds are often complex. We argue that understanding gleaned from the often anecdotal literature on nonstandard and standard means of long distance dispersal is the foundation for both statistical and mechanistic models of long-distance dispersal. Such models hold exciting promise for the development of a quantitative ecology of long-distance dispersal.

Long-distance dispersal (LDD) includes events in which propagules arrive, but do not necessarily establish, at a site far removed from their origin. Although important in a variety of ecological contexts, the system-specific nature of LDD makes ''far removed'' difficult to quantify, partly, but not exclusively, because of inherent uncertainty typically involved with the highly stochastic LDD processes. We critically review the main methods employed in studies of dispersal, in order to facilitate the evaluation of their pertinence to specific aspects of LDD research. Using a novel classification framework, we identify six main methodological groups: biogeographical; Eulerian and Lagrangian movement/redistributional; short-term and long-term genetic analyses; and modeling. We briefly discuss the strengths and weaknesses of the most promising methods available for estimation of LDD, illustrating them with examples from current studies.The rarity of LDD events will continue to make collecting, analyzing, and interpreting the necessary data difficult, and a simple and comprehensive definition of LDD will remain elusive. However, considerable advances have been made in some methodological areas, such as miniaturization of tracking devices, elaboration of stable isotope and genetic analyses, and refinement of mechanistic models. Combinations of methods are increasingly used to provide improved insight on LDD from multiple angles. However, human activities substantially increase the variety of long-distance transport avenues, making the estimation of LDD even more challenging.

The distribution of many woodland herbs extends 1000-2000 km in a north-south direction, yet the majority of these species grow clonally, have little recruitment by seed, and possess no obvious mechanism for long-distance seed dispersal. Although aware that woodland herbs disperse poorly, ecologists have tacitly assumed that, given long periods of time, even small dispersal distances would allow woodland herbs to colonize the vast geographic region they now occupy. We examined this assumption for the understory herb Asarum canadense. To estimate long-term rates of spread by seed, we calibrated seed-dispersal diffusion models with life history data and with data on seed carries by ants. We supplemented our field observations and modeling results for A. canadense with a literature survey on the dispersal capabilities of other plant species. Ants transported A. canadense seeds up to 35 m, the largest distance ants are known to move the seeds of any woodland herb. Empirically calibrated diffusion models indicated that over the last 16 000 yr A. canadense should only have traveled 10-11 km from its glacial refugia. In actuality, A. canadense moved hundreds of kilometers during this time. Models that examined the tail of A. canadense's seed-dispersal curve indicated that occasional dispersal events had to have a high frequency (0.001 on a per seed basis) and a large magnitude (dispersal distance 1 km) for A. canadense to have traveled over 200 km in 16 000 yr. The literature survey showed that most woodland herbs and many other forest, desert, coastal, and open-habitat plant species have limited seed-dispersal capabilities , similar to those in A. canadense. We conclude that woodland herbs, as well as many other plants, disperse so slowly that there is no documented mechanism by which most of these species could have reached their present geographical range since the last glacial maximum. This suggests that occasional events leading to long-distance dispersal dominate the Holocene colonization of northern temperate forest by woodland herbs, and this, in turn, has implications for issues ranging from the importance of genetic analyses to the structure of metapopulation models.

The distribution of many woodland herbs extends 1000-2000 km in a northsouth direction, yet the majority of these species grow clonally, have little recruitment by seed, and possess no obvious mechanism for long-distance seed dispersal. Although aware that woodland herbs disperse poorly, ecologists have tacitly assumed that, given long periods of time, even small dispersal distances would allow woodland herbs to colonize the vast geographic region they now occupy. We examined this assumption for the understory herb Asarum canadense. To estimate long-term rates of spread by seed, we calibrated seeddispersal diffusion models with life history data and with data on seed carries by ants. We supplemented our field observations and modeling results for A. canadense with a literature survey on the dispersal capabilities of other plant species.Ants transported A. canadense seeds up to 35 m, the largest distance ants are known to move the seeds of any woodland herb. Empirically calibrated diffusion models indicated that over the last 16 000 yr A. canadense should only have traveled 10-11 km from its glacial refugia. In actuality, A. canadense moved hundreds of kilometers during this time. Models that examined the tail of A. canadense's seed-dispersal curve indicated that occasional dispersal events had to have a high frequency (Ն0.001 on a per seed basis) and a large magnitude (dispersal distance Ͼ1 km) for A. canadense to have traveled over 200 km in 16 000 yr. The literature survey showed that most woodland herbs and many other forest, desert, coastal, and open-habitat plant species have limited seed-dispersal capabilities, similar to those in A. canadense. We conclude that woodland herbs, as well as many other plants, disperse so slowly that there is no documented mechanism by which most of these species could have reached their present geographical range since the last glacial maximum. This suggests that occasional events leading to long-distance dispersal dominate the Holocene colonization of northern temperate forest by woodland herbs, and this, in turn, has implications for issues ranging from the importance of genetic analyses to the structure of metapopulation models.

A stochastic simulation model that relates the success of various herbivore searching behaviors to the density and arrangement of food plants was developed. It was designed to be fieldtestable with all model parameters directly measurable in field experiments. The model uses probability distributions and parameter ranges that are in accordance with published herbivore movement data.In the model, herbivore insect searching success depends in a complex manner upon parameters such as insect move lengths, turning angle concentration, reactive distance, risk of mortality while searching, plant density, plant dispersion, and plant stand size. Simulation results preclude simple generalizations and indicate that the outcome of interactions between plant density, plant dispersion, and herbivore searching behavior depends heavily upon parameter values. I argue that species-tospecies shifts in the details of the simple searching process that I simulate could explain the (experimentally observed) high variation in herbivore response to vegetation texture.I discuss plant and insect strategies that are "optimal" within biological constraints. For example, results from the model suggest that under many conditions clumping by plants can be an effective escape strategy from searching herbivores. This contradicts the claim that plant spatial distribution does not affect insect searching success, a claim that is justified only if it is assumed that organisms search (infinitely long) without biological cost. The importance of considering the cost of ineffective search is emphasized.

There are few studies in natural ecosystems on how spatial maps of soil attributes change within a growing season. In part, this is due to methodological diculties associated with sampling the same spatial locations repeatedly over time. We describe the use of ion exchange membrane spikes, a relatively nondestructive way to measure how soil resources at a given point in space¯uctuate over time. We used this method to examine spatial patterns of soil ammonium (NH 4 ) and nitrate (NO

Through plasticity in traits controlling clonal morphology, clonal plants possess the potential to selectively place ramets within a heterogeneous environment or "forage." Although many studies document plant responses that are consistent with foraging, few studies test directly whether plants can preferentially locate "good" patches or avoid "bad " patches when grown in a heterogeneous environment. We conducted such a test for Hydrocotyle bonariensis, a clonal dune species that inhabits soils known to be extremely patchy on both time and space. We subjected Hydrocotyle to a temporally constant but spatially variable soil resource environment created by clipped patches of grass. Clonal morphology on the patch treatment was compared to that in two homogeneous treatments: no grass and full grass. In order to predict the effect of our treatments on the long—term expansion of clones, we calibrated diffusion models of clonal growth with data on the morphological response of Hydrocotyle to our treatments. Clones of Hydrocotyle bonariensis were able to respond to the presence of patches by selectively placing ramets outside of grass patches, thus providing direct evidence of an effective foraging response. For each of the three traits identified as potentially contributing to an overall foraging response (branching, internode distance, and direction of rhizome growth), the response in and out of grass patches was substantially different from the degree of plasticity manifested by the two homogeneous treatments. For example, no difference in main rhizome internode lengths was found between the two homogeneous treatments. In contrast, ramets in the heterogeneous (patchy grass) treatment responded to their local environment by increasing internode lengths when in the unfavorable (grass patch) portion of the environment. Empirically calculated diffusion models indicate that habitat complexity has considerable impact on the long—term expansion of clones: only in the heterogeneous treatment did clones expand less rapidly in favorable habitat than they did in unfavorable habitat. In the heterogeneous treatment, Hydrocotyle rhizomes exhibited a previously undocumented behavior: they appeared to veer away from patches of grass. Finally, we discuss how the foraging response of Hydrocotyle ramets may be enhanced by its previously documented capacity for resource integration.