3.5.6 Mark-recapture/resight

Use mark-recapture/resight (MRR) techniques for calculating Absolute Abundance for non-breeding birds when counts are not feasible. It is one of the methods recommended for swans, geese, and Sandhill Cranes. It has very specialized and limited application and is a useful census method under only two conditions:

1. to census localized wintering populations of a species where the limits of the range are known and during a period when immigration and emigration are known not to occur (e.g. Swans in the Comox area); or
2. to census resident, nonmigratory populations with a known distribution (e.g. Canada geese on Southern Vancouver Island).

In addition, Mark and Recapture/Resight can be used for two non inventory purposes:

1. Mark-recapture techniques can be used to determine geographic patterns of distribution during migration, calculate indexes of population size and production, or estimate survival and assess harvest pressure on different age/sex classes (Brownie et al. 1985; Bibby et al. 1992). Therefore they can be much more useful in providing insight into mechanisms of population change than in measuring the actual size of changes.

   These purposes require international cooperation and these kinds of studies are likely to be used only by federal wildlife authorities; or

2. To mark breeding birds for local breeding biology studies.

Because the focus of this manual is inventory methods, mark and recapture will be discussed only in regard to the first two points.

Mark-recapture/resight is recommended for assessing absolute abundance only when counts are not feasible, and only in the special situations listed above. Estimating population size using mark-recapture/resight techniques is based on the assumption that, if a proportion of the population is marked in some way and is then returned to the original population, complete mixing occurs. A second sample is then taken. The number of marked individuals in the second sample should have the same ratio to the total numbers in the second sample as the total number of marked individuals originally released has to the total population. By knowing the number originally caught, the number marked and the number of marked individuals in the second sample, it is possible to estimate the total population.

There are many considerations involved in determining whether mark-recapture/resight is an appropriate technique for estimating population size: (1) it must be possible to catch enough individuals and all individuals should be equally catchable. It is likely that some species/sexes/age classes will be easier to catch than others; (2) it is important that the mark will not harm the bird (e.g., make it more vulnerable to predators); (3) the chances of recapture or resighting must not be affected by the fact that it was caught in the first place. The population should be sampled randomly in recapture/resighting events; (4) marks must be permanent and detectable; and (5) the population must be closed or immigration and emigration must be measured. These assumptions and detailed methods for estimating population size can be found in Bibby et al. 1992. Mark-resight recapture requires adherence to these numerous assumptions, large numbers of marked individuals, and high recapture/resight probabilities.

Recapture/Resight/Recovery

Once a large enough proportion of the bird population has been marked (first sample), then the second sample is taken. This may be done by recapturing the birds, or by resight and/or recovery.

Advantages:

• Data can be collected over a large window of time.
• Provides extensive information about habitat selection, dispersal and other movements and survival rates.
• Even low returns and low rates of resighting provide information on distribution and habitat use.

Disadvantages:

• Tends to require extensive time and effort and is best done on a local scale (does not monitor populations over a large geographic range).
• A large sample of birds (5 to 10% of the estimated population size) need to be banded and recovered/resighted to get good estimates.
• There are numerous assumptions required to calculate population size and survival rates (Bibby et al. 1992) that can be violated but they are often not tested.
• Behaviour of many species of birds may make it difficult to read bands and since large numbers of recoveries are necessary, additional markers, such as nasal disks, are essential.
• Marking birds with markers other than bands may increase their mortality or alter their social status.
• Recaptured sample is rarely random - some birds may be more vulnerable to recapture than others, and there is a tendency to capture those birds.

There are numerous variations on the MRR method. However, all have common assumptions that must be met, or approximated, in order for subsequent data analysis and abundance estimates to be valid. These assumptions are:

• Demographic and geographic closure (i.e., The sample population is not significantly altered by births, deaths, immigration or emigration during the time of sampling).
• All members of a population must have an equal or known probability of being captured (i.e., Any sample should be representative of the population being sampled).

Only in rare situations are both of these assumptions ever met and for this reason use of this census method is not encouraged. However, if the method is used, it is important that continued efforts be made to approximate the assumptions, if accurate estimates of population size are to be attained. It is possible to minimize the effects of violating assumptions by modifying the general approaches discussed below. It can be useful to review some modifications to the basic formulas in literature such as Eberhardt (1969), Cormack (1972), O'Farrell et al. (1977), Pollock (1982), Krebs and Boonstra (1984), Nichols et al. (1984), Kenneth and Anderson (1985), Wilson and Anderson (1985), and Chao (1988).

Other disadvantages of using this method are the complexities of analysing the data. Computer programs such as CAPTURE attempt to integrate all of these formulae into one comprehensive package, but even such programs require considerable expertise to ensure meaningful results. Researchers are advised to consult documents which deal with the suite of models used in CAPTURE, such as Otis et al. (1978), White et al. (1982), or Rexstat and Burnham (1991). See also the introductory manual, Species Inventory Fundamentals, No. 1.

If a density estimate is required then at least four sessions should be conducted in a brief time period (to minimize violations of the assumption of closure). For this reason, recapture of birds, many of which become wary of traps while others are repeatedly caught, is ineffective, so resighting methods must be used. Data from this design can be used with program CAPTURE (see White et al. 1982).

As an alternative, the Jolly-Seber model will allow calculation of survival estimates, as well as a population size; however, biologists should be aware that this is not a true measure of density even though it provides a good measure for comparison over time or among areas. Sampling sessions can be conducted with longer durations between resighting periods (i.e., a few sessions each month). Many analysis options exist for the Jolly-Seber model; these are discussed in Species Inventory Fundamentals, No. 1.

Both of the designs above can be combined to allow density and survival estimates. This design is called "Pollock's robust design" and is also discussed in Species Inventory Fundamentals, No.1.

Office Procedures

• Review the introductory manual No. 1 Species Inventory Fundamentals.
• Obtain relevant maps of the project area (e.g., 1:50 000 air photo maps, 1:20 000 forest cover maps, 1:20 000 TRIM maps, 1:50 000 NTS topographic maps). Any map which is used to record data should be referenced to NAD83.
• Outline the Project Area on a map and determine Biogeoclimatic zones and subzones, Ecoregion, Ecosection, and Broad Ecosystem Units for the Project Area from maps.
• Delineate one to many Study Areas within this Project Area. Study Areas should be representative of the Project Area if conclusions are to be made about the Project Area. For example, this means if a system of stratification is used in the Sampling Design then strata within the Study Areas should represent relevant strata in the larger Project Area.
• Select banding sites in areas that will ensure that the banded sample is representative of the population of interest, e.g., areas where birds concentrate to moult.
• Acquire banding permits and special permits to mark birds.