species snapshot

s part of my PhD I am collecting the second full season of repro- ductive data at Blue Hill Escape Ain the Kouga mountains of the Western Cape. The broad overview of my research comes down to one question: why are populations of the Cape an unidentified thorn penetrated its ab- Chaetops frenatus declining, mostly in domen, and one nest failed because the areas where its habitat is warming? While male parent disappeared and the lone I am also looking at the physiology, be- female could not keep up with feeding haviour and genetics of this endemic and and nest maintenance. With only one Near Threatened species, it seems likely parent to perform nest-care duties, this that the main reason is as a result of its nest was overrun by ants on a hot day. low reproductive success. The nestlings, which were only 10 days The 2017 breeding season proved dis- into their 21-day nestling period, tried michael mason astrous for Cape . There was to throw themselves out of the nest to es- an inordinate degree of nest failure, pos- cape the ants and did not survive. When sibly as a result of the continuing drought both parents are in attendance, occasion- or early heatwaves (or both). Out of a ally with an additional adult helper, they potential 24 fledglings from 14 nests are able to share not only provisioning that reached the two-egg stage, our team but also ‘cleaning’, and they will often eat above Weighing the slightly older nestlings recorded only one successful fledging. any nearby ants to avoid just such an out- could be nerve-wracking; this one was 15 days rocky road While it is accepted that ground-nesting come, while also gaining a tasty snack. old and already a bit feisty and active. By the Cape Rockjumper have high levels of nest failure, 95.8 end of day 16 we stopped approaching the per cent is extreme. e also collected footage of a nest, as the nestlings were so active we wor- My research was initially intended to fair amount of direct preda- ried about them fledging prematurely. Text & photographs krista oswald examine whether declining rockjumper tion. Although we had one in- populations were as a result of its inabil- stanceW of a nest being raided by a honey top, left Like many altricial birds, nestling I watched as a male Cape Rockjumper finally showed himself, calling and standing proudly on a prominent ity to maintain provisioning rates during badger and a few by grey mongooses, by Cape Rockjumpers seem to have a face only rock along a montane fynbos ridge, his elegant white malar stripe pointed directly at me. Worried that high temperatures, but after 2017 the fo- far the main culprits were . We a mother could love. This nestling was esti- cus has necessitated a slight shift. Provi- began erecting trail cameras opportunis- mated to be about two days old. the slightest move would send him skittering over the edge, I froze, binoculars propped on my knees and sioning data are being collected by way of tically at nests in 2015, before the start of a sharp rock digging into my back as usual. When a female followed him down to a tiny rock and ducked high-quality video cameras placed at ac- my current study, and I had seen footage top, right A Cape Rockjumper nestling about tive nests supporting nestlings at different of a boomslang preying on a rockjumper underneath, I knew they had to be at a nest. 15 days old. After being filmed on day 15, nest- stages of development, but such data are nest in that season. At the time I assumed lings were ringed and not monitored again I quickly made my way through the fynbos to check under the rock and had two surprises. Firstly, there obviously hard to collect when nestlings this was unusual and had no idea how until after they had fledged. were three eggs (until then I had only seen nests of two), and secondly, one was slightly damp, indicating are not surviving long enough to be filmed. prevalent these predation events were. Nest failure came about for different In 2017, as a result of equipment fail- opposite Young birds are supplementarily that the female had just laid. reasons; one nestling (the sibling of the ure, we only obtained definitive footage fed by the resident adults for up to three I thought yet again how lucky I am to call what I do ‘work’. one fledgling that survived) died when of predation at 11 of the 14 nests. Of > months after fledging.

50 african birdlife march/april 2019 cape rockjumper 51 next generation. Admittedly it would not take much to improve on the results from the previous season, and the presence of at least four fledglings at the time of writ- ing already shows an advance. Hopes are high that this season’s crop of Cape Rockjumpers at Blue Hill will boost the numbers back to the level of that in Janu- ary 2015.

t is possible that cool (or downright Whenever possible we the early stages, once the eggs are laid above After filming on days 15 and 16, we cold) temperatures early this season catch and ring the rock- the parents tend to remain, due to high fitted the nestlings with a SAFRING ring and may have limited the activity of some female investment in egg production. As a unique three-colour combination and only jumpers, using unique col- ofI the ’s prey items, such as midday soon as there are nestlings, it seems the returned to check the nest after day 20. This agamas and crag lizards, and a shortage our combinations so we adults do not abandon the nest unless the fledgling provided me with some of the of food may compromise the birds’ suc- can accurately account youngsters have gone. cutest footage of the project. cess. As a ground-nesting species, rock- The main rockjumper breeding pair in for re-nesting attempts jumpers face many threats during their each territory share parental duties: both above, left Snow early in the 2018 breeding breeding cycle, especially given their and overall reproductive male and female incubate, brood and season was a disaster for nests that were long incubation and nestling periods; ac- success. It also enables us provision. Occasionally young from the still at the egg stage, but in this instance the cording to a previous study, both average to track dispersal move- previous season will stay to assist, with eggs hatched before the snows fell in early 21 days. some males remaining as helpers for an September. The three adults – one female ments between territories Despite this encouraging start, I have additional two seasons before moving (pictured) and two males – successfully raised already recorded two nests being preyed away to establish their own territory. De- and fledged the two nestlings. upon by grey mongooses. It seems Cape became unviable as a result of the low spite this, most of the territories I moni- Rockjumpers may go from facing mon- temperatures. tor have only the one main breeding pair out) that Cape Rockjumpers lay two-egg goose predation early in the season when Some studies suggest that, especially and rarely have any helpers. This could be clutches. Larger clutches may be result- michael mason it is still cool, to snake (and ongoing in long-lived species, it is more advan- due to a change in overall population lev- ing from higher food availability due to A male Cape Rockjumper in a typical alert po- Despite the generally held perception mongoose) predation late in the season tageous for individuals to prioritise els, since in 2015 when I arrived many of more precipitation in 2018. As with many sition mid-call, possibly either defending his that occur only in trees, they when it is hot. Spending 42 days on the their own preservation over that of their the territories we monitored comprised birds, it seems the rockjumpers will keep territory or communicating some unknown seem quite at home in montane fynbos, ground in an area with many such op- young. On a warm day (which for rock- three to five individuals. breeding as long as food is available and rockjumper message to others in the vicinity. where the tallest plant is at most a me- portunistic predators is a perilous affair. jumpers means above about 15 °C) they Rockjumpers begin breeding as early the female is capable of producing more tre high. While snake predation on bird This season I also had the first record- may not return to incubate for hours, as late July and the season continues into eggs. those, seven showed boomslang pre- nests is well known, we were unaware ed instance of nest failure in this species relying on the ambient warmth of the January, so by mid-September the birds are Whenever possible we catch and ring dation and it only ever occurred after that boomslangs are active ground-nest as a consequence of extreme weather. insulating nest material to help eggs re- in various stages of the cycle: some territo- the rockjumpers, using unique colour the chicks had hatched. Two of these hunters and had assumed that rockjump- While the rockjumpers generally build tain heat. Leaving the nest unattended ries have nests that failed, some pairs have combinations so we can accurately ac- predation events were caught on our er nests would be more vulnerable to their nests under rock overhangs fac- for so long seems to be a risky strategy, nestlings ready to fledge, while others are count for re-nesting attempts and overall high-quality video cameras, leaving us mammal predation. It is interesting that ing downslope, two did not provide ad- although in most territories it seems that starting to re-build and try again. reproductive success. It also enables us with stunning footage of exactly how a during the non-breeding season rock- equate cover for the intense snowstorm at least one of the parents keeps the nest One pair in particular surprised me by to track dispersal movements between boomslang repeatedly checks the nest to jumpers have been observed reasonably that occurred in early September and within sight, as evidenced by the alarm- laying their second three-egg clutch for territories; for example, a young male ensure that there are no survivors. Adult close to boomslangs with nary an alarm trail camera images show that the eggs calling whenever I approached the nest. the year after the first was one of those ringed as a juvenile in 2014 is now domi- rockjumpers, aware of what is happening call being given. were covered in snow for more than 24 The likelihood of rockjumpers aban- covered by snow. A pair in a nearby ter- nant in a nearby territory. Currently, the but helpless to deter a two-metre-long Despite a massive, unseasonal snow- hours. The parents were unable to ac- doning a nest decreases exponentially ritory also produced a three-egg clutch, oldest rockjumper at Blue Hill is a male venomous snake, alarm call and display storm in early September 2018, the rock- cess the eggs properly in order to keep depending on the stage. While the spe- which is interesting as the literature ringed as an adult in 2013, making him at in the background to no effect. jumpers were intent on producing the them warm and it is likely the clutch cies can be prone to deserting nests in states (and my previous experience bears least six years old. >

52 african birdlife march/april 2019 cape rockjumper 53 michael mason n a positive note, recent fires was predation (by boomslang in par­ appear to have increased ticular), but data from one season are suitable rockjumper habitat. not sufficient to draw any firm conclu- OStudies have shown that the species sions. Hopefully, a second full season does best about three years post-fire and of data will help determine whether the it has not taken the birds long to move predation of 2017 was normal or simply into the newly burned areas around a combina­tion of circumstances lining Blue Hill. While the birds are ‘territo- up against this iconic South African rial’ in the broader sense of the word, endemic.  their boundaries and group composi- tion seem to shift at a far greater level Acknowledgements than previously thought. On a prac- This research occurs at Blue Hill Escape tical level, that means rockjumpers are with the cooperation of the Lee family now coming within 200 metres of the and is done in collaboration with Drs Ben farmstead; this not only makes my job Smit, Alan Lee and Susie Cunningham, easier as I have far less distance to carry along with help from many volunteers. camera setups, but we may be close to Financial support comes from the NRF, having rockjumpers in the backyard! Tygerberg Bird Club and the Be- The increased probability of both havior Society. fire and drought due to future climate change scenarios will be a mixed bag for References rockjumpers. Whereas a more regular Holmes RT, Frauenknecht BD, Du Plessis above One of our provisioning cameras fire regime across the fynbos will in all MA. 2002. ‘Breeding system of the Cape caught this image of a boomslang preying likelihood increase their habitat, if more Rockjumper, a South African fynbos upon seven-day-old nestlings. frequent droughts result in increased endemic.’ The Condor 104: 188‒192. snake activity, their overall population Lee AT, Barnard P. 2016. ‘Endemic birds of top A male on his way back to provision his may continue to decrease. the Fynbos biome: a conservation assess- young brood warily surveys the area around Our results from 2017 suggested the ment and impacts of climate change.’ Bird the nest . main threat to rockjumper reproduction Conservation International 26: 52‒68.

54 african birdlife