Abstract from BTO Research Report 324:

Joys, A., Clark, J.A., Clark, N.A. & Robinson, R. (2005)

An investigation of the effectiveness of rehabilitation of birds as shown by ringing recoveries.

1. This report aims to assess post-release survival of rehabilitated birds and compare this to the survival of non-rehabilitated birds using distance and time elapsed between ringing and finding. Comparisons are made between selected species and ecological groups of species to enable an assessment of rehabilitation success. From this it is possible to prioritise species best suited for rehabilitation in terms of post-release survival if resources available are limited. This report does not consider the moral, ethical or conservation value of rehabilitating individuals of a species.

2. The BTO National Ringing Scheme is concerned with the analysis and collation of reports (recoveries) of ringed birds, both healthy wild birds and those ringed following rehabilitation. Recoveries include both live recaptures and birds found dead. Details of ringing: species, age, sex, date (day, month, year), place of ringing (latitude and longitude) and details for finding; finding condition (dead, alive, alive and released etc), finding circumstances (natural causes, predation, shot etc) and finding status (unknown, nesting, roost, colony etc) are recorded. From these it is possible to assess the distance travelled and the duration (time elapsed) between ringing and finding.

3. After removal of records deemed inaccurate, a total of 1,483 recoveries of dead adults or juvenile birds were available from the BTO National Ringing database for rehabilitated birds (RH). These covered 73 species of which only 17 had more than 10 recoveries. Mute Swan, Guillemot and Kestrel had more than 100 recoveries each. Time elapsed and distance travelled from ringing to recovery were compared between RH and non-rehabilitated recoveries (non-RH). The two datasets were mutually exclusive. Sample sizes for non-RH birds were considerably larger than RH samples.

4. Analyses focused on 11 selected species with more than 20 RH recoveries each, these were: Barn Owl, Buzzard, Gannet, Guillemot, Herring Gull, Kestrel, Little Owl, Mallard, Mute Swan, Sparrowhawk and Tawny Owl. Where RH sample sizes for a species were less than 20 birds, these species were placed in seven groups of ecologically similar species. The groups were: corvids, seabirds, wildfowl, raptors, herons, passerines and gulls. In addition to the 11 species comparisons, RH recoveries for Common Scoter were compared with non-RH for eider (an ecologically similar species), as there were only eight non-RH recoveries for Common Scoter.

5. Mann-Whitney U tests showed that there were significant differences in median time elapsed and distance travelled between RH and non-RH birds for the majority of the 11 species. Median time elapsed and distance travelled was greater for non-RH recoveries. Buzzard was the only species not to show a significant difference in median time elapsed and distance travelled. The seven of species groups also showed a significant difference in median time elapsed between RH and non-RH recoveries. Non-RH recoveries had a longer time elapsed between ringing and finding. Less clear cut differences existed between RH and non-RH for median distance, but differences were most pronounced for seabirds and wildfowl.

6. The significance of the differences in distribution of recoveries against time elapsed between RH and non-RH birds were tested using the Kolmogorov-Smirnov 2-sample test. This was conducted for three time periods of recoveries:
(1) all recoveries;
(2) all recoveries excluding the first two months and
(3) those within the first year only.
For all recoveries within species comparisons, large differences between RH and non-RH birds occur within the first year, especially the first two months post-release (highest number for RH birds) and differences are significant for all species except Buzzard, Mallard, juvenile Kestrel and juvenile Tawny Owl. Beyond two months, the distribution is more similar between RH and non-RH bird. At least 90% of the RH recoveries for Gannet, Guillemot, Common Scoter, Little Owl and adult Barn Owl occur within the first year.

7. Where a species sample size was sufficient, separate analysis of juveniles and adults showed recoveries of RH were broadly similar between the two. However, despite the rehabilitation success of adults and juveniles being similar, the maximum difference between RH and non-RH occurred with adults as non-RH juveniles had a greater percentage of rapid recoveries than adults. This was especially true for Barn Owl, Tawny Owl and Kestrel. Gannet, Guillemot and Little Owl have approximately twice the number of recoveries within the first year for RH compared to non-RH, as do adult and juvenile Herring Gull, adult Kestrel and adult Tawny Owl. Species with little difference between RH and non-RH include Sparrowhawk, Mallard and Mute Swan. There was also little difference between RH and non-RH for juvenile Barn Owl and juvenile Kestrel. Relationships are similar when the first two months are excluded, except that significance of the differences between RH and non-RH are reduced, although the difference for Guillemot is still apparent.

8. Within the first year only, recoveries of RH and non-RH are similar for Mallard, Kestrel, Tawny Owl and Mute Swan. In contrast, recoveries of RH are higher compared to non-RH recoveries for Buzzard, Gannet, Guillemot, Common Scoter, Herring Gull, Little Owl, Sparrowhawk and Barn Owl, especially within the first two months. The percent of recoveries for RH birds in the first year is three times that of non-RH recoveries for Guillemot, with nearly 90% of the recoveries in the first year occurring within the first month.

9. Small sample sizes for RH recoveries relative to non-RH recoveries for the species groups meant analysis of percentage of recoveries with time elapsed was restricted to adults and juveniles combined. For these groups of species, similar results were obtained when considering all recoveries and exclusion of those recovered within the first two months. All groups of species had significant differences (Kolmogorov-Smirnov 2-sample test) in cumulative percentage of recoveries with time elapsed when considering all recoveries. Most of the RH recoveries occur within the first year and the difference in percentage of recoveries between RH and non-RH birds is greatest within the first two months. Approximately 90% of the total recoveries for RH for corvids, herons, passerines, raptors and seabirds occur within the first year, the difference between RH and non-RH being most marked in seabirds and corvids. Wildfowl and herons show the least difference between RH and non-RH recoveries.

10. The vastly different sample sizes between RH and non-RH recoveries, especially for the species groups, suggests that caution should be used in interpreting the results. However, for some of the species and groups where differences between RH and non-RH are large, these differences are likely to be ‘real’ and act as a useful indicator of the post-release survival of rehabilitated birds.

11. The report has shown that for some species, such as Guillemot and Common Scoter, post release survival is so low that the majority of the rehabilitated individuals are unlikely to re-enter the wild population. While for other species such as Mute Swan a high percentage enter the wild population. Birds can only be classed as ‘successfully rehabilitated’ in the long term if they re-enter the population and have a similar chance of survival to wild birds.

Back to Research Reports 289 - 344