Crag Martin neontology complements taphonomy at the Gorham's Cave Complex

We carried out monitoring of ECM wintering at a roost in Gibraltar that consists of a series of caves alongside each other at sea level, primarily during the autumn–winter period of 2019-2020. The monitoring consisted of weekly counts of birds returning to roost, and of regular ringing and measuring sessions. The ringing data for 2019–2020 were augmented with data collected at the site between 2016 and 2018.

The eight caves at the site are all micro-sites within a single ECM roost and three of them, which lie just above the current sea level and are the only ones accessible from land, were studied: Gorham’s Cave (36° 07′ 13.86″ N 5° 20′ 32.57″ W UTM 30 N 289190.5 3999856.5), Vanguard Cave (36° 07′ 18.89″ N, 005° 20′ 31.62″ W UTM 30 N 289218.0 4000011.0) and Cave F (36° 07′ 19.8″ N, 005° 20′ 30.088″ W UTM 30 N 289257.0 4000038.1) (Figs. 4 and 5).

Twenty-three weekly evening counts were conducted of birds returning to the roosting site during the period 4th October, 2019 to 12th March, 2020, from a fixed point overlooking the study site. We attempted to space these evenly in time, but adjustments were made due to unfavorable weather (mean number of days between counts = 6.96 ± 1.45 SE). The approach of the birds as they return to roost is described elsewhere¹². It occurs along a fixed trajectory and our vantage point optimized the viewing of these movements. All ECM returned to the site along a common trajectory and birds only broke up and headed towards the different caves once at the site, so that in principle, every bird had equivalent opportunities to access each cave on arrival to the site.

We used a combination of the results of the counts and the Schnabel Index for mark-release-recapture data from a series of dates³³ to estimate the total roosting population size of ECM at the site during the 2019–2020 period. The latter was achieved by estimating the number of birds roosting at each cave and then combining these for a total population size, although we recognize that birds also use other micro-sites⁷; (pers. obs.). Due to differences in sample sizes of birds recaptured, 95% confidence limits for the estimated roosting population size at each cave were drawn from the t-distribution for Gorham’s Cave and the Poisson distribution for Vanguard Cave and Cave F³³.

Trapping and ringing were carried out at the three caves at least once a week. All licences required under the laws of Gibraltar were obtained and protocols were approved by the Ethics Committee of the University of Gibraltar. Ringing and handling of birds was carried out under the auspices of the Gibraltar Ornithological & Natural History Society (GONHS), which carries out its bird ringing under licence from the Ministry for the Environment, HM Government of Gibraltar, under the 1991 Nature Protection Act. Gibraltar-based ringers are licensed by the British Trust for Ornithology (BTO), and we adhered closely to the technical and ethical standards of the BTO for handling and ringing birds³⁴. Routinely, birds are released without ringing when their condition is poor. One bird was captured in a condition that was too poor for it to be ringed. The reporting recommendations of the ARRIVE guidelines³⁵ were followed.

The majority of the data used in this study were collected between October 29th 2019 and March 4th 2020. In addition, trapping and ringing had taken place intermittently at Vanguard and Cave F during the winter since 2016, and trapping took place at the site throughout autumn 2020. We used the BTO A-sized rings, in accordance with guidelines for other European hirundines³⁴. Due to the different dimensions of the caves, we used different mist net sizes at each one. A 6m-length net was used at Vanguard Cave, 12 m and 3 m nets at Cave F, and 3 × 6 m nets mounted vertically on triple high poles at Gorham’s Cave.

The number of trapping sessions, and the range of dates of these at each cave during the 2019–2020 autumn-winter season, was: 10 Gorham’s Cave (29/10/2019–04/03/2020; mean number of days between sessions 14.11 ± 2.23 SE), 11 Vanguard Cave (13/11/2019–04/03/2020; mean number of days between sessions 11.20 ± 1.81 SE), 11 Cave F (13/11/2019–04/03/2020; mean number of days between sessions 11.20 ± 1.81 SE). Seven extra trapping sessions took place at Vanguard Cave and Cave F before the 2019-2020 autumn-winter season, on: 01/28/2016, 02/16/2016, 02/13/2018, 02/21/2018, 12/04/2018, 01/08/2019 and 02/21/2019. There were eight additional trapping sessions during the autumn of 2020, on: 10/29/2020, 11/02/2020, 11/12/2020, 11/15/2020, 11/19/2020, 11/24/2020, 12/02/2020 and 12/03/2020. 1511 different birds were processed between 2016–2020, of which 156 were captured at least twice. 796 individuals were processed during the 2019–2020 autumn-winter season, the period for which most of our analyses are based: 369 at Gorham’s Cave, 221 at Vanguard Cave and 206 at Cave F. Of the birds recaptured that had been ringed at the site during previous seasons, eighteen were from the 2019–2020 season (ten ringed at Gorham’s Cave, two at Vanguard Cave, seven at Cave F), fifteen were from the 2018–2019 season (eight at Vanguard Cave, six at Cave F), four were from the 2017–2018 season (three at Vanguard Cave, one at Cave F), and one was from the 2015–2016 season (from either Vanguard Cave or Cave F; unspecified and excluded from the analysis). A bird was recaptured that had been ringed elsewhere in Gibraltar (the GONHS Jews’ Gate Field Centre) on the 14/01/2014, 2233 days before it was captured again on the 25/02/2020.

Biometric measurement of all birds was carried out by a single person (CP) in order to maximize consistency. We followed the standard processing procedure of the BTO³⁴, which includes recording the weight of birds in grams (g) to 0.1 g and length of wing in millimeters (mm) to 0.5 mm. Birds were aged whenever this was possible but ageing of ECM became increasingly difficult towards the end of the winter period, increasing the possibility of confusion with adults⁹. For this reason, age was excluded from most of the analyses. Birds could not be sexed because sexes are similar in appearance, including size^4,36. We captured birds only during the evening, to ensure that condition of birds was not a factor of weight-loss whilst roosting, since ECM at the site are known to weigh less during mornings than the evenings¹³. Birds captured were roosted in boxes and released at the site the following morning.

Although Elkins & Etheridge¹² assumed that movement of birds between different parts of the roost at Gibraltar is considerable, this was never tested. The proximity of different parts of the roost from each other means that all micro-sites are potentially equally accessible to ECM using the site. It is expected that they should be able to use micro-sites interchangeably, given especially their approach during evenings along a fixed narrow route. Any fidelity to micro-sites must thus be explained by factors other than distance between individual micro-sites. The multiple cavities at the roosting site, and the ease with which we were able to access these, allowed a unique opportunity to test whether individual birds repeatedly used the same micro-sites within the roost, both within and between winters. We used the data gathered to test the following hypotheses: (1) that a degree of fidelity to different spaces within the roost (‘micro-sites’) exists among ECM, with individuals more likely to be recaptured at the same cave than in a different cave, (2) that any fidelity observed will translate to a difference in quality of roosting sites, as indicated by differences in condition of birds according to micro-site, and (3) that the incidence of recapture should be highest at the cave at which birds are in the best condition.

Statistical analyses followed Sokal & Rohlf³⁷ and were carried out on SPSS statistical software (IBM). We used a binomial Z test to analyze whether recaptured birds that were initially ringed during the 2019–2020 season were returning to the cave where they were first trapped/ringed, (1) within the 2019–2020 season and (2) between this and separate seasons. We also used a 3 × 2 Fisher’s exact test to test for differences, between caves, in the frequency with which birds ringed at one cave were captured at another. Multiple recaptures of birds were excluded from all of these analyses on fidelity in order to avoid bias.

We explored the relationship between wing length and weight using linear regression analysis, to control for the possible effect of body size on weight—on the basis that wing length provides a good measure of body size in passerines³⁸—using only data collected during the 2019–2020 season. For individual birds that were trapped more than once, we used wing length and weight on the date of first capture. We then grouped, by cave, the residuals of the regression and used a one-way ANOVA to explore differences in mean condition of birds between caves, with condition expressed as the relationship between wing length and weight. We also used linear regression to explore the relationship between daily recapture rate at all caves and the number of days from the first day of trapping at each cave, with the latter as the explanatory factor. Again, we segregated the residuals of the regression by cave and used a one-way ANOVA to explore differences in recapture rates between caves. We used Pearson’s chi-squared test with Yates’s correction for small sample sizes³⁹ to explore differences in the likelihood of recapture of birds on more than one occasion at each cave. Because differences in weight and wing length have been recorded between adult and juvenile ECM in Gibraltar¹³, we used Pearson’s chi-squared test to explore the relationship between age and use of the different micro-sites for all the birds that we were able to age (n = 395 of 796 birds processed), to see whether this was consistent with our other findings.

Source: Ecology - nature.com