Generating an accurate picture of the entire process of dispersal in a species involves detailed demographic analyses in addition to tracking emigrating and immigrating individuals (see Spatial Distribution and Spatial Distribution Models). In order to know what demographic parameters drive effective dispersal, it is important to know how many individuals leave, survive the exploratory process, and breed successfully in the new area. Ideally, all parameters of dispersal should be quantified. However, because the dynamics of a population are directly driven by effective dispersal, it may be unnecessary to conduct detailed studies of each stage of dispersal, depending on the particular goals of the researcher.
Mark-recapture methods and demographic analyses can assist in the estimation of many dispersal-related parameters, and though the route traveled by the individual captured in a new patch is often unknown, it is still possible to gain an estimate of immigration and emigration rates. Many studies focus on relatively local effects of dispersal, studying population dynamics in a few interconnected populations that are spatially tractable. These studies involve either mark-recapture methods, genetic methods, seed traps (for plants), or radio- or satellite-tracking methods. Animals and seeds can both be marked using tags, paint, or dyes. Tracking methods, such as by radio telemetry or satellite, show great promise for obtaining detailed information on dispersal patterns, especially on the tail of the dispersal curve. Genetic methods are becoming popular as well, because they can detect effects of very low rates of dispersal over long distances. Each method has advantages and disadvantages, and all these methods have assumptions and uncertainties associated with them, which must be taken into account when analyzing data and estimating dispersal curves.
Collecting data from the tail of the dispersal curve can be difficult, either hampered by the difficulty of maintaining sampling densities, or due simply to the rare and stochastic nature of long-distance dispersal events. The importance of long-distance dispersal in estimating the spread of populations was highlighted in scientific literature when the rate of post-Pleistocene expansion of treesin Europe estimated with models neglecting long-distance dispersal could not account for the rapid expansion rates observed as the glaciers receded. Incorporating longdistance dispersal by modeling spread with leptokurtic dispersal curves matched the estimated rates of spread more closely. Unfortunately long-distance dispersal events are extremely difficult to measure empirically, and hence estimating them has since received much attention.
For animals, one way of estimating dispersal patterns involves marking and releasing animals, then observing the animals when they are collected, usually during an annual harvest. In the case of mark-harvest methods, animals are only viewed twice, once during the marking process, and once when harvested. This type of data may be useful for estimating mortality rates associated with movement from one site to another if it is possible to assume that the animals in question always return to either the original marking site or to the final capture site. Prior knowledge of movement patterns is important; mortality cannot be estimated if a significant number of animals disperse outside of the sample area. If multiple mark-capture episodes are accomplished in one season, it is possible to estimate probabilities of survival and movement for a specific area. However, no models currently can estimate dispersal from one area to another using this type of data.
In a similar method, animals are marked, released, then re-sighted or recaptured and released again. Animals may be sighted multiple times with this method, and with a robust sampling design, immigration and temporary emigration rates can be estimated. If this type of method is employed on multiple sites, with site-specific markers, immigration and emigration probabilities in addition to transition rates can be estimated. If possible to employ, this type of design is quite useful, as it provides data necessary for estimating population dynamic parameters associated with dispersal. Long-distance dispersal events are difficult to detect with this method, due to the stochasticity associated with the occurrence and detection ofsuch events.
Seed dispersal is often measured using seed traps to capture seeds at varying distances from the source. Seed traps usually involve pit traps or sticky traps placed in or near the ground. To identify seeds' origins, individual fruits can be marked, a chemical tagging method can be used, or a rare genetic variant can be used as a marker. The most common method is to measure the densities of seed deposited at various distances from a source. Because individual plants are not identified when only density of seeds can be recorded, likelihood methods are used to model dispersal curves. Seed traps work well for estimating dispersal curves near the source, but as distance from the source increases it becomes more and more difficult to detect dispersal events. If enough traps are used, longdistance dispersal events can be detected; however, such events will be rare, and their detection will be dependent on the resources available to the researcher.
Radio telemetry and satellite tracking provide excellent data, when practical. Such studies have documented that long-distance dispersal events are more common than estimates from mark-recapture methods suggest. Most studies involve large- to medium-sized animals, including marine mammals. Invaluable information about the long-distance travels ofthese animals has been collected, including information about movements of some seabirds. Ideally, a large proportion of a population could be followed individually, and detailed analyses made of their movements. In order to accomplish this, the radio or satellite transmission units should not inhibit movement or survival, and the batteries should be strong enough to allow signal detection at a distance for a long period of time. As technology advances, smaller tags can be used. For example, very small radar tags have lately been adapted for use on bumblebees, showing promise for generating detailed dispersal data for larger insects.
Was this article helpful?