Deterministic camera trap monitoring
In Switzerland, the deterministic camera trap monitoring was developed for the focus species lynx, but the principle can be applied to any species with individually recognisable external markings. The deterministic camera trap monitoring uses camera traps for a pre-defined duration and distributed systematically over an area. This approach allows KORA to estimate the population abundance and density of lynx and wildcat, and with regular repetition of the monitoring to follow the development of the populations. This is done with the capture-recapture method.
The capture-recapture method with systematic sampling allows us to estimate to the total population of species which are individually identifiable due to natural or artificial markings, e.g. spotted or striped cats. Through repeated photographs (“captures”), we can estimate their numbers, their capture probability, and the corresponding statistical errors with capture-recapture models. This allows for the best possible estimate of the true population size, and evaluating the precision of the estimate. Usually, the model Mh is used which enables the use of different capture probabilities. In exceptional cases, the models M0, Mth or Mbh are used. The model M0 uses the same capture probability for all individuals. The model Mth allows the use of different individual capture probabilities, that can change over time. The model Mbh uses individual capture probabilities, that change after the first capture.
In Switzerland, camera trapping has been used since 1998 to estimate the abundance and density of lynx. KORA performs the deterministic camera trap monitoring sessions in reference areas, which are distributed in the large carnivore management compartments and cover a large part of the current lynx distribution. Each session is coordinated in advance with the respective cantonal offices and game wardens. In each reference area, camera traps are installed every three or four years and run for 60 nights to get the samples for the capture-recapture method. Each lynx must be individually identified from its fur pattern. For a definitive identification, both flanks of the individual must be photographed, as the fur pattern differs. The deterministic camera trap sessions are supported by the “opportunistic monitoring” with which as much information as possible is already gained on the present lynx and their fur pattern. From the number of lynx images and the number of recorded individuals, KORA can estimate the total number of lynx in the reference area with the capture-recapture model. The population size is expressed as the number of “independent lynx”, i.e. resident adult and not yet resident subadult individuals. Juveniles that are still with their mother are not counted. The estimated number of lynx in the reference area is then converted into the lynx density in the reference area. Additionally, the density per suitable habitat is calculated to compare the densities of different reference areas. The suitable habitat in Switzerland and in the whole of the Alps was identified in 2004 as part of a doctoral thesis (Zimmermann 2004). The results for each session are reported in individual KORA reports.
The calculated population densities within the reference areas are used to calculate the annual abundance within the management sub-compartments. A grid of 10 x 10 km is put over Switzerland. Each cell of the grid is counted as permanently populated if there are observations of the SCALP categories C1 or C2 for two of the last three years. For all permanently settled cells, the sum of the suitable habitat is calculated from the habitat model (Zimmermann 2004). The density that was estimated during the latest deterministic camera trap monitoring session is then applied to this sum of suitable habitat to estimate the population size within the management sub-compartment. This extrapolation assumes (1) that KORA used the correct habitat variables, (2) that the density of lynx in suitable habitat within the management sub-compartment is similar inside and outside of the reference area, and (3) that the density did not change considerably since the last camera trap session. There are some sub-compartments where no reference area has yet been established. In those sub-compartments, the estimated lynx abundance is based on the minimum number of lynx that have been identified, or on expert opinion if there has been no opportunistic camera trap monitoring.
During KORA’s camera trap monitoring of lynx in the Jura mountains, wildcats are regularly captured, too. These are important data on the occurrence of wildcat in the research areas. During a pilot project, a deterministic camera trap monitoring was installed for the wildcat synergetic to the lynx camera trap monitoring in the Jura mountains in order to estimate the abundance of wildcats. This pilot project showed camera traps to be suitable for the monitoring of wildcats. Wildcats and domestic cats can be distinguished using several characteristics in the pelage (DE, FR). Individual wildcats can be distinguished based on their fur patterns, too. Since then, KORA used the deterministic camera trap monitoring in reference areas specifically defined for the wildcat to assess the population size and trend of the species. A genetic monitoring is performed simultaneously with the camera trap monitoring. To this end, a coarse wooden pole is set up at every camera trap location and sprayed with valerian as an attractant. Wildcats attracted by the smell will rub their cheeks on the pole and leave some hairs on the coarse wood. These hairs are collected and then used for the genetic identification of the species, the individual, as well as the hybridisation rate.