5.4.1. What techniques are available to estimate the number of dogs?

If information on the number of dogs present in the community is not available, it is recommended, but not required, that a dog population survey be conducted before implementing a canine rabies control programme. These surveys assist with more accurate campaign planning, assessing the needs of dog population management programmes, and evaluating the effectiveness of intervention. If the campaigns need to be implemented with some urgency, first rapid population estimates can be made, as described here, and additional surveys can be implemented post-vaccination (e.g. combined with surveys for estimation of vaccination coverage, described here).

The options for estimating the number of dogs to vaccinate are as follows:

 Expert opinion based on historical data of previous campaigns or on registration records if available.

 Expert opinion based on estimations made in other geographic areas/demographic settings.

 Commonly used census techniques:

• Questionnaire surveys can be used to establish the mean number of owned dogs per household and dog:human ratios. Since the total human population or number of households is generally known through national population censuses, an estimate of the owned dog population can then be extrapolated. These surveys can be conducted before, during or after campaigns (e.g. combined with post-vaccination surveys to estimate vaccination coverage, described here). Households for interview should be selected randomly. Additional information can be obtained on:

  (1) Dog characteristics (e.g. sex, age, population turnover, dog handling/keeping practices, vaccination status), features of the animal that are helpful for understanding reproductive patterns, annual recruitment into the dog population (critical to determine the required frequency of campaigns) and level of supervision and accessibility.

  (2) Household characteristics including determinants of dog ownership (e.g. socio-economic status, livestock ownership, religion and sex of the head of households), which can be used as predictors of dog population size and distribution (as shown in this study) and to gather information on accessibility for vaccination.

  (3) Knowledge of rabies in the household, which can be useful for targeting public awareness campaigns, as described here.

  It should be understood that the additional information is not always necessary, and priorities should be based on resources available to carry out the surveys.

Photo courtesy of Serengeti Carnivore Disease Project

• Methods are also available to assess the number of roaming dogs (both owned and unowned dogs, but not accompanied by an owner) such as:
  • Indicator counts which consist of counting dogs (e.g. males, females and pups) along selected representative routes. Counts can be repeated every year (at the same time of the year) to evaluate changes in population over time (i.e. if the number of dogs has increased or decreased). Click here to learn more about this method.
  • Capture-mark-recapture methods consist of temporarily marking dogs, e.g. with a dye or distinctive collars (click here for examples of devices to mark dogs), and then subsequently recording the proportion of marked individuals in the population during a “visual recapture” effort. From the number of dogs marked and the observed ratio of marked to unmarked dogs the total number of street dogs is calculated. Marking can conveniently be done during vaccination campaigns. However, it is important to perform marking and recapture within a time period of a few days in order to minimize mark loss, dog movement and mortality effects. Also it is better to work in a defined area of only 0.5 to 2 km2 than in transects. The observations can also be combined with estimates of the number of owned dogs obtained through household questionnaires to estimate the number of unmarked dogs as shown in these studies. Estimates of population densities can also be obtained from the rates of capture during a marking campaign lasting a few days. Dogs may also be given permanent marks if the intervention includes anesthesia for additional studies.
    

    

    Photo courtesy of Serengeti Carnivore Disease Project

• • Population estimates can be obtained by extrapolating counts made in a sample (e.g. randomly selected subregions to whole cities). These surveys can also be repeated to detect changes in the number of roaming dogs. Click here to learn more about this method and how to estimate the total number of dogs from the sample.

    See here for examples of dog population estimates made in various settings.