Providing evidence to improve practice

Action: Control mammalian predators on islands for seabirds Bird Conservation

Key messages

  • We found 16 before-and-after studies, one paired sites study and one literature review from around the world, all describing positive seabird responses to the removal or control of mammalian predators (mainly rats Rattus spp. and feral cats Felis catus) from islands.
  • Of these 18 studies, seven found either large population increases or recolonisations following predator eradication or control. Two of these found only partial population increases or recolonisations: a study from Alaska.
  • Twelve studies found increases in reproductive success and survival or decreases in predation and mortality following predator control. In one case there was also a small population increase.
  • Rats and mice Mus musculus were controlled in twelve studies, mostly examining burrow-nesting seabirds; cats in eight, mostly on ground or cliff-nesting seabirds; and other species in two.


Supporting evidence from individual studies


A before-and-after study on the sub-Antarctic Marion Island (290 km2), South Africa (Cooper & Fourie 1991), found that breeding success of great-winged petrels Pterodroma macroptera, increased from 0-21% to 56-60% following 14 years of cat Felis catus control. In addition, no signs of cat predation were found in 1990, but at least 28% of chicks were predated in 1983.  Nests were monitored in 1979-80, 1982 and 1984 (between 17 and 53 nests studied) and in 1990 (50 nests). Control consisted of the release of the disease panleucopaenia, shooting and trapping.



Two before-and-after studies on two Galápagos Islands, Ecuador (Cruz & Cruz 1996), found that dark-rumped petrel Pterodroma phaeopygia fledging success increased (on Floreana Island, 173 km2) and predation of adults decreased (on Santiago Island, 585 km2) following black rat Rattus rattus and feral animal control. On Floreana, fledging success increased (and nestling predation decreased) from 31% in 1981-2 to 46% (1983) and 72% (1984) with control in 1983-4. It declined in 1985 (to 23%) with increased feral cat Felis catus predation of young but increased again following cat control to 70-80% in 1986-8. Between 83 and 104 nests were studied each year. On Santiago, 55% of 510 monitored adult petrels were predated in 1985, but only six were in 1986, following an 80% reduction in the feral pig Sus scrofa population. Rats were intensively controlled by poisoning with coumatetralyl, whilst pigs, cats, goats Capra hircus and donkeys Equus asnus were shot. This study is also discussed in ‘Provide artificial nesting sites’ and ‘Use vocalisations to attract birds to safe areas’.



A replicated and controlled paired sites study in 1993-4 in six areas of shrubland on Sand Island, Midway Atoll, Hawaii, USA (Seto & Conant 1996), found that Bonin petrel Pterodroma hypoleuca reproductive success was higher in two out of three areas with black rat Rattus rattus removal (treatment sites), compared to paired control sites (65-92% success for 144 nests in treatment sites vs. 0-96% for 144 in controls). In addition, the number of nests that failed because of rat predation was significantly lower in treatment sites (overall, 41% of 17 failed nests in treatment areas were due to rats vs. 95% of 41 nests in control areas, 48 nests monitored at each site). Rats were controlled with the rodenticide Bromothalin dispensed from bait stations. The authors note that early in the 1993 breeding season, there was limited poisoning in the control sites as well, which may have resulted in higher than expected breeding success in these sites.



A controlled before-and-after study on Simeonof (4,000 ha) and Chernabura (3,000 ha) Islands in the Shumagin Islands, Alaska (Byrd et al. 1997) found that the average number of pigeon guillemots Cepphus columba recorded increased from 28 to 46 individuals on Chernabura, between 1994 and 1995, following the eradication of introduced arctic foxes Alopex lagopus. There was no increase on Simeonof Island, with an average of four recorded in both years. Guillemot densities were significantly lower than on islands without foxes.



A before-and-after study on Madeira (741 km2), Portugal between 1985 and 2000 (Zino et al. 2001) found that the population of Zino’s petrel Pterodroma madeira on the five known breeding ledges increased from 12 pairs in 1987 to 29 pairs in 2000, following black rats Rattus rattus and cats Felis catus control. In 1987 (prior to cat control), ten adult petrels were killed by cats but only four were found dead between 1992 and 2000, of which two had been partially eaten by cats. Rats were controlled using 65 brodifacoum bait boxes, first set in 1986; cats were caught in traps, with eight placed in 1991, increasing to 17 by 2000.



A replicated before-and-after study on Natividad Island (7.4 ha), Baja California Sur, Mexico (Keitt & Tershy 2003), found that mortality rates of black-vented shearwaters Puffinus opisthomelas in four survey plots fell by 90%, from 7.4 carcasses/month (April-July 1997) to 0.7 carcasses/month (March-June 2002) following the eradication of feral cats Felis catus from the island. Extrapolated to the entire colony, this suggests a decrease in mortality from over 1,000 birds/month to fewer than 90 birds/month, with each of 25 cats killing approximately 37 birds/month.



A before-and-after study on The Chicken Islands, North Island, New Zealand (Parrish 2005) found that nesting success of Pycroft’s petrel Pterodroma pycrofti and little shearwater Puffinus assimilis increased from 20% in 1992-3 to 75% in 1994-5, following the eradication of Pacific rat Rattus exulans from Lady Alice Island (1.4 km2) in 1994. Rats were eradicated using brodifacoum-impregnated cattle feed.



A before-and-after study on Congreso (25.6 ha), Chafarinas Islands, Spain, between 1997 and 2004 (Igual et al. 2005) found breeding success of Cory’s shearwaters Calonectris diomedea at two sub-colonies increased when black rats Rattus rattus were intensively controlled and fell when control was relaxed. Success was 27-44% and 51% during two periods of little or no control (1997-8, 2000), compared with 70% and 71% during intensive control (2000 and 2002-4). Increases were due to decreased chick mortality from 52% and 23% in 1999 and 2001 to 11% in 2000. The increase in breeding success after control was greater at the more easily accessible colony, although that colony had lower breeding success overall. Between 208 and 280 nests were studied each year.



A before-and-after study on Lundy Island (445 ha), southwest UK (Lock 2006) found that Manx shearwaters Puffinus puffinus and Atlantic puffins Fratercula arctica both returned to breed on the island after an absence of 45 and 20 years respectively, following the successful eradication (by poisoning) of brown rats Rattus norvegicus and black rats Rattus rattus in 2004.



A before-and-after study on Isla Isabelle (194 ha), western Mexico (Rodriguez et al. 2006) found that mortality rates of adult sooty terns Onychoprion fuscata fell from an estimated 23-33% of the nesting population over the breeding season in 1991-4 to 5% in 1996 and then to 2% in 2002-4, following the eradication of cats Felis catus through trapping, poisoning (with 1080 sodium monofluoroacetate) and shooting. An attempted eradication of black rats Rattus rattus at the same time, using brodifacoum poisoning failed.



A before-and-after study on Mokoli’i Island (1.6 ha), Hawaii, USA (Smith et al. 2006), found that the reproductive output of wedge-tailed shearwaters Puffinus pacificus increased from a single chick between 1999 and 2001 to 126 chicks in 2002 and 185 in 2003, following the eradication of black rats Rattus rattus between March and October 2002. Rats were eradicated using diphacinone bait and snap and cage traps.



A controlled before-and-after study on Langara Island (3,270 ha), British Columbia, Canada (Regher et al. 2007), found that the estimated population of ancient murrelets Synthliboramphus antiquus increased from 13,000 (1999) to 24,000 (2004), following the eradication of brown rats Rattus norvegicus in 1995 through brodifacoum poisoning. This followed a period of population decline between 1981 and 1999. On islands without eradication programmes (and on Langara Island between 1981-99), ancient murrelet populations declined if rats were present (22% over ten years on Lyell Island, 50% over seven years followed by extirpation on Kunghit Island) and were stable or increasing on islands without rats. A population of Cassin’s auklets Ptychoramphus aleuticus (previously extirpated on Langara) also re-established following rat eradication.



A before-and-after study over 17 years (1990-2007) on Ascension Island (88 km2), South Atlantic (Hughes et al. 2008), found that the sooty tern Onychoprion fuscata population increased (though not significantly) from 302,000-417,000 birds (1990 and 2001) to 420,000 (2007), following feral cat Felis catus eradication in 2002-4. The authors note that no increase would be expected until 2008 due to the breeding cycle of the terns. Predation by cats fell from 33 birds/night (early 1990s) to zero  birds/night (2003), and overall nesting success rose from 54% of 233 nests to 68% of 656 nests. Following cat eradication, there was a significant increase in the number of tern chicks being predated by rats, from zero prior to cat eradication (473 days of monitoring) to 46% of 200 chicks ringed in 2005 (40 days of monitoring). Cats were removed through poisoning (with 1080 sodium monofluoroacetate), trapping and other methods (started in 2002, completed 2004).



A before-and-after study on Juan de Nova Island (4.4 km2), Mozambique Channel in July-August 2006 (Peck et al. 2008) found that predation rates on sooty terns Sterna fuscata fell from an estimated 2,205 terns/week to 416 terns/week following the control of feral cats Felis catus. Forty three cats were removed through trapping and shooting, leaving an estimated ten remaining. Predation rates were based on the recovery of 122 tern carcasses, of which 89 (73%) showed signs of predation and consumption and 27 (22%) showed signs of predation but not consumption (‘surplus killing’).



A before-and-after study on Selvagem Grande (245 ha), Madeira Archipelago, Portugal (Zino et al. 2008), found that breeding success and productivity of Cory’s shearwaters Calonectris diomedia borealis was significantly higher, following the eradication of rabbits Oryctolagus cuniculus and house mice Mus musculus from the island. This paper is discussed in detail in ‘Control or remove habitat-altering mammals’.



A review of rat Rattus spp. eradication programmes on five islands around the UK between 1959 and 2006 (Ratcliffe et al. 2009) found that bird populations increased on four of the islands: Atlantic puffins Fratercula arctica have recolonised Ailsa Craig, southwest Scotland and their range and population increased on Handa, north Scotland. Manx shearwaters Puffinus puffinus increased from 700 to 2,000 pairs on Ramsey, west Wales and 308 to 1,120 pairs on Lundy, southwest England (see Lock 2006). However, they did not recolonise Cardigan Island, southwest Wales.



A before-and-after study on Feno Islet (1.6 ha), Azores, Portugal between 1997 and 2009 (Amaral et al. 2010) found that both roseate tern Sterna dougallii and common tern S. hirundo populations returned to the islet following the eradication of black rats Rattus rattus in March-April 2007. Pre-rat populations of roseate and common terns were estimated at 340 and 280 pairs respectively, but none bred in 2005. Common terns returned in 2007 and increased to approximately 120 pairs by 2009. Few roseate terns returned in 2007-8, but 260 pairs bred in 2009. A total of nine rats were trapped on the island.



A before-and-after study on Ascension Island (88km2), South Atlantic (Ratcliffe et al. 2010) found that five species of ground-nesting seabird recolonised the island in small but increasing numbers, following the eradication of cats Felis catus in 2004: white-tailed tropicbird Phaethon lepturus (25 pairs in 2007), red-billed tropicbird Phaethon aethereus eight pairs), brown noddy Anous stolidus (79 pairs), masked booby Sula dactylatra (152 pairs) and brown booby Sula leucogastor (29 pairs). All recolonising populations were small (less than 18% of the population breeding in cat-inaccessible sites in 2002) and breeding success for boobies and brown noddy was lower than other populations. The study also found that sooty tern Onychoprion fuscata numbers increased (see reference Hughes 2008). Cats were poisoned with 1080 sodium monofluoroacetate and trapped (488 cats poisoned, 70 captured in live traps, nine caught in various methods).


Referenced papers

Please cite as:

Williams, D.R., Child, M.F., Dicks, L.V., Ockendon, N., Pople, R.G., Showler, D.A., Walsh, J.C., zu Ermgassen, E.K.H.J. & Sutherland, W.J. (2017) Bird Conservation. Pages 95-244 in: W.J. Sutherland, L.V. Dicks, N. Ockendon & R.K. Smith (eds) What Works in Conservation 2017. Open Book Publishers, Cambridge, UK.