|Specific Name: Eudyptes chrysocome (Southern Rockhopper), Eudyptes Filholi (Eastern), Eudyptes moseleyi (Northern)|
|Adult Height: 45-58cm|
|Adult Weight: 2.5-3.5kg|
|Adult Flipper Length: 15-19cm|
|Estimated Population: 2 Million breeding pairs|
Rockhopper penguins, which are the smallest of the Eudyptid penguins, breed on many subantarctic and southern temperate zone islands. The latter include many of the southerly Islands of Chile and Argentina.
Species / Subspecies Distinctions
Rockhoppers are presently split into 2 or 3 different species. These are the Southern (Short-crested) Rockhopper (E. chrysocome), Eastern Rockhopper (E. filholi (or E. c. filholi (if merely considered subspecies of Southern)) and Northern (Long-crested) Rockhopper (E. moseleyi). Populations of the different species are found in (i) Falkland isl. and Cape Horn Archipelago, (ii) P Edward, Marion, Crozet, Kerguelen, Macquarie, Heard, Auckland, Campbell, Antipodes and Bounty islands, and (iii) Tristan da Cunha, Gough, Amsterdam & St Paul islands, respectively, as indicated on the map. This page covers all three species, although an additional page focuses specifically on the Northern Rockhopper.
Separation into Northern and Southern species has been long suggested based on differing crest and song characteristics, both of which are relevant to breeding (Jouventin 1982. Visual and Vocal Signs in Penguins, their Evolution and Adaptive Characters. P. Parey Ed. Berlin). Genetic studies by the same author based on differences in the mitochondrial control region and ND2 genes between Rockhoppers from N and E (referred to as S in paper as S/E populations considered to be same), confirmed that at least a N and a S population could be distinguished (Jouventin et al., 2006. Mol. Ecol. 15, p.3413-3423). The species were estimated to have split as little as 140000 years ago, but more certainly less than 1 Million years ago (inaccuracy as rate of DNA mutation upon which calculation is based can only be estimated).
The further separation into 3 species was based on analysis of differences in sequences in the small ribosomal subunit (12S), cytochrome B, and cytochrome oxidase subunit 1 genes between populations from N, S and E areas (Banks et al., 2006. Polar Biol. 30, p.61-67). These differences were considered sufficient to merit separation into three species. The N species is genetically less related to the S and E populations than the latter pair to each other. In fact, the N and S genetic divergence was higher than that between Galapagos and Humboldt penguins (two recognized species) in two out of the three genes analysed. The Northern species is also easily distinguishable from the other two based on morphological features, whereas S and E species are visually indistinguishable. For example, the Northern Rockhopper has a far longer crest, a larger area of black plumage on the tip of the underside of the flipper and a thick black leading edge of the flipper (the other species have a thin grey leading edge on the underside of the flipper). Northern Rockhoppers are also marginally longer bills and flippers (Tremblay and Cherel 2003. Mar. Ecol. Prog. Ser. 251, p.279-297).
It should be noted that using a more classical definition of a species as a breeding population unable to fertively interbreed with other species, the Rockhoppers, and indeed probably all Eudyptid penguins would have to be grouped together, since successfully breeding apparent hybrids of e.g. Rockhopper and Macaroni Penguins have been reported on several occasions (White and Clausen 2002. Mar. Ornithol. 30, p.40-42).
Distribution and Population Sizes
Rockhopper penguin populations appear to be declining worldwide. Presently, the number of breeding pairs approximates to 800000 (Southern), 800000 (Eastern), and 350000 (Northern), although the figures may be inaccurate, since populations are not monitored yearly, and in numerous colonies significant population fluctuations have been recorded in recent years.
An important breeding site for the Southern Rockhopper is Isla Noir (Chile) with numbers of nests in 2005 estimated at 158200 (Oehler et al., 2008. Wilson J. Ornithol. 120(3), p.575-581). This level is higher than previous figures and is in line with reported increases of Rockhopper numbers at other Chilean / Argentinean colonies. However, the same author reported a complete disappearance of the historically recorded colonies on Isla Recalada (Oehler et al., 2007. Wilson J. Ornithol. 119(3), p.502-506). Two other main Chilean breeding sites are the Ildefonso and Diego Ramirez Islands at the Southern tip of the South American continent, with approx. 85000 and 130000 breeding pairs in 2002, respectively (Kirkwood et al., 2007. Waterbirds 30(2), p.259-267). Large numbers of Southern Rockhoppers are also found further to the E on Argentinian Ano Nuevo and Staten Islands, with the latter having over 165000 breeding pairs in 1999 (Schiavini 2000. Waterbirds 23(2), p.286-291), making it the probably largest of all S. Rockhopper breeding sites.
Numbers on the Falklands appeared to be stable at census sites between 1996 and 2001, with about 285000 breeding pairs in the latter year (Tremblay and Cherel, 2003. Mar. Ecol. Prog. Ser. 251, p.279-297), although further declines have been recently reported (Huin 2004. Wildlife Conserv. Falkl. Isl. 4, p.15-16) and a 2005/06 census only counted 210000 breeding pairs. The largest colonies are found on the Jason Islands to the NW and Beauchene Island to the S, with 150000 and 75000 breeding pairs respectively in 95/96. The 2005/06 Falklands census suggests that numbers at the Jason Islands had fallen to about 70000. At least 1.5 Million breeding pairs existed on the Falklands in 1933 (Puetz et al., 2003. Waterbirds 26(2), p.169-175), meaning that the population has dropped by 80% since then. The cause(s) of this decline, which was most rapid in the 1980s and 1990s and is presently continuing at a slower rate, is a subject of much debate. It has been at least partly attributed by some authors to the effect of large-scale commercial fisheries operating in Falkland Island waters (Bingham 2002. Rev. Chilena de Hist. Nat. 75(4), p.805-818). However, others argue that there is no direct evidence of competition with commercial fisheries. It is likely that multiple factors are involved, with the most dramatic annual declines resulting from climatic anomalies (see threats section).
An extremely detailed overview of significant Southern Rockhopper breeding sites is provided by Bingham and Mejias, 1999 (Sci. Mar., 63(Suppl.1), p.485-493).
Numbers of Eastern Rockhoppers are in decline at Marion Island according to comparison of 2008 figures with those of 1995 (Crawford et al., 2008. Afr. J. Mar. Sci. 30(1), p.185-188). Numbers at Campbell Island are considered to have dropped by 94% since the 1940s, with only about 100000 breeding pairs left in 1985. Interestingly, the decline was strongest in the 1945-55 period, when sea temperatures were high, whereas a small temporary increase at one colony was recorded in the 1960s when temperatures were temporarily lower (Cunningham and Moors, 1994. Emu 94(1), p.27-36). The populations at Campbell and Antipodes Islands apparently continue to fall (Thomson and Sagar, 2002. Water & Atmosphere 10(3), p.10-12).
Northern Rockhopper populations at Gough were estimated at 145000 in 2000 and were considered stable at the time (Cuthbert and Sommer 2004. Mar. Ornithol. 32(1), p.97-103). However, 2009 press reports citing RSPB sources suggest that numbers at Gough and Tristan are now in significant decline. A 1995 census at Amsterdam and St. Paul Islands showed that numbers on Amsterdam Island had declined compared to 1971 (58000 down to 25000 breeding pairs), whereas numbers on St. Paul had increased (4000 up to 9000 breeding pairs), possibly as a result of long-term recovery from exploitation in the 1930s (Guinard et al., 1998. Col. Waterbirds 21(2), p.222-228).
The diet and foraging strategies of Rockhopper penguins have been studied at a wide variety of sites. This has revealed broadly similar foraging strategies with certain adaptations to the local environment. Diet varies not only with location but also temporally due to seasonal changes in relative and total prey availability. Further, significant interannual changes in diet have been recorded at several sites, correlating to yearly fluctuations in prey biomass. Rockhopper foraging patterns can be studied in a variety of ways. Dietary studies are performed on breeding birds returning to colonies and generally involve flushing the stomach contents of birds and analysing their composition. Further, depth-loggers or satellite-tracking devices can be attached to the backs of penguins to provide data on diving behaviour and foraging ranges.
Rockhopper penguin diets are highly variable and demonstrate that these penguins are opportunist feeders. Dietary studies conducted during the chick-rearing period at various sites on the Falklands from 1986/87 to 1998/99 showed average proportions of diet (by mass) to be 53% crustaceans, 29% cephalopods (e.g. squid), and 18% fish during the breeding season. Penguins at W Falkland sites appeared to consume largely squid and crustaceans, those at the NW (Seal Bay) consumed negligible amounts of squid, but from nearly 100% crustaceans in 97/98 to nearly 90% fish in 98/99 (Pütz et al., 2001. Polar Biol. 24, p.793-807). Studies at Staten Island (Argentina) of females returning to feed chicks during the brood period in the 99/00 to 01/02 seasons revealed mean stomach contents of about 160 grams with significant interannual fluctuations (131 g in 99/00, 203 g in 00/01). The average diet consisted (by mass) of about 25% crustaceans (T. gregaria 10%, E. vallentini (8.5%), and T. gaudichaudii (4.5%) and over 75% cephalopods (L. gahi (27.5%), E. megalocyathus (21.5%), G. antarcticus (17%), and M. ingens (10%)). A variety of fish were consumed in small numbers (although these may be underestimated as fish are rapidly digested), with H. bispinnis being by far the most common. If prey is analysed by number as opposed to mass, the relative contribution of the smaller species increases. For example, 85% of prey items are crustaceans, but only 9% are cephalopods. In contrast, by number, the squid L. gahi makes up less than 0.5% of the diet, whilst it is the main prey component by mass. In 99/00, prey items were generally more digested, suggesting that birds may have had to forage over larger distances.
At Macquarie Island, Rockhopper stomach contents were analysed in the 93/94 to 95/96 breeding seasons at Brother Point on the E coast (Hull 1999, J. Zool (London) 247, p.507-529). Birds returning to the colony before breeding had little stomach contents and only slightly more was brought ashore during the incubation period. Once chicks had hatched birds started returning with full stomachs with an average of about 160 g (max. 485 g) per trip during the guard and creche phases. The average meal was composed of 60% (by weight) Euphausiids (mainly E. vallentini), 25% fish (mainly Krefftichthys anderssoni), and small amounts of crustaceans and cephalopods. Compared to nearby-nesting Royal Penguins, the Rockhoppers consumed twice as much Euphasiids and half the amount of fish (it is known that larger crested penguins tend to consume more fish and a generally larger variety of prey). Of the fish consumed, a greater proportion were benthic, suggesting that Rockhoppers were foraging nearer to shore.
The diet of E Rockhopper penguins has been studied at Kerguelen and Crozet Islands in 98/99, and compared to the diet of the N Rockhopper at Amsterdam Island 96/97, located about 1000 km further North (Tremblay and Cheret 2003. Mar. Ecol. Prog. Ser. 251, p.279-297). Birds at Amsterdam Island fed mainly on crustaceans (43.7% by mass) and juvenile squid (43.9%), with fish making up 12.2% of the diet. This was in stark contrast to diets at Crozet (95.1% / 0.6% / 4.3%) and Kerguelen (97.1% / 0% / 2.9%). At Amsterdam Island, the euphausiid shrimp T. gregaria was the main prey by number (83.6%), whereas, E. vallentini dominated at Crozet (87.8%) and Kerguelen (82.5%). The proportion of squid increased over the study period at Amsterdam Island and the birds diving behaviour also showed a progressive trend towards more numerous but less deep dives. No significant dietary fluctuations were observed at the other sites.
The varied diet of N Rockhoppers at Amsterdam Island had also been recorded in an earlier study during the 94/95 breeding season (Tremblay et al., 1997. Polar Biol. 17, p.119-122). In Nov. 44% (by mass) squid, 31% crustaceans, and 25% fish were consumed. In contrast, in Dec. values of 15, 21 and 64%, respectively were obtained. Thus, there was a significant shift in diet from squid to fish within a relatively short space of time during the chick provisioning period. A corresponding shift from inshore to offshore foraging and towards shallower dives was noted and may have had an influence on the type of prey consumed. In contrast to the Amsterdam Island birds, Northern Rockhoppers at Gough feed predominantly (over 90% mass) of crustaceans (Klages et al., 1988. Ostrich 59, p.162-165).
Foraging Ranges / Diving Parameters
Foraging ranges and diving parameters have also been studied at various sites during the breeding season. A study of diving behaviour of female N. Rockhopper Penguins at Amsterdam Island provided compelling evidence for cooperative hunting behaviour in this species (Tremblay and Cherel 1999. Condor 101, p.179-185). Two females were observed diving synchronously over a period of 7 hours including 286 dives. Dives were generally initiated within seconds of each other, reached similar depths and terminated at similar times. About 92% of dives were synchronous during this period. The two birds apparently met at sea and did not leave or return to the colony together or dive in synchrony on several other days studied. It is however noted that birds often leave or return to shore in groups.
Using time/depth loggers the frequency and depth of dives can be established. At Amsterdam Island, diving of Northern Rockhoppers was studied during the guard phase in 1995 (Cherel et al., 1999. Mar. Biol. 134, p.375-385) and during the creche phase in 1994 (Tremblay et al., 1997. Polar Biol. 17, p.119-122). It is noted that figures from the two studies are not entirely comparable due to interannual variations. During the guard phase, in which only the female forages, birds were found to leave the colony at dawn, returning about 12 hours later. Transit speed to the foraging areas is approx. 7.4 km/h based on earlier studies (Brown 1987. J. Field. Ornithol. 58, p.118-125) and transit time was about 45 min either way. Most birds stayed within 6 km of the coast and performed about 550 dives. Mean dive depth was 18.4 m, but dives up to a depth of 109 m were recorded. Average dives lasted for 57 sec, with max. of 168 sec, and the mean surface time between dives was 21 sec. This means that birds spent nearly 70% of time at sea under water. The penguins usually descended at a rate of about 1.2 m/sec, foraged at varying speeds for an average of 28.4 sec within a narrow depth range ("bottom time"), and then ascended at 1 m/sec (3.6 km/h). The angle of descent was steeper during longer dives, thus increasing the rate of descent/ascent without changing swimming speed (which is kept at a constant, probably physiologically optimal level). The studied birds had average stomach contents of 123.6 g (58% crustaceans by mass). As the brood phase progressed, a trend towards more numerous shallow dives was noted. Of 29 foraging trips analysed, 2 involved birds remaining at sea overnight. Diving activity was however low and shallow at night. Since penguins are visual predators, maximum foraging depths are usually attained at around noon, when the sun is at its highest position. During the creche phase, birds at Amsterdam Island were observed diving deeper and shifting from a squid- to fish-dominated diet when the early and late creche phase was compared. Mean depths of 66 m were recorded, with a max. depth of 168 m being achieved by one female. Dive depths of males and females were generally comparable.
In a further study (Tremblay and Cherel 2003. Mar. Ecol. Prog. Ser. 251, p.279-297), Northern Rockhoppers at Amsterdam Island were compared to Eastern Rockhoppers at Mayes Island (Kerguelen) and Possession Isl. (Crozet) during the 1996 brood period. The birds at all sites were found to have shorter outward transit times than return transits, suggesting that they gradually move away from the colony as they forage. Overnight trips were observed at Amsterdam and Crozet, but not Kerguelen. Dive depth averages were 18.7, 33.9, 26.3 m, respectively (with 93, 104 and 85 m max. dive depths) with the deepest dives performed at the middle of the day at all sites. The average stomach contents were 107.5 g (Ams.), 150.6 g (Croz.) and 214.8 g (Kerg.) and the relative rates of chick growth varied accordingly. The amount of stomach contents corresponded to the relative levels of zooplankton biomass at the three sites.
Diving of Southern Rockhoppers has been studied at Staten Island, Argentina (Schiavini and Rey 2004. Mar. Ecol. Prog. Ser. 275, p.251-262). Female birds brooding chicks were studied in Dec. 2000 as they fed mainly on juvenile squid. Mean dive depths of 28.9 m (max. 113 m) and durations of 79 sec were recorded, with 30.4 foraging dives/h and 66% of time spent under water. Trips were significantly longer than in the above studies, with 30% longer day trips and 60% longer overnight trips (which were also more frequent: about 50% of trips as opposed to 10% in the above studies). It is considered that the longer daylight periods compared to at more temperate latitudes account for the significantly longer but slightly less intense foraging trips. In this study, no difference in the departure time of birds doing day trips or overnight trips was noted, in contrast to birds at Amsterdam or Crozet Islands. A further study on Southern Rockhoppers compared birds at Staten Island with those at Berkeley Sound in the E Falklands during both incubation and brood stages. A detailed table comparing the results with those of many of the above studies is provided (Pütz et al., 2006. Mar. Biol. 149, p.125-137). Incubation phase foraging trips were long and Falkland male birds tended to swim about 300 km (max. 500 km) NE, whereas Staten Isl. birds swam up to 900 km E or 550 km S. Interestingly, diving was most frequent around dawn and dusk in birds from Staten Island, and in male incubating birds from the Falklands. Only the female incubating birds from the Falklands showed the usual preference for midday. Presumably the birds were feeding on prey that is most accessible around nighttime. Brooding birds were generally found to be diving with a higher frequency and an increased proportional bottom time than incubating birds.
Interestingly, in 1998 it was observed that birds at Mayes Island (Crozet) perform two different types of foraging (Tremblay and Cherel 2000. Mar. Ecol. Prog. Ser. 204, p.257-267). Birds not only performed normal pelagic series of dives, but also performed benthic dives (i.e. dives at the sea-floor). The latter can be recognized as a series of dives to highly consistent depths (which could be correlated to the depth of the sea-floor at the study location). The benthic dives appeared to be targeting high concentrations of crustaceans that accumulate at the sea bed during the day, and consequently the higher the proportion of benthic dives the greater the stomach contents of returning birds. The dives involved higher ascent/descent rates, were generally deeper and involved a longer bottom time (mean of 66 v 53 sec) compared to pelagic dives in the same study. This type of diving is not possible at many sites since it requires the presence of shallow oceanic shelf regions around the breeding site.
After moulting, Rockhopper penguins usually spend extended periods at sea until returning to their colonies for the next breeding season. Satellite tracking of penguins from several colonies on the Falklands and from Staten Island (Argentina) has revealed wide dispersal, but also the location of preferred wintering grounds. Staten Island birds were tracked for an average of 50 days and reached a mean maximum distance from the colony of 442 km, with a maximum of over 2000 km, travelled on average at about 3 km/h. Most Rockhoppers foraged to the N of the colony off the NE coast of Tierra del Fuego, or E of the colony in the Burdwood Bank area, located S of the Falklands, although in both cases birds didn't tend to head directly to these areas but swam various patterns covering up to several thousand km before proceeding to the wintering grounds (Pütz et al., 2006. Polar Biol. 29, p.735-744). A number of birds also appeared to winter in the Drake Passage, with several reaching polar waters near the South Shetland Islands. Birds from the Falklands tend to stay on the Patagonian shelf with major concentrations to the NW of the Islands, about 50 km off the Argentinian coast near Puerto Deseado (about 400 km N of the preferred feeding grounds for the Staten Island penguins), whilst others travel S and join Staten Island birds at Burwood Bank (Pütz et al., 2002. Mar. Ecol. Prog. Ser. 240, p.273-284). Thus, preferred feeding areas depended on colony location, with birds from S Falklands colonies frequenting Burwood Bank and those from N colonies frequenting the areas to the W and NW. Interestingly, in 1999 in particular, a number of Falklands birds made repeated trips of several days, returning to the colony in between, before finally leaving about a month after the moult. This is possibly explained by a high local abundance of prey in that year, as signified by unusually high commercial catches of squid (Illex argentinus), which in turn suggest high levels of crustaceans, upon which they and the penguins prey.
It was calculated that in total about 90000 birds from the Falklands or Staten Island winter at Burwood Bank, and over 170000 off the NE coast of Tierra del Fuego.
Timing of Breeding
The breeding season commences as birds return to the colonies from their long winter foraging trips. This occurs at different times at different locations within the breeding range. Mass of the birds on arrival can be a critical determinant of subsequent breeding success (Crawford et al., 2007. Afr. J. Mar. Sci. 30(1), p.185-188; Raya Rey et al., 2007. Ibis 149, p.826-835). at different times throughout the wide breeding range of the penguin. Generally, the more N colonies in warmer locations start to breed earlier than those further S (approx. 10 days later for each degree C lower water temperature). However, other factors may also play a role, since there may be several week differences between different colonies on the Falklands or between the Falklands and the nearby S. American colonies.
Nest & Partner Selection
Male birds return to the colonies first to establish nest sites, whilst females return up to 10 days later. Interannual fidelity to nest site (>50%) and mates is considered high in Eudyptid penguins. The colonies are dense, usually in rocky areas largely stripped of vegetation, and often shared with Cormorants or Albatrosses. Simple nests are made using stones, twigs and other available materials and usually occupy areas near rocks or in hollows beneath them. However, nesting in tussock grass areas is also observed and may be more successful due to more protection from predation (C & R Cassady St. Clair. 1996. Oikos 77, p.459-466). The colony is usually reached by a single or small number of tracks leading up from the coast which may be extremely steep and require the penguins formidable climbing skills. The claws of the penguins may cause rock striations over time and it has been suggested that these could be used to analyse the sites of former colonies (Splettstoesser 1985. Arctic Alpine Res. 17(1), p.107-111).
Courtship and Copulation
Courtship involves advertising behaviour such as bowing and also the ecstatic display, in which all Eudyptid penguins point their heads skywards, extend their flippers and makes a series of cries with their bills wide open whilst swinging their heads from side to side. As courtship proceeds, the partners may perform the mutual display where both birds bow, quiver or perform synchronous forms of the ecstatic display.
Copulation occurs at the nest and can be instigated by either partner. The female may squat on the nest or the male may indicate willingness by beating the female with his flipper. The male then huddles up to the female, nibbles at her neck and pats her with one flipper. The male then proceeds to mount the female which lies with a raised head, and the male edges backwards towards the cloaca whilst beating the females flanks with his flippers. Both birds move their cloaca towards each other to make contact. The male holds its position briefly, stabilized by its flippers and often held by the neck by the female. The male then dismounts and takes a subdued hunched attitude. Both birds may then proceed to preen or simply rest for a short period.
Laying / Brood Reduction Mechanism
First eggs are usually laid at the end of October in the Falklands, at a time when eggs have already largely hatched at Tristan da Cunha. Rockhopper Penguins, like the other Eudyptid penguins species display an unusual form of brood reduction which involves the successive laying of two eggs over a period of about 4 days, wherein the first egg is smaller and usually hatches (if at all) after the later laid second egg. This reversed hatching asynchrony is most pronounced in the Macaroni, Royal and Erect-Crested Penguins, where the first (A) egg is generally displaced from the nest before laying of the second (B) egg, whilst the Rockhopper, along with the Snares and Fiordland Crested Penguins, displays a milder form of this behaviour, wherein both eggs are usually retained until hatching, but the advantage conferred by larger egg and earlier hatching thereof give the (B) chick a strong selective advantage, which along with preferential feeding of the larger chick usually leads to starvation of the A chick within days of hatching. Reports of Rockhopper Penguins sometimes laying 3 eggs have been largely discounted and attributed to adoption of additional eggs (Williams 1981. Emu 81(2), p.87) although some evidence for laying of 3 egg clutches in Macaroni and Rockhopper Penguins has been found (Gwynn 1993. Emu 93, p.287-290).
Recent studies at New Island (Falklands) have shown that both eggs are potentially equally viable in Rockhopper Penguins (Poisbleau et al., 2008. Polar Biol. 31, p.925-932). At the study site, (A) eggs had a mass of 91 g, compared to 116 g for (B) eggs, making a size difference of 28%. Incubation, which only commences after arrival of the (B) egg, took an average 32 days for (B) eggs, but 33 days for (A) eggs, when both eggs were present. Removal of the (B) egg from the nest after laying resulted in faster development of the (A) egg, which hatched after only 32 days. Conversely, removal of the (A) egg had no effect on incubation duration of the (B) egg. Hatching mass was also higher for (A) chicks derived from single egg nests compared to mixed (A+B) nests (76 v. 61 g). In contrast, (B) chick hatching mass was unaffected by the presence of (A) eggs during incubation (86 g, 37% heavier than A chicks in mixed nests). Since the (B) chick hatches first, is larger and likely to be fed even before hatching of the (A) chick, the B chick has a massive selective advantage. As a result, in mixed nests only 11% of (A) chicks survived for the first 10 days, whilst 85% of the (A) chicks without sibling survived. No overt sibling aggression is involved, but the smaller chick is simply unable to compete effectively for food from the returning parents and consequently starves. In some cases the (B) chick has been observed displacing the (A) egg from the nest before it hatches although this is considered accidental as the chick moves around in the nest trying to get into the most comfortable position under the brood patch. In total, only 2 of 114 unmanipulated nests observed gave rise to two chicks. In the absence of a sibling, (A) chicks had the same chance of reaching the end of the brood period as (B) chicks, suggesting that the lower hatching weight, per se, confers no significant disadvantage on the chick. Hence, in Southern Rockhoppers, the (A) egg/chick clearly has a potentially significant insurance value in case the (B) egg fails to hatch or the (B) egg or chick is lost due to e.g. predation.
The frequency of egg loss and level of utilization of the insurance value of the (A) egg has been studied at both New Island and Macquarie Isl. (C & R Cassady St. Clair. 1996. Oikos 77, p.459-466). Egg losses in general could be attributed to (i) predation, largely by Skuas, (ii) loss due to aggression between penguins, and (iii) movement of the (B) hatchling. Egg loss of (A) eggs peaked within the first 10 days of laying, in particular around the time of arrival of the (B) egg, and then after a few days of transiently low loss rates gradually increased again towards the time of hatching. Other studies had found most mortality around the hatching date. The anterior eggs are most vulnerable to Skua predation (the most common cause of egg loss) with the (A) eggs being statistically more likely to be lost in this way due to their tendency to be found in this position. Egg loss due to intraspecific aggression accounted for 18% of losses and displacement by siblings for a lesser number.
Unfortunately, in most cases where (B) eggs were lost, the (A) egg was no longer present and could provide no insurance value (97/110 nests). In about 10% of nests, no (B) egg was ever seen, in which case the (A) egg may be able to serve as insurance against failure to lay two eggs. The viability of the (A) egg in such single egg nests could however not be confirmed. Overall, hatching of (A) eggs was up to 10.4% on Macquarie and 4.2% on New Island. However, in nests where the (B) egg was lost or never appeared to be laid, (A) eggs served their insurance role in at best 33% of nests on Macquarie and 10% of nests on New Island. Hence, only in about a fifth of nests, the (A) egg may have prevented total clutch failure.
An interesting question is why the (B) egg hatches before the (A) egg. Incubation usually only commences after arrival of the (B) egg, so the (A) egg presumably remains developmentally inert in the first days after laying. During incubation, egg temperatures of up to 35'C may be reached with the (A) egg having a slightly lower (about 4'C less on average) and more variable temperature (Burger and Williams 1979. Auk 96(1), p.100-105). The rear egg contacts the broadest part of the brood patch and is less likely to be exposed if the parent adopts a semi-upright position. Studies at New Island (Falklands) showed the tendency of the smaller (A) eggs to occupy the thermally less favourable anterior position in the nest, which may explain the lower temperature observed (C Cassady St Clair 1996. J. Animal Ecol. 65, p.485-494). The higher temperature may largely account for the faster development of the (B) egg, since no significant differences in the gross composition of the two eggs were detected. (B) egg incubation periods were lengthened if these eggs were incubated together with larger artificial eggs. This study also found that eggs in manipulated nests with two (A) eggs needed nearly 33 days for hatching, with those in double (B) egg nests needing under 32 days. Further, eggs laid earlier in the season were generally slower to hatch. This can be explained by a later study at the same site which showed that the incubation temperature of the eggs increased from an average of about 20'C after laying of the (B) egg to 30 degrees at the end of the first week (in birds laying relatively early in the laying period), suggesting that the brood patch was not fully developed at the onset of incubation (C. Cassady St. Clair 1998. The Auk 115(2), p.478-482). The study also showed removal of the (A) egg did not inhibit brood patch development (discounting the hypothesis that the (A) egg serves to stimulate development of the brood patch), and further that incubation temperature was largely correlated to the time of laying, with birds laying later in the laying period immediately reaching higher incubation temperatures. This suggested that the brood patch develops largely in response to a social stimulus from the colony as a whole.
Division of Parental Duties
The division of labour between male and female birds follows clear patterns during the breeding season. A study over 3 seasons (93/94, 94/95 and 95/96) at Macquarie Island assessed arrival / departure times of parents and weight fluctuations in both parental birds and chicks over the whole breeding season (Hull et al., 2004. Polar Biol. 27, p.711-720). The male arrives first and is then joined by the female. Both birds remain on land (fasting) until the eggs have been laid which occurs about 1 month after male arrival. The male then departs on a long foraging trip (mean: 9 days) in which he may gain 40% in body weight (after having lost about 30% whilst on land in the preceding weeks). It is noted that this seems short for a male foraging trip, since other studies on colonies in the NE of the Falklands showed trips of 11-15 days, reaching to about 400 km N of the colony (Pütz et al., 2003. J. Avian Biol. 34, p.139-144). Upon return, the male takes up incubation duties and the female almost immediately goes on a nearly 2 week foraging trip in which it may gain 20% weight, making up for most of its 30% weight loss whilst on land. The female returns to land in time for hatching (so its trip length is largely determined by the time of return of the male) and whilst the male remains on the nest fasting during the brood/guard period (resulting in an about 20% mass loss), the female makes usually daily foraging trips to provision the chicks. The female also loses weight in spite of the fact that it is foraging, since the foraging effort is energetically demanding but most of the food caught is provided to the chick(s). Only once the chicks join creches do both parents start to forage. They may perform single or multi-day foraging trips during this period. To what extent they complement each other at this stage is still subject of debate.
Studies of E Rockhoppers at Kerguelen and Crozet using an automated system recording penguin movements to and from the colony, interestingly revealed that female birds feeding less than 5 day old chicks may actually sometimes perform multiple short trips during a single day (Tremblay and Cherel 2005. J. Avian Biol. 36, p.135-145). This can probably be explained by the female adapting its foraging to the limited stomach capacity of the small chick.
Development of Young Chicks
Chicks hatch with a thin coat of down and develop a thicker 2nd down (mesoptyl plumage) within 2-3 weeks. Young chicks are brooded but rapidly become too large to fit under it the parent. These chicks lie partially covered (usually with head and shoulders covered) by the parent or lean closely against it. The 2nd down is crucial for thermoemancipation of the chicks, and is a prerequisite for the transition from the guard to the creche phase. The water-proof juvenile plumage develops shortly before fledging. Chicks are more vulnerable to predation, in spite of parental care, during the guard stage than during the creche phase (Raya Rey et al., 2007. Ibis 149, p.826-845). In the creche phase, predation is usually focused on smaller chicks and most chicks are relatively safe from predation after 1-2 weeks of the creche phase.
Chicks in the 1993-96 study at Macquarie had a hatching weight of approx. 60-90 g (Hull et al., 2004. Polar Biol. 27, p.711-720). Their weight increased nearly linearly for the first about 70 days (reaching about 1 kg at the onset of the creche phase at about 23-26 days), after which a slight drop in weight was sometimes recorded. Mean fledging weight was about 2.3 kg. Studies covering chick development at New Island (Falklands) in 2006/07 showed high growth rates with mean chick mass peaking at about 2.5 kg after only 40 days, after which it stabilized and even slightly declined (Poisbleau et al., 2008. Polar Biol. 31, p.925-932). Flipper development was rapid in the first 35 days and then flattened out, whereas bill length continued to increase until the chicks fledged after around 65 days, albeit with a minimally slowing rate after about 30 days. Development of surviving (A) or (B) chicks was essentially indistinguishable. In spite of the high growth rate, suggesting good food availability, only 0.69 chicks were raised per nest.
Creche Phase / Acoustic Parent-Chick Recognition
During the creche period, chicks usually remain in the vicinity of their nests and recognize returning parents by acoustic signals. When arriving at the nest area, adult Southern Rockhopper Penguins emit about 0.45 sec calls made up of a long 1st sound component (syllable), followed by about 8 further shorter syllables (Searby and Jouventin 2005. J. Avian Biol. 36, p.449-460). The pattern (or tempo) of the syllables together with the harmonic content make the signal distinctive, meaning that Rockhoppers have a double vocal signature. This is less complex than the double vocal signature of Aptenodytes penguins (e.g. Emperor), but more complex than the purely harmonic system in Pygoscelis penguins (e.g. Adelie, Gentoo). The higher complexity of the signal in non-nesting penguins is accounted for by the need for parental recognition in absence of a visual focal point. In the Rockhoppers, playback experiments showed that even partial calls can be recognized. Recognition is usually from a distance of just under 2 m and probably is supplemented by a visual recognition of a penguin at the nest since responses to playbacks were under 50%, but 100% responses were recorded for calls from returning parents. On hearing the call, the chicks call and approach the parent, after which feeding commences within about 40 sec. Parents recognize their own chicks acoustically and will peck other chicks that try to get fed.
The calls of the Northern species are much lower-pitched and have an entirely different structure with fewer syllables and longer modulated phrases (Jouventin 1982. Visual and Vocal Signals in Penguins. Parey Ed.).
Towards the end of the creche phase, chicks start to develop the waterproof juvenile plumage. As this emerges, the mesoptyl down is gradually lost. Once this moult is completed, birds are essentially ready to fledge. The exact fledging date is subject to interannual and intersite fluctuation, but will be after about 70 days. The juvenile birds only moult into adult plumage about 13 months after fledging. Breeding is usually thought to start at about 5 years of age, although sexual maturity is probably reached earlier.
At the end of the breeding season, once chicks have fledged, adult birds need to forage intensively to rapidly build up weight in preparation for the annual moult. Usually, these trips take 3-4 weeks although in years of food shortages penguins can apparently delay the moult for another month or two (e.g. 85/86 in the Falklands). The moult lasts for about 25 days, during which the birds cannot forage as their plumage is not water-proof. Consequently, if the moult starts before the birds have gained a high enough pre-moult weight, starvation is the likely outcome. In a normal season, moulting birds lose about 40% of their body mass during the moult period (Brown 1985. Environ. Physiol. 155(4), p.515-520).
Rockhopper Penguins are considered quite fierce if disturbed and are known to peck and grab onto the legs of researchers moving through colonies. When threatened, the black crown feathers may be erected and the cheeks slightly puffed out. Rockhoppers may turn there heads from side to side and/or up and down whilst facing an adversary. This is generally accompanied by raising of flippers and sometimes cries. They may also jab towards an opponent with an open bill. If fighting ensues, birds may try to grab each other at the back of the neck and then beat each other with their flippers. Retreating birds may be followed and bitten in the hind as they flee. Aggressive behaviour is more readily displayed by males (which may account for the fact that the male is responsible for the chicks during the guard phase, where they are most vulnerable). It is noted that the Rockhopper usually only shows aggressive behaviour if it feels in danger or if its territory is challenged, especially by other males, although bullying of other birds, especially juveniles or chicks may be observed. Human beings that keep a reasonable distance from the nests and maintain calm are more likely to be met with curiosity than aggression.
The male is generally responsible for collecting nest material, whilst the female is more likely to be responsible for arranging the nest contents, especially at the egg-laying phase. Nest relief may be accompanied by loud braying (trumpeting) behaviour which may involve simple vertical trumpeting with the head pointing skywards and the flippers moving up and down, or may involve some swinging of the head from side to side, as in the ecstatic display. After this greeting behaviour the birds may switch positions at the nest (although this may not occur for some time). The penguins tend to take a relatively subdued hunched posture during this process and the returning bird gently shuffles onto the nest where it may make a couple of stepping motions whilst remaining essentially stationary before settling on the nest.
Like other penguins, Rockhoppers like to congregate near sources of fresh water and can sometimes be seen bathing or drinking in streams. Cleanliness is sometimes achieved by a combination of washing / preening during short trips to the sea, during which penguins tend to stay close to the coast. Preening and allopreening also serve to maintain the plumage in good condition and remove parasites, and in the latter case additionally to reinforce the pair bond.
When the penguins are not foraging or cleaning themselves, most time is spent roosting in a standing or lying position.
Rockhopper penguin numbers have declined significantly in the last 50 years, probably largely due to competition with commercial fisheries and maybe also due to increasing fluctuations in local sea temperatures (which impact on prey numbers and location), possibly resulting from global warming. Since many of the colonies of N and E Rockhoppers are located on small isolated islands, birds at these colonies are particularly vulnerable since they can not relocate to other nearby sites in response to changes in local prey availability.
The effect of commercial fisheries is controversial. However, the increase in fishery activities between 1985 and 1995 at the Falkland Islands is likely to be a contributing factor to the massive reduction in the local penguin population during this period since the fisheries target Illex and Loligo species of squid which also form part of the penguin's diet. Whilst it is argued that the fishery targets adult squid, clearly a reduction in the adult population will affect levels of juvenile squid upon which the penguins feed. Bingham, 2002 (Rev. Chilena de Hist. Nat. 75(4), p.805-818), reported that between 1984 and 1995 numbers of Southern Rockhopper Penguins dropped from 2.5 Million breeding pairs to only about 300000. Removing biomass from the feeding grounds of the penguins, especially near their colonies, may exacerbate food shortages in naturally occurring years with low biomass yields, leading to starvation, poor breeding success, and some migration to nearby South American colonies. In the 1985/86 breeding season in particular large numbers of penguins starved to death prior to or during the moult, probably due to a severe shortage of krill (Keymer et al., 2001. Dis Aquat. Org. 45, p.159-169). It has been suggested that El Nino-like climatic conditions in the Atlantic that summer, which was extremely hot and windy in the Falklands, led to the food shortage (Boersma 1987. Nature 327, p.96). Large-scale starvation would have presumably occurred even in the total absence of fisheries activity. A further large-scale starvation event was recorded in 2002 and correlated to a 10-fold drop in commercial catches of squid (Illex argentinus) in the area, suggesting general food shortages (Pütz et al., 2006. Polar Biol. 29, p.735-744). Squid levels were also depressed in 1985/86. It is also argued that global reductions in marine ecosystem productivity are responsible for the decline. However, the much-cited Hilton et al., 2006 paper (Global Change Biol. 12, p.611-625), whilst showing an average small drop in productivity, showed that patterns in the last 160 years were not consistent across sites, and concluded that no single clear global explanation could account for the decline in Rockhopper populations. It is however noted that the paper did not analyse specimens from the Falklands.
Fisheries may also directly cause penguin fatalities. For example, Rockhoppers have been reported to get entangled in the nets of drift-net fisheries at Tristan da Cunha (Ryan and Cooper 1991. Oryx 25(2), p.76-79).
In the last 200 years, Penguins and their eggs were exploited on a massive scale in the Falklands. Whilst due to its small size the Rockhopper may not have been the primary target of the penguin oil industry, it is probably that large numbers of birds were rendered to make oil. Further, Rockhopper eggs were collected by the thousand for food. For example, a single (now abandoned) colony of Rockhoppers at Sparrow Cove near Stanley provided 25000 eggs in 1871 alone (Falklands Conservation Website). Fortunately, industrial scale egg collection ceased many years ago, and in 1999 all collection of Rockhopper eggs was made illegal. Historically, Rockhoppers have also been used as bait by Cray fishermen. For example, fishermen decimated numbers of penguins on St Paul Island in the 1930s. At Tristan da Cunha the practice was reported as recently as the 1970s (Richardson 1984. Cormorant 12, p.123-201). Fortunately, this practice appears to have stopped and Cray Fisheries appear to have little other direct impact on the penguins.
A possibly greater threat is global climate change. Sea temperature changes of even only a few degrees appeared to coincide with changes in penguin numbers at Campbell Island. At this location the recorded decline in penguin numbers was strongest in the 1945-55 period, when sea temperatures were high (max. peak (highest summer temp.) of 9.7 'C in 1949), whereas a small temporary increase was recorded in the 1960s when temperatures were lower (min. peak of 8.6'C in 1965). Declines were again observed from 1970-1993, coinciding with peak averages of 9.7'C (Cunningham and Moors, 1994. Emu 94(1), p.27-36). If even such small fluctuations in sea temperatures generally have such significant effects, global warming is likely to pose a significant threat to penguin populations. Recent research which reassessed the findings of Cunningham and Moors has confirmed that a gradual reduction in prey abundance occurred from the time of onset of the population decline (Thomson and Sagar, 2002. Water & Atmosphere 10(3), p.10-12). Carbon and Nitrogen isotopes were analysed in living penguins and museum specimens from the same area, collected since 1880. Whilst the N isotopes indicated that prey composition had changed little, the C isotope analysis revealed an alarming decline in the amount of nutrition (largely krill) available to the penguins. The decline has been significant since about 1950.
It has also been hypothesized that a local drop in sea temperatures around Amsterdam Island could have led to prey migrating to more northerly waters and leading to a drop in penguin numbers. In both cases the effects may have been coincidental but no other explanatory factors could be found. More recently, the warming of waters around Tristan and Gough is being linked to alarmingly declining N Rockhopper numbers. Fisheries competition, predation, disease or pollution all can apparently be ruled out as responsible factors in this case. The warmer waters are considered less nutrient-rich and this may be affecting the whole food-chain in the area.
Predation may play a role in the sustainability of colonies. The Amsterdam Island Sub-antarctic Fur Seal population increased from 5000 to 35000 in the 1970s at a time when the penguin population was falling (Guinard et al., 1998. Col. Waterbirds 21(2), p.222-228). Seals are known to hunt penguins, although these don't usually make up a significant proportion of their prey. Sharks presumably also prey on penguins, yet little is known about at sea predation. As with many other penguin species, avian predation of eggs and young chicks is common at breeding colonies. Eggs may be taken by a variety of predators, including Skuas, Sheathbills, Gulls. Only large predators such as Skuas take chicks. Due to the relatively small size of Rockhopper chicks, these suffer proportionally higher predation by Skuas in mixed colonies (Williams 1980. Auk 97, p.754-759). At New Island Reserve in the Falklands, Skua predation accounts for significant loss of eggs and young chicks. Interestingly, not only poorly guarded eggs were snatched. On several occasions, Skuas were observed to fly straight into the breasts of incubating birds to unbalance them, after which eggs that became exposed were quickly snatched (C. & R. Cassady St.Clair 1996. Oikos 77, p.459-466). At Staten Island (Argentina), the main predators are Striated Caracaras. These often wait in the dense tussock grass at the edge of the colony until opportunities arise to prey on an egg or young chick (Liljesthroem et al., 2008. Polar Biol. 31, p.465-474). When colonies are shrinking and fragmenting due to other factors, predation may accelerate the decline, since smaller fragmented colonies have relatively more peripheral nests than large colonies and all avian predators tend to prefer hunting at the colony periphery (Jackson et al., 2005. Oikos 111, p.473-478). Interestingly, avian predation at sea was recently observed at Nightingale Island, Tristan da Cunha (Ryan et al., 2008. Ardea 96(1), p.129-134). Mature male Northern Giant Petrels were observed patrolling around the penguins landing area. The Petrels tried to grab penguins by the neck, and if successful in gaining a strong grasp would hold the penguins head under water for 5-6 minutes until it drowned. Several penguins were killed by individually hunting Petrels which either pattered across the water after penguins or lunged towards penguins that came close to them. Whilst the former technique was the most common, only 1/88 attacks observed resulted in a kill as penguins often dived to avoid the approaching Petrel. The lunging technique was more successful with 4/50 attempts resulting in a kill (although 7 birds that were initially grabbed managed to escape). One penguin was also killed after it had been dragged off the landing rock by a Petrel.
Disease may also cause mortality in penguins. For example, studies in 1985-86 showed that a number of penguins on Campbell Island had succumbed to avian cholera (Pasteurella multocida). Avian Pox virus may also threaten penguin populations, although recent outbreaks on the Falklands only caused significant mortality in Gentoo penguins. Further, algal bloom ("red tide") events can threaten all seabirds in the affected area. The algae produce toxins of the Paralytic Shellfish Poisoning group. Smaller marine organisms feed on the algae and introduce it into the food chain, along which it eventually reaches the penguins. Algal blooms were responsible for significant penguin mortality at the Falklands in 1994 (Ingham and Morris 2004. Falkl. Cons. Newslett. 4) and again in 2002/2003, when large numbers of birds were lost at New Island and other nearby sites. Penguins often carry ectoparasites such as lice, but whilst these may be implicated in disease transmission, the direct effect of such parasites is generally considered minimal.
Pollution is also a threat to all penguins. Increased oil exploration activity near the Falklands may lead to more frequent spills in the future. Rockhopper penguins examined in the wake of the 85/86 mass starvation were found to have high levels of various heavy metals, with levels of Cadmium raising concerns (Keymer et al., 2001. Dis. Aquat. Org. 45, p.159-169). The source of the Cadmium was unknown. In 2011, numerous Northern Rockhopper penguins were lost when the bulk Carrier MS Oliva stranded and broke up at the Nightingale Island breeding site, releasing large amounts of fuel oil (see Northern Rockhopper page).
Where To See:
Southern Rockhopper penguins are probably most easily seen in the wild by visiting one of the numerous nature reserves on the Falkland Islands. A number of companies offer cruises which combine the Falklands with other destinations, classically South Georgia and the Antarctic. These cruises often make stops at sites like New Island or West Point Island, both of which have significant Rockhopper rookeries. Visitors who wish to spend longer in the Falklands should contact the visitor board of the Falkland Islands.
Colonies of Eastern Rockhoppers, which are visually virtually indistinguishable from the Southern type, are presently extremely difficult to visit. A small colony is found on rocks at the Macquarie Island Isthmus and it may be possible to briefly zodiac cruise offshore of the colony during trips to the Island (e.g. with Heritage Expeditions, NZ). For Northern Rockhoppers, relatively long boat passages to Tristan da Cunha or the Amsterdam and St. Paul Islands are the only options. However, only few trips have these Islands in their itineraries.
Tourism is presently not considered to pose a threat to Rockhopper penguins, which are naturally quite inquisitive and will approach humans if they remain calm and do not make rapid movements.