The rockhopper penguins are the smallest of the crested penguins and although outnumbered by Erect-crested penguins (Eudyptes sclateri) within New Zealand, are nevertheless globally the most widespread and most numerous of the New Zealand breeding Eudyptes penguins. Rockhopper penguins have a circumpolar range, breeding on many islands in the sub-Antarctic Zone. Currently three taxa are recognised with just one of these, the eastern rockhopper penguin (Eudyptes filholi) breeding in the New Zealand Region; on Campbell, Antipodes and Auckland Islands as well as nearby Macquarie Island. In the early to mid-20th century Campbell Island was the stronghold for the eastern rockhopper penguin, but numbers there have declined by about 94% since the 1940s. Eastern rockhopper populations have also declined in numbers at both Antipodes and Auckland Islands, although the extent of decline is not adequately quantified for those islands. Rockhopper penguin populations have also declined at most other breeding islands elsewhere in the Southern Ocean. A reduction in food availability due to ocean warming associated with climate change is implicated in the declines on Campbell Island and presumably elsewhere, although at some colonies on Campbell Island other threats including predation and harassment by New Zealand sea lions (Phocarctos hookeri) have contributed to the declines observed. The breeding biology, foods and marine ecology of rockhopper penguins has been studied to varying extents at Campbell Island, but there is little information on these birds from the Antipodes Islands. For the Auckland Islands even the distribution of colonies is inadequately mapped and estimations of their numbers there are rough at best.

The chapter by Pütz et al.⁴ presents a good review of what is known about the southern rockhopper penguin, presenting information on both eastern and western taxa. Marchant & Higgins¹ is more encyclopaedic drawing together snippets of information often overlooked in other reviews but as they treat all three rockhopper taxa as a single species, information presented there needs to be used with care. The comparative review of the crested penguin species by John Warham¹⁰ is still a useful detailed overview particularly of those populations breeding in the New Zealand Region. Other descriptions of the rockhopper penguin^11,12 were written primarily for a general audience and provide rather brief introductions to the species.

Two recent workshops, one for seabirds in general¹³ the other for New Zealand penguins¹⁴ identified research required for the conservation of New Zealand penguins. Neither went through peer review and both are less detailed than required for our purposes. Research and conservation priorities for Rockhopper Penguins have been listed by Birdlife International¹⁵, Graeme Taylor¹⁶ and with less detail by Pütz et al.⁴ and Karen Baird¹⁷.

Our review and the research and conservation priorities listed herein relate specifically to the New Zealand populations of the eastern rockhopper penguin.

The Department of Conservation lists the eastern rockhopper penguin as ‘nationally vulnerable’¹⁸ and IUCN as ‘vulnerable’².

All rockhopper penguins were once considered to be a single species (E. chrysocome)¹. Birdlife International² and the Global Penguin Society recognise two species, the northern rockhopper penguin (E. moseleyi)³ and the southern rockhopper penguin (E. chrysocome), the latter with two sub-species^4,5. Although many recent papers treat the resident eastern rockhoppers as a subspecies (E. chrysocome filholi)⁶ the New Zealand checklist elevates these to full species status, the resident eastern rockhoppers as E. filholi and the western rockhoppers as E. chrysocome⁷. Both northern and western rockhopper penguins have been recorded in New Zealand as very rare vagrants. In this report we follow Gill et al.⁷ and Banks et al.⁸ in treating the eastern rockhopper penguin as a full species. A guide to distinguishing between the three rockhopper taxa is given by Heather & Robertson⁹.

The northern rockhopper penguin breeds only on the Tristan da Cunha Islands and Gough Island in the South Atlantic Ocean and Ile Amsterdam and St Paul Islands in the Indian Ocean³. The western rockhopper penguin breeds in southern South America and the Falkland Islands and the more widespread eastern rockhopper on sub-Antarctic Islands in the Indian and Pacific Oceans⁴. In the wider New Zealand Region, they breed on the Campbell, Auckland, Antipodes and Macquarie Island⁴.

All rockhopper taxa have declined in numbers with the global populations of the southern taxon (including both eastern and western rockhoppers) by 34% and northern rockhopper penguins by 57% in the last 37 years².

Within the New Zealand Region eastern rockhopper penguins breed on Campbell, Auckland and Antipodes Islands and on nearby Macquarie Island^1,4. The distribution of rockhopper penguin colonies on Campbell Island has been mapped three times, the early 1940’s, in 1984^19,20 and 2012⁶. Morrison et al.⁶ includes a table showing the history of occupation of each Campbell Island colony in 1958, 1975, 1984, 1985-87 and 2010-12. Most colonies are either on the western most peninsula or on the exposed west coast of the Island.

On the Antipodes Islands rockhopper penguins breed in amongst the much more numerous erect-crested penguins. The maps of crested penguin colonies in Hiscock and Chilvers²¹ and Hiscock²² show 103 colonies fairly evenly distributed around the main island and on Bollons, Archway and the eastern most of the Windward Islands. There is an implied assumption that eastern rockhopper penguins are present in most if not all of these colonies.

The distribution on the Auckland Islands is poorly documented. Known colonies have been mapped by Bell²³ and by Cooper²⁴, but both surveys were primarily boat-based with observations made as the vessel cruised slowly along the cliff-bound north, west and south coasts where most rockhopper penguins breed. As most colonies are small and surveying these rugged, cliff-bound, exposed coasts is challenging, it would be easy to miss entire colonies. The 14 known breeding colonies are mostly on the southern half of the west coast and the western sector of the north coast of the main Auckland Island, with just one colony on the south coast of Adams Island, one on Disappointment Island, and four colonies between Chambres Inlet and Falla Peninsula on the east coast^23,24.

Rockhopper penguins occur as vagrants or beach-cast individuals on the New Zealand mainland and at The Snares^1,25 but few sightings are identified to species level and as all three rockhopper taxa have been positively identified in the New Zealand Region⁹ these records do not provide reliable insights into the non-breeding distributions of eastern rockhopper penguins. Cole et al.²⁶ genetically identified >40 prehistoric or archaeological midden crested penguin bones collected from North and South Islands, none were from rockhopper penguins. Given the abundance of eastern rockhopper penguins on the New Zealand sub-Antarctic Islands and Macquarie Island, the paucity of records from the New Zealand mainland may be indicative of a more southerly non-breeding foraging range than for the other New Zealand crested penguins.

Campbell Island, once the stronghold of the eastern rockhopper penguin, indeed one of the largest populations of any of the rockhopper taxa, has suffered massive declines since the first population estimate was made by J.H. Sorensen who was a coast watcher there during the Second World War. Since then his initial estimate of five million birds has been reassessed, considering counts, observations, nesting densities and photographs (mostly unpublished) made by more recent observers. Sorensen’s estimate was initially revised downward to about 1.6 million breeding rockhopper penguins (814,550 pairs)^19,20, then Morrison et al.⁶ again adjusted the estimate using knowledge not available to the earlier researchers, suggesting there were 619,925 breeding pairs on Campbell Island in the early 1940’s. Based on colony photos taken in 1942 and re-photographed in 1987, Cunningham and Moors²⁰ estimated that by 1984-87 numbers had declined by 94% to about 103,100 individuals which would translate to a maximum of 51,550 pairs. The 1984 estimate further adjusted to 42,528 pairs⁶ (Table 1). Cunningham and Moors²⁰ calculated the area occupied by penguin colonies in the 1940’s was 406,600 m2 but just 25,500 m2 in the 1980s. They thought that the decline had begun by 1945 (with a major reduction in numbers until 1956) and attributed the decline to warming ocean temperatures, with a temporary increase in penguin numbers in the 1960’s during a cooling period^20,27.

The Campbell population was assessed again in 2012 using a combination of ground counts and photographs, this time estimating the population to be 33,239 breeding pairs, a decline of 21.8% from an adjusted population of 42,528 pairs in 1984 and a 94.6% decline from an adjusted estimate of 619,925 breeding pairs in 1942⁶ (Table 1). The recent declines were greatest between 1984 and 1996 after which time the overall population increased concurrent with lower sea surface temperatures (SST)⁶. They calculated the decline between 1984 and 2012 to be slower (k = 0.991) than that between 1942 and 1984 (k = 0.940). The population trends were not identical across all colonies, the differences linked to predation by New Zealand sea lions (Phocarctos hookeri) at those colonies most accessible to the sea lions²⁷.

In November 1950 R.A. Falla²⁸ suggested that rockhopper penguins outnumbered erect-crested penguins on the Antipodes Island, the reverse to that recorded by all later surveys. In 1978 a total of 50,000 pairs spread between 86 colonies were estimated to breed on the Antipodes Islands¹. No information for the basis for this estimate is given. A survey in 1995 found a maximum of 3,400 pairs on Antipodes and Bollons Islands¹⁶ with a total of 4,000 when the Archway Islands were also included²⁹. Photos of the Ringdove Bay colony showed a huge decline from tens of thousands in 1950 to just a few individuals in 1995¹⁶.

A further estimate of rockhopper numbers was made for the Antipodes Island in 2011; of the 39,701 penguins counted in ground counts, 2475 (7%) were rockhoppers, the rest being erect-crested penguins²¹. In total 42,689 nests were counted, the remainder from observation points or from a boat, suggesting there were about 2,988 pairs of rockhopper penguins (7% of 42,689) on the Antipodes in 2011 (Table 1).

A partial survey of penguin colonies at the Antipodes Island in 2014 found that the breeding population in those colonies resurveyed had declined on average by 19% since 2011³⁰. In January 2014, a major storm caused extensive landslides on the Antipodes Islands with 44% of the penguin colonies losing area due to landslides or were partially buried by landslide debris³⁰. The magnitude of the decline was roughly proportional to the area of colony affected, with an 11.7% decline in colonies not impacted by landslides, but 39.9% in colonies that had lost at least 75% of their area to landslides³⁰. With global climate change major storms such as the one in January 2014 are predicted to become more frequent.

There are no accurate estimates of the numbers of Rockhopper Penguins at the Auckland Islands. Based on a single day, boat-based survey of the west coast of the main island, plus surveys of parts of the east coast from a dingy during the 1972-73 summer, Bell (1975) estimated there to be between 5,000 and 10,000 pairs at the Auckland Islands. The only other estimate is that by Cooper²⁴ based on a two-day visit to the islands in January 1990. They surveyed the northern and western coasts of the main island and the southern coast of Adams Island from a boat and visited the four known colonies on the east coast. They located nine breeding colonies and based on an apparent reduction in area occupied estimated the total population to be 2,700 – 3,600 pairs²⁴. R. Russ was present during both surveys and thought the colonies were much smaller than he remembered from 1972-73.

While undertaking actual counts is logistically challenging, molecular based monitoring using population genetics and genomics may provide an alternative way to estimate population trends, perhaps even allowing effective population sizes to be estimated (T. Cole pers. comm.).

Numbers have also declined at Macquarie Island where in 2007 the estimate of 32,000–43,000 pairs was significantly less than past estimates².

Cunningham and Moors²⁰ were the first to link population trends in Campbell Island rockhopper penguins to changes in sea surface temperature (SST), but then the best SST data available were weekly measurements taken near the entrance of Perseverance Harbour, not necessarily indicative of SST where the penguins fed. Average SST were already increasing in the 1940’s and the 5-year running means peaked at 9.7oC in 1948-49 and 9.6oC in 1953-54. Temperatures cooled after 1957 to a low of 8.9oC in 1965, then rose to 10.2oC by 1970²⁰. The 1946-56 warm period was when they believed penguin populations declined most rapidly. There was a temporary resurgence in penguin numbers in the one colony where the best data is available following that cool period, the population increase lagging behind the temperature decline, the lag equating to the years it takes fledglings produced to recruit into the breeding population²⁰.

Recent advances in satellite technology, the online availability of oceanographic data and computing power has allowed much more robust analyses of the relationship between SST and changes in penguin populations. Morrison et al.⁶ were able to compare a 100-year time series (1913-2012) of extended reconstructed sea surface temperatures (ERSST) with changes in penguin populations. The ERSST data were based on ship and buoy SST measurements prior to 1985 and since then also included satellite measurements. Monthly ERSST data were downloaded for a rectangular area centred on Campbell Island, encompassing the expected foraging radius of the penguins during chick-rearing⁶. While they also used a 5-year moving average for ERSST in their analysis, they could do this for winter as well as summer ERSSTs. In addition to the average ERSST, they were able to compare population changes against temperature anomalies⁶.

Morrison et al.⁶ showed that on Campbell Island rockhopper penguin numbers declined during warm periods and recovered during cool periods, although the initial decline began before the regional ERSST began to increase. Since 1996 penguin numbers have rebounded a little, coincident to the current global warming hiatus, lower ERSST and increased abundance of a key prey species⁶. The timing of population fluctuations and changes in ERSST are not exact, one changing a few years before the other, partly due to the lag before fledglings join the breeding population but complicated by other marine factors⁶. They predict the long-term population decline will begin again once global climate warming resumes.

ERSST for both summer and winter from 1913 to 2012 reveal three periods of contrasting trends: a cool period 1913–1950, a warm middle era 1951–1989 with cooler years in the late 1960s, and a return to cooler temperatures 1990–2012, with a few warmer years in the early 2000s⁶. The average difference in monthly ERSST values was just 0.30oC between the 1913–1944 era when rockhopper populations were large and the 1945–1995 period of decline, and this was driven more by temperature increases over winter than summer. The increase in penguin numbers between 1996 and 2012 corresponds to a decline of just 0.06°C relative to the preceding warm period, driven primarily by lower temperatures over spring and summer⁶.

The frequency of seasonal temperature anomalies is probably more significant than the ERSST means. During the 1913-1944 cool period just 3.1 % of years had ERSST temperature anomalies, whereas during the 1945–1995 warm period 17.6 % of years were unusually warm, but none were as warm as those anomalous years in the 1996–2012 cool period⁶.

The penguins are not responding to SST directly but to changes in food availability within foraging range from their breeding colonies, with better foraging conditions during years with low SST than warmer SST. It appears that rockhoppers experienced greater food availability, consequently higher reproductive success and survival during the cooler period after 1990. The average year class strength of Southern blue whiting (Micromesistius australis), the main prey of the Campbell Island rockhopper penguins was c.3.8 times greater during the recent cool era (1990–2009) than the warmer years of 1977–1989⁶.

Stable isotope ratio analysis (SIA) of carbon and nitrogen can provide information on the trophic level targeted. SIA is based on the predictable and quantifiable ways that the ratio of carbon isotopes 13C and 12C, expressed as δ13C and nitrogen isotopes 15N/14N expressed as δ15N change at different trophic levels (Hilton et al. 2006). Hilton et al.³¹ compared δ15N and δ13C from the feathers from living penguins with those from museum specimens, some collected during the 19th century, to see how diet may have changed over time.

Hilton et al.³¹ treated all rockhopper penguins as a single taxon, thus including in their analysis the northern rockhoppers from Tristan du Cuna and Amsterdam Islands, western rockhoppers from the Atlantic and eastern rockhoppers from the Indian, and Pacific Oceans. They include data from rockhopper penguins from Campbell and Antipodes Islands and it is the results from these two sites discussed here. δ13C decreased significantly over time at seven breeding sites, including the Antipodes, but not at Campbell Island where no significant isotopic trends were evident³¹. The decrease in δ13C values at Antipodes Islands, and those other sites where δ13C declined, is indicative of decreases in primary productivity in the seas exploited by penguins from those sites, correlated with the observed declines in those penguin populations³¹. There was some evidence of a long-term decline in δ15N at some sites, indicative of a shift in diet to prey of lower trophic status, although this was not significant for Antipodes Island³¹.

Except for a few open nests, tawaki are cryptic breeders that prefer to nest in inaccessible locations. Along open coast sites like Jackson Head, the penguins tend to breed in dense, impenetrable kiekie shrub often in small clusters of nests². At Gorge River, the penguins breed in a more open forest setting, primarily under tree roots or fallen trees, in amongst kiekie or tangles of supplejack (Ripogonum scandens)³³. In Harrison Cove, Milford Sound less than a third of all nests are under rock overhangs or upturned tree roots. Most of the nests are in rock crevices or in cavities and caves under rock falls. At Sinbad Gully, Milford Sound, the penguins occupy a steep slope that is dominated by windfall of rimu (Dacrydium cupressinum) and other large trees; a few kilometres up the fjord, the majority of the penguins nest in an extensive warren under remnants of a glacial moraine³². On Codfish Island/Whenua Hou, the penguins breed in dense tree fern groves, mainly in dug out, deep burrows in the soft peaty soil or in deep hollows under tree roots, while along the north-east coast of Stewart Island the penguins breed in sea caves and fissures in cliffs only accessible from the sea (Mattern, pers. obs.).

Nests usually consist of shallow bowls lined with twigs and stones (Fig. 1).

Figure 1. Eastern Rockhopper nest, Penguin Bay, Campbell Island, November 2010 (Photo: Kyle Morrison, NZ Bird Online)

Rockhopper penguins breed in colonies that range in size from a few tens to thousands, in some colonies nest density may exceed two nests per square metre. On the Antipodes Islands, rockhopper penguins nest in mixed colonies alongside erect-crested penguins, the rockhoppers often with some form of overhead shelter, tending to be above, inland from and in rougher terrain than the larger erect-crested penguins^10,28. Rockhoppers may nest in small caves, under boulders and among tussocks up to 200 m inland and up to 100 m above sea level¹¹.

As with other crested penguins rockhoppers lay a single clutch of two eggs each year. The smaller first laid A egg followed 4-5 days later by the larger B egg^4,32. The breeding biology has been studied intensively at Macquarie Island³³ and references cited therein). Incubation begins once the B egg is laid and normally the A egg, or the chick from the A egg is lost during incubation or soon after hatching. Proportionality more rockhopper penguin A eggs hatch than in most other crested penguins, but only under exceptional circumstances do both chicks fledge. The incubation period is 32-34 days with the female taking the first incubation spell while the male is at sea, his opportunity to replenish body reserves after a month ashore. At Macquarie Island the male’s foraging trip lasts on average 12 days (range 9-17 days)³², at Campbell Island 13-16 days²⁷ during which time the female must remain ashore incubating the eggs. After hatching the male remains at the nest guarding his chick for about three weeks, during which time the female alone provisions the chick, usually returning with food daily. On first leaving his chick the male usually makes a multi-day foraging trip, typically of 7-11 days, replenishing his own body reserves after a fast of over three weeks when he was confined to the nest^4,34.

The cause and function of egg-size dimorphism in crested penguins has been debated by many researchers over the last 50 years but has yet to be satisfactorily resolved. The debate continues, and the current theories are discussed by Morrison³⁵.

Once the guard stage is over, chicks form creches while waiting for their parents to return with food. Once the male ends his post guard foraging trip, both parents return to feed their chick at or close to the nest every few days, until the chick fledges about 70 days after hatching^4,34. Well-fed rockhopper penguin chicks can reach a peak weight greater than that of their parents, slimming down before fledging¹⁰. As with other crested penguins there is delayed maturity and few breed before they are five years of age, but contrary to Warham’s¹⁰ assertion that they breed annually thereafter, more recent research at Macquarie Island showed just 8% of males and 3% of females bred in all three study years³³. Their breeding biology is similar at Campbell Island²⁷.

At Macquarie Island the first males arrived at their breeding colonies between 14 and 17 October in 1960, 1961, 1993, 1994 and 1995^32,33, with the timing of the breeding cycle constant between years³³. The females joined their mates at the nest on average 6.5 days later. Laying was well synchronised as in other crested penguins¹⁰. At Macquarie Island eggs were laid between 7 and 18 November³²; the median date of laying the A egg was 13 or 14 November (1993-1995) and the B egg 17-19 November³³. The interval between arrival and laying of the second egg was 17 to 21 days³².

At Macquarie Island the median hatching date was 18-21 December (1993-1995)³³ with both parents present at the time of hatching. At Antipodes Island the guard stage lasted until at least 27 December 2002, the median end of the guard stage being 30 December³⁶. During the guard stage females usually returned from mid-afternoon each day to feed chicks, while during post-guard stage males returned most evenings with female parents returning less often; the chicks being fed most days in both guard and post-guard stages⁶. At Macquarie Island chicks fledged between 24 February and 10 March³², the median date over three years ranging from 21 to 24 February³³. The timing of the breeding season is similar at Campbell island^34,35,37 with the dates for each stage of the cycle in Table 1 of Morrison et al.³⁷.

At the Antipodes Islands rockhopper penguins breed about 12 days earlier than at Macquarie Island, probably arriving about 7 October, a month after the erect-crested penguins with whom they share colonies. Antipodes rockhoppers laid about 1 November with eggs hatching about 4 December^28,38. By late January chicks were well grown, some were fully feathered by 5 February; the first chicks left the island on 9 February 1969 and most had fledged by 20 February^28,38. Breeding at Campbell Island is earlier than Macquarie but later than at Antipodes Island, with the peak of egg-laying about 9 November³⁸. At Campbell Island chicks fledge mid-February²⁷.

Morrison et al.³⁴ compared breeding success and foraging trip durations, the frequency at which chicks were fed and chick growth in rockhopper penguins at Campbell Island in 2011 when food was abundant, with 2012 when food was scarce. In 2012, both hatching success (0.69 in 2011, 0.58 in 2012) and reproductive success (0.60 in 2011, 0.35in 2012) were lower, and mean foraging trips during guard and creche stages longer than in 2011 when feeding conditions were more favourable. Chicks were fed about half as often in 2012 and mean chick mass at 30-31 days of age in 2011 was 1466 +/-201 g compared to 1025 +/- 233 g in 2012³⁴.

It has been suggested that the breeding regime of crested penguins, where the male has an unbroken three- to five-week period fasting ashore during late incubation and the guard stage, followed by a multi-day foraging trip before he can begin to provision the chick, limits these penguins ability to respond when food is in short supply³⁴. This they postulate, renders crested penguins more vulnerable to climate change than other penguin species where both sexes alternate brooding and feeding during the guard stage.

At Macquarie Island year to year nest site and mate fidelity has been reported to be about 50%.⁴, a statement at odds with Hull et al.’s³³ finding that few individuals breed in consecutive years. At Campbell Island nest site and mate fidelity was high (Kyle Morrison unpubl. data).

At Macquarie Island a few non-breeding birds were ashore from the start of the breeding season until the end of the guard stage. They had limited success at finding partners until incubation and chick rearing was underway when fewer aggressive breeding penguins were present, and non-breeders had greater freedom to wander through the colony seeking partners³². Some non-breeding birds managed to acquire nest sites vacated by failed breeders and form partnerships³².

At Campbell Island first year non-breeding penguins began coming ashore in early December and were common from mid-December through January. The last count was made just before observers left the islands in late January in both 2011 and 2012 was the highest suggesting numbers ashore may have continued to increase later in the summer³⁴. At Macquarie Island yearlings were first seen ashore about 10 December, they were quieter than and dominated by older penguins, some formed short-lived liaisons with other immatures, non-breeding adults or even chicks.

At Campbell Island sub-adults moult in January²⁷ while at Antipodes Island most yearlings had moulted by about 15 February and all adults appeared to be at sea on their pre-moult excursion on 12 March²⁸. On Macquarie Island the pre-moult absence lasts 28-35 days suggesting that adults would return to the Antipodes to moult about 5-10 April²⁸. At Macquarie Island yearlings began moult around 16 January, non-breeding adults in late February and breeding birds from late March³².

Figure 2. Eastern Rockhopper penguins moulting on Campbell Island, january 2012 (Photo: Kyle Morrison, NZ Birds Online)

Rockhopper Penguins are opportunistic foragers, with variation in their diet and foraging ecology between the different breeding populations, presumably reflecting availability more so than preference⁴.

The most common prey taken by eastern rockhopper penguins at Campbell Island in 1985 and 1986 were dwarf cod (Austrophycis marginata), and juveniles of southern blue whiting and hake (Merluccius australis), in contrast to a euphausiid dominated diet elsewhere^20,36,39. At Macquarie Island, euphausiids principally Euphausia vallentini comprised 62% and 70% of food by weight in two different studies^40–42 in which the euphausiid Thysanoessa gregaria⁴⁰ and the fish Krefftichthys anderssoni⁴² were other important prey. At Macquarie Island one Zanclorhynchus spinifer was 60 mm long and five Nothothenia sp were <100 mm in length⁴⁰.

Cooper et al.³⁹ reviewed what was then known about crested penguin diet and compare the foods of southern rockhopper penguins across their range. The prey species taken by Campbell Island rockhopper penguins are listed in Table 2. Xavier et al.⁴³ recorded eight species of juvenile and sub-adult cephalopods taken by rockhopper penguins at Campbell Island, with Onykia ingens, Martialia hyadesi and Octopus campbelli being the most important species by frequency of occurrence, number and mass. There is little other information on the diet of rockhopper penguins from the New Zealand Region.

The mean fledgling weights of Campbell Island chicks in 1987 and 1988 was 1,992g (1,560-2,400g), significantly lighter than western rockhopper penguins at the Falkland Islands, the difference attributed to their fish diet²⁰.

A fish-based, high tropic level diet is often assumed to be better for penguins than a diet based on lower trophic level cephalopod and crustacean prey. However, contrary to predictions a study using stable isotopes found that during incubation male Campbell Island rockhopper penguins were heavier in 2011 when lower trophic level, offshore, pelagic, zoo-plankton dominated their diet than in 2012 when their diet was primarily benthic, inshore, fish and cephalopods³⁷. Similarly, in 2011 average chick mass was higher than in 2012. They attributed this unexpected result to the low energy density of the southern blue whiting that was thought to dominate their diet in 2012³⁷.

Information of foods consumed as well as parasites, bacteria and the sex and genotype of the bird can now be obtained by molecular analysis of scats. This is non-invasive and samples can be obtained even during very brief visits. Any study wishing to utilise this method will need to ensure a genetic database (such as GenBank’s BLAST; https://blast.ncbi.nlm.nih.gov/Blast.cgi) of all potential prey items is available to compare sequence data to, as a reduced database will limit the power of the analysis and miss potential food species.

Within the New Zealand Region there has been just one study tracking rockhopper penguins at sea. This was conducted on the Antipodes Island from 18 December 2002 to 3 January 2003³⁶. During the guard stage eight tracked females foraged 22-54 km (mean 36.4 km) from their colony. The mean distance travelled on a feeding trip was 81.6 km (range 50-114 km), going NNE to feed where the sea was 500-1,500 m deep, or east to waters >1,500 m in depth³⁶. During the guard stage the mean duration of feeding trips for tracked birds was 1.37 days, significantly longer than that of birds unimpeded by tracking devices. Two female penguins were tracked on post-guard trips, for one the trip duration was 5.67 days, the maximum distance from the colony 104 km, in total traveling an estimated 243 km. The other was at sea for 6.92 days, at furthest was 119 km from the colony and travelled about 325 km³⁶. Both foraged along the subantarctic slope in waters >1,500 m deep.

Rockhopper Penguins spend the five to six months between breeding seasons at sea, seldom if ever touching land. Campbell Island adult Rockhopper penguins spent the winter south-east or east of the island some birds travelling about 15,000 km during that period⁴⁴. On average the maximum distance from the Island was about 2,000 km, with one bird 4,000 km from the island.

Of the three islands in the New Zealand Region where rockhopper penguins breed, Antipodes Island had only mice (Mus musculus, eradicated in 2016). At Campbell Island sheep (Ovis aries), cattle (Bos taurus), Norway rats (Rattus norvegicus) and cats (Felis catus), all since eradicated were present. At the Auckland Islands feral cats, pigs (Sus scrofa) and mice are present on the main island but there are no introduced mammals on Disappointment and Adams Islands, the only other islands in the Auckland archipelago where rockhopper penguins breed. On Campbell Island disturbance by sheep or people did not contribute to population declines as population trends were similar in colonies accessible and inaccessible to these mammals^19,20. Rats did eat penguin eggs and small chicks, but only cracked or broken eggs could be taken by rats and it was not determined if the chicks eaten were preyed upon or scavenged by the rats^19,20. There was no evidence to suggest cats preyed on the penguins. Today, within the New Zealand Region, introduced predators co-occur with breeding rockhopper penguins only on the main Auckland Island, but as most penguin colonies are at the base of cliffs, introduced predators appear to have little if any impact.

Genetic means can now be used to detect prey items in the scats of predatory mammals (http://www.ecogene.co.nz). This method may provide clues as to the scale of impact invasive mammals have on the Auckland Island rockhopper penguin populations.

Native predators do however kill rockhopper penguin eggs, chicks and adults. Subantarctic skuas (Catharacta antarctica lonnbergi) are present on all islands where these penguins breed and they take both eggs and chicks. Skuas were the main cause of egg and chick loss at Campbell Island in 2011 and 2012 when skua predation was studied at the Penguin Bay rockhopper colonies²⁷. In the smallest of the four colonies (mean: 26 pairs/year) 44% of eggs were taken by skuas, compared with just 7% in the largest of the colonies (mean: 1,476 pairs/year) where there were proportionality fewer peripheral nests. Many of the A eggs were scavenged rather than predated and whether they were taken by skuas, they would not have resulted in a fledged chick. Proportionality fewer B eggs were taken by skuas in larger than the smaller colonies²⁷. Skuas were better able to steal eggs when the smaller females were incubating than the larger more aggressive male penguins²⁷.

At the Antipodes Islands northern giant petrels (Macronectes halli) gathered on the shoreline when penguins fledged, they ate dead chicks but were not seen to actually kill penguin chicks²⁸. On Campbell Island northern giant petrels were observed or implicated in killing a small number (<10 per breeding season) of sub-adult and adult rockhopper penguins²⁷.

At Macquarie Island rockhopper penguins were seen bearing injuries from attacks by New Zealand fur seals (Arctocephalus forsteri)³². Fur seals were seen preying on rockhopper penguins at Campbell Island in the 1940’s⁴⁵ but not between 2010 and 2012; conversely New Zealand sea lions regularly preyed on sub-adult and adult penguins in the latter but not in the former period^6,27. Sea lions were uncommon at Campbell Island in the 1940’s (<20 pups born/year) but more common in the 21st century (&rt;681 pups born in 2009)⁶. The sea lions could only access some of the Campbell Island rockhopper penguin colonies but at those colonies they can contribute to a decline in penguin numbers. At the Penguin Bay colonies sea lions were estimated to kill 6% of the adult penguins present in 2011 and 3.6% in 2012²⁷, and their calculations of adult survival suggest that sea lions accounted for most if not all the adult mortality during chick rearing. Most of the sea lion predation was probably by just one or two male sea lions, although at least six individuals were seen to kill penguins. Predation occurred only in the water or during chases where the penguin managed to jump ashore and was then caught by the pursuing sea lion. Sea lions also contributed to egg, chick and adult deaths by trampling when the mammals blundered their way through penguin colonies²⁷. While sea lions are contributing to the ongoing decline of the Penguin Bay colonies, these colonies represent just 9% of the Campbell Island rockhopper penguin population. Thus, control of the sea lions is not recommended²⁷.

Land-based threats

Marine-based threats

1. Population monitoring

2. Marine Ecology

3. Breeding biology

4. Predators and diseases

Thanks to Klemens Pütz, Kyle Morrison and Tess Cole for their comments on early drafts of this paper. Graeme Taylor reviewed the research priorities section of this review. Louise Chilvers and Jo Hiscock kindly allowed us to include observations of their paper currently in review. This review was kindly funded by the T-Gear Charitable Trust through the Birds New Zealand Research Fund. We are especially grateful to Peter Gaze who facilitated our work and gave encouragement where it was required.

1. Marchant, S & Higgins, PJ. (1990) Handbook of Australian, New Zealand and Antarctic Birds. in Handbook of Australian, New Zealand and Antarctic Birds. Vol. 1 1A: (Oxford University Press).
2. Birdlife International. (2019) Southern Rockhopper Penguin (Eudyptes chrysocome) - BirdLife species factsheet. Species factsheet: Eudyptes chrysocome Available at: http://datazone.birdlife.org/species/factsheet/southern-rockhopper-penguin-eudyptes-chrysocome. (Accessed: 30th July 2019)
3. Cuthbert, RJ. (2013) Northern rockhopper penguin (Eudyptes moseleyi). in Penguins; natural history and conservation (eds. Borboroglu, P. G. & Boersma, P. D.) 131–143 (University of Washington Press).
4. Pütz, K, Rey, AR & Otley, H. (2013) Southern rockhopper penguin (Eudyptes chrysocome). in Penguins; natural history and conservation (eds. Borboroglu, P. G. & Boersma, P. D.) 113–129 (University of Washington Press).
5. del Hoyo Calduch, J & Collar, NJ. (2014) Illustrated Checklist of the Birds Od the World: Non-passerines. (Lynx Edicions).
6. Morrison, KW, Battley, PF, Sagar, PM & Thompson, DR. (2015) Population dynamics of Eastern Rockhopper Penguins on Campbell Island in relation to sea surface temperature 1942–2012: current warming hiatus pauses a long-term decline. Polar Biology 38: 163–177
7. Gill, B et al. (2010) Checklist of the Birds of New Zealand, Norfolk and Macquarie Islands, and the Ross Dependency, Antarctica. (Te Papa Press).
8. Banks, J, Van Buren, A, Cherel, Y & Whitfield, JB. (2006) Genetic evidence for three species of rockhopper penguins, Eudyptes chrysocome. Polar Biology 30: 61–67 Available at: https://doi.org/10.1007/s00300-006-0160-3.
9. Robertson, HA & Heather, BD. (2015) The field guide to the birds of New Zealand. Revised edition. (Penguin Books).
10. Warham, J. (1975) The crested penguins. in The biology of penguins (ed. Stonehouse, B.) 189–269 (Macmillan).
11. Morrison, KW. (2013) Eastern rockhopper penguin [Updated 2017]. Miskelly, C.M. (ed.) New Zealand Birds Online Available at: http://nzbirdsonline.org.nz/species/eastern-rockhopper-penguin. (Accessed: 30th July 2019)
12. De Roy, T, Jones, M & Cornthwaite, J. (2013) Penguins: Their World, Their Ways. (CSIRO Publishing).
13. Wilson, K-J & Waugh, S. (2013) New Zealand Seabird Research Priorities Workshop, Te Papa, Wellington, 5-6 May 2013. Te Papa. Wellington. Available at: http://www.birdlife.org.au/images/uploads/branches/documents/ASG-Seabird-Priorities-may13.pdf.
14. Wilson, K-J & Otley, H. (2014) Research and Management priorities for New Zealand Penguins.
15. BirdLife International. (2018) Eudyptes chrysocome. The IUCN Red List of Threatened Species 2018: e.T22735250A132664584. Available at: http://dx.doi.org/10.2305/IUCN.UK.2018-2.RLTS.T22735250A132664584.en. (Accessed: 31st July 2019)
16. Taylor, GA. (2000) Action Plan for Seabird Conservation in New Zealand. Part A: Threatened Seabirds. Department of Conservation, New Zealand. Wellington.
17. Baird, K. (2016) Reversing the penguin decline in New Zealand. Forest & Bird. Wellington.
18. Robertson, HA et al. (2017) Conservation status of New Zealand birds, 2016. Department of Conservation. Wellington. Available at: https://www.doc.govt.nz/Documents/science-and-technical/nztcs19entire.pdf. (Accessed: 29th July 2019)
19. Moors, PJ. (1986) Decline in numbers of rockhopper penguins at Campbell Island. Polar record 23: 69–73 Available at: https://doi.org/10.1017/S0032247400006835.
20. Cunningham, DM & Moors, PJ. (1994) The decline of Rockhopper Penguins Eudyptes chrysocome at Campbell Island, Southern Ocean and the influence of rising sea temperatures. Emu 94: 27–36 Available at: http://www.publish.csiro.au/MU/MU9940027.
21. Hiscock, JA & Chilvers, BL. (2014) Declining eastern rockhopper (Eudyptes filholi) and erect-crested (E. sclateri) penguins on the Antipodes Islands, New Zealand. New Zealand Journal of Ecology 38: 124
22. Hiscock, J. (2013) Monitoring penguins in the Antipodes Island Group. Department of Conservation Technical Series 37: Department of Conservation. Wellington. Available at: https://www.doc.govt.nz/globalassets/documents/science-and-technical/docts37entire.pdf.
23. Bell, BD. (1975) Report on the birds of the Auckland Islands expedition 1972-73. in Preliminary results of the Auckland Islands Expedition 1972-1973 136–142 (Department of Lands and Survey).
24. Cooper, W. (1992) Rockhopper penguins at the Auckland Islands. Notornis 39: 66–67 Available at: http://notornis.osnz.org.nz/system/files/Notornis_39_1_66.pdf.
25. Powlesland, RG. (1984) Seabirds found dead on New Zealand beaches in 1982 and a review of penguin recoveries since 1960. Notornis 31: 155–171
26. Cole, TL et al. (2019) Ancient DNA of crested penguins: Testing for temporal genetic shifts in the world’s most diverse penguin clade. Molecular Phylogenetics and Evolution 131: 72–79 DOI: https://doi.org/10.1016/j.ympev.2018.10.025
27. Morrison, KW, Armstrong, DP, Battley, PF, Jamieson, SE & Thompson, DR. (2017) Predation by New Zealand sea lions and Brown Skuas is causing the continued decline of an Eastern Rockhopper Penguin colony on Campbell Island. Polar Biology 40: 735–751 Available at: https://doi.org/10.1007/s00300-016-1996-9%0A.
28. Warham, J & Bell, BD. (1979) The birds of Antipodes Island, New Zealand. Notornis 26: 121–169
29. Tennyson, AJD, Taylor, RH, Taylor, G, Imber, M & Greene, T. (2002) Unusual bird records from the Antipodes Islands in 1978-. I 995, with a summary of other species recorded at the island group. Notornis 49: 241–245 Available at: https://www.researchgate.net/profile/Alan_Tennyson/publication/281642541_Unusual_bird_records_from_the_Antipodes_Islands_in_1978-1995_with_a_summary_of_other_species_recorded_at_the_island_group/links/565cae1c08aefe619b2539c8/Unusual-bird-records-from-the.
30. Chilvers, BL & Hiscock, JA. (2019) Significant decline of endangered Antipodes Island penguins: Is extreme weather an additional impact? Aquatic Conservation: Marine and Freshwater Ecosystems 29: 546–553 DOI: https://doi.org/https://doi.org/10.1002/aqc.3034
31. Hilton, GM et al. (2006) A stable isotopic investigation into the causes of decline in a sub‐Antarctic predator, the rockhopper penguin Eudyptes chrysocome. Global Change Biology 12: 611–625 DOI: https://doi.org/https://doi.org/10.1111/j.1365-2486.2006.01130.x
32. Warham, J. (1963) The rockhopper penguin, Eudyptes chrysocome, at Macquarie Island. The Auk 80: 229–256 Available at: https://sora.unm.edu/sites/default/files/journals/auk/v080n03/p0229-p0256.pdf.
33. Hull, CL et al. (2004) The breeding biology and factors affecting reproductive success in rockhopper penguins Eudyptes chrysocome at Macquarie Island. Polar Biology 27: 711–720 Available at: https://doi.org/10.1007/s00300-004-0643-z.
34. Morrison, KW et al. (2016) The canalized parental roles of a Eudyptes penguin constrain provisioning and growth of chicks during nutritional stress. Behavioral ecology and sociobiology 70: 467–479 Available at: https://doi.org/10.1007/s00265-016-2060-z%0A.
35. Morrison, KW. (2016) Individual repeatability in laying behaviour does not support the migratory carry‐over effect hypothesis of egg‐size dimorphism in Eudyptes penguins. Journal of Avian Biology 47: 466–475 DOI: https://doi.org/https://doi.org/10.1111/jav.00740
36. Sagar, PM, Murdoch, R, Sagar, MW & Thompson, DR. (2005) Rockhopper penguin (Eudyptes chrysocome filholi) foraging at Antipodes Islands. Notornis 52: 75–80 Available at: https://notornis.osnz.org.nz/system/files/Notornis_52_2_75.pdf.
37. Morrison, KW, Bury, SJ & Thompson, DR. (2014) Higher trophic level prey does not represent a higher quality diet in a threatened seabird: implications for relating population dynamics to diet shifts inferred from stable isotopes. Marine biology 161: 2243–2255 Available at: https://link.springer.com/article/10.1007/s00227-014-2502-y.
38. Warham, J. (1972) Breeding seasons and sexual dimorphism in Rockhopper Penguins. The Auk 89: 86–105 Available at: https://sora.unm.edu/sites/default/files/journals/auk/v089n01/p0086-p0105.pdf.
39. Cooper, J et al. (1990) Diets and dietary segregation of crested penguins (Eudyptes). in Penguin Biology 131–156 (Academic Press).
40. Horne, RSC. (1985) Diet of royal and rockhopper penguins at Macquarie Island. Emu 85: 150–156 Available at: https://doi.org/10.1071/MU9850150%0A%0A.
41. Hindell, MA. (1988) The diet of the rockhopper penguin Eudyptes chrysocome at Macquarie Island. Emu 88: 227–233 Available at: https://doi.org/10.1071/MU9880227%0A%0A.
42. Hindell, MA, Robertson, GG & Williams, R. (1995) Resource partitioning in four species of sympatrically breeding penguins. in The penguins: ecology and management. 196–215 (Surrey Beatty). Available at: https://www.researchgate.net/profile/Mark_Hindell/publication/236621979_Resource_partioning_in_four_sympatrically_breeding_species_of_penguin/links/58bf5ca6aca272bd2a3ad46b/Resource-partioning-in-four-sympatrically-breeding-species-of-penguin.pdf.
43. Xavier, JC et al. (2018) Eastern rockhopper penguins Eudyptes filholi as biological samplers of juvenile and sub-adult cephalopods around Campbell Island, New Zealand. Polar Biology 41: 1937–1949 Available at: https://doi.org/10.1007/s00300-018-2333-2%0A.
44. Thompson, DR. (2016) Penguins reveal unknown swimming talents. NIWA media release https://www.niwa.co.nz/news/penguins-reveal-unknow Available at: https://www.niwa.co.nz/news/penguins-reveal-unknown-swimming-talents.
45. Bailey, AM & Sorensen, JH. (1962) Subantarctic Campbell Island. in (Denver Museum of Natural History).
46. Duignan, PJ. (2001) Diseases of penguins. Surveillance 24: 5–11
47. Crawford, R et al. (2017) Tangled and drowned: a global review of penguin bycatch in fisheries. Endangered Species Research 34: 373–396 Available at: https://www.int-res.com/articles/esr2017/34/n034p373.pdf.
48. Black, C, Rey, AR & Hart, T. (2017) Peeking into the bleak midwinter: Investigating nonbreeding strategies of Gentoo Penguins using a camera network. The Auk: Ornithological Advances 134: 520–529 Available at: https://doi.org/10.1642/AUK-16-69.1.
49. Black, C, Southwell, C, Emmerson, L, Lunn, D & Hart, T. (2018) Time-lapse imagery of Adélie penguins reveals differential winter strategies and breeding site occupation. PloS one 13: e0193532 Available at: https://doi.org/10.1371/journal.pone.0193532.
50. Block, WM, Franklin, AB, Ward Jr, JP, Ganey, JL & White, GC. (2001) Design and implementation of monitoring studies to evaluate the success of ecological restoration on wildlife. Restoration Ecology 9: 293–303 Available at: https://onlinelibrary.wiley.com/doi/abs/10.1046/j.1526-100x.2001.009003293.x.
51. Weimerskirch, H, Prudor, A & Schull, Q. (2018) Flights of drones over sub-Antarctic seabirds show species- and status-specific behavioural and physiological responses. Polar Biology 41: 259–266 DOI: https://doi.org/10.1007/s00300-017-2187-z
52. Hodgson, JC et al. (2018) Drones count wildlife more accurately and precisely than humans. Methods in Ecology and Evolution 9: 1160–1167 Available at: https://doi.org/10.1111/2041-210X.12974.
53. Brisson-Curadeau, É et al. (2017) Seabird species vary in behavioural response to drone census. Scientific Reports 7: 17884 DOI: https://doi.org/10.1038/s41598-017-18202-3
54. Gendner, J-P, Gauthier-Clerc, M, Le Bohec, C, Descamps, S & Le Maho, Y. (2005) A new application for transponders in studying penguins. Journal of Field Ornithology 76: 138–143 Available at: https://doi.org/10.1648/0273-8570-76.2.138.
55. Deagle, BE, Chiaradia, A, McInnes, J & Jarman, SN. (2010) Pyrosequencing faecal DNA to determine diet of little penguins: is what goes in what comes out? Conservation Genetics 11: 2039–2048 Available at: https://doi.org/10.1007/s10592-010-0096-6.
56. Mattern, T, McPherson, MD, Ellenberg, U, van Heezik, Y & Seddon, PJ. (2017) High definition video loggers provide new insights into behaviour, physiology, and the oceanic habitat of marine top predators. PeerJ Preprints 5: e2765v1