Providing evidence to improve practice

The journal, Conservation Evidence

Our online journal publishes research, monitoring results and case studies on the effects of conservation interventions. All papers include some monitoring of the effects of the intervention and are written by, or in partnership with, those who did the conservation work. It includes interventions such as habitat creation, habitat restoration, translocations, reintroductions, invasive species control, and education or integrated conservation development programmes, from anywhere around the world.

A volume is created each year with peer-reviewed papers published throughout the year. We now accept Short Communications as well as standard papers.

Special issues contain new papers on a specific topic.

Virtual collections collate papers published in the journal on specific topics such as management of particular groups of species.

To search for papers on a specific topic within the journal select 'Advanced search' on the Home page, enter your keyword(s) and within the Source box type: "conservation evidence".

Invasive animals



This virtual collection contains 27 papers on the management of invasive animals.

The Asian musk shrew Suncus murinus is an invasive insectivore that first colonised Mauritius in the eighteenth century. It is a significant predator and poses a threat to terrestrial endemic reptiles in Mauritius. On the islet nature reserve Ile Aux Aigrettes, Mauritius, shrews predate juvenile Telfair’s skink Leiolopisma telfairii, limiting the recruitment of this threatened species. It is therefore important to reduce numbers of Asian musk shrews, and live and fatal trapping are potential methods that can be used to control or eradicate invasive vertebrates. This study tested whether Asian musk shrews preferred the bait currently used for trapping shrews on Ile Aux Aigrettes compared to a novel bait, crushed cockroach. We also tested whether shrews preferred bait in the presence or absence of musk, a chemical attractant. Shrews were observed in a specially designed choice box where a behavioural tally recorded their activity. Their strongest preference was for compartments containing no bait, although they also displayed a significant preference for crushed cockroach in the presence of musk, and a lack of interest in the existing bait. These results suggest that the use of this novel bait plus musk could improve the success of trapping shrews on Ile aux Aigrettes and elsewhere.

 

Rat eradication resulted in prolonged presence of the anticoagulant rodenticide brodifacoum in exposed lizards, likely significantly contributing to the deaths of secondarily exposed raptors up to at least 773 days after bait application.

Possum control and bird recovery in an urban landscape, New Zealand
Johnstone Macleod L., Dickson R., Leckie C., Stephenson B.M. & Glen A. S. (2015), 12, 44-47

In New Zealand invasive brushtail possums Trichosurus vulpecula reduce nesting success of native birds and compete with them for food. As an urban biodiversity initiative, intensive possum control was carried out in a residential area on Napier Hill, North Island. Bird species were monitored using  five-minute point counts, conducted once before the possum control programme and then annually for a further five years afterwards. Significant increases in the relative abundance of bellbird Anthornis melanura and tui Prosthemadera novaeseelandiae were attributed to an increase in food supply due to reduced competiton from possums. Kereru Hemiphaga novaeseelandiae numbers remained relatively stable and a significant decline was recorded in the relative abundance of silvereyes Zosterops lateralis. Management of possum populations will be continued to try to further improve native bird abundance on Napier Hill.

 

The use of helicopters for spreading bait pellets in rodent eradication operations for conservation programmes is increasing. While aerial applications generally have a high success rate at eradicating rodents, operations that tackle extensive areas of steep terrain (slopes >50°) are more challenging, as the effectiveness of spreading bait pellets at the targeted density in these areas is unknown. We undertook an aerial baiting trial on Gough Island, where predation by the non-native house mouse Mus musculus is devastating the globally important seabird populations. It is therefore critical to deliver bait to the island’s large areas of vegetated cliffs that contain burrowing petrels and mice. Using a helicopter and bait hopper we spread non-toxic bait pellets on two areas of coastal cliffs and the adjoining flat ground, and measured the resulting density of pellets using teams of roped climbers and distance sampling. Compared with adjacent flat areas, the vegetated cliff areas retained an average 66-76% of pellets (lower 95% confidence interval 45-60%). While baiting rates on cliffs were lower than adjoining flat areas, the recommended best practice for aerial eradications prescribes applying two additional drops on steep areas. Consequently, current best practice would be sufficient to ensure coverage at densities at or above the targeted baiting rate. While these trials were focused on Gough Island, the results should be useful for eradication operations on other islands with cliffs with similar terrain and vegetation cover.

A quarry pond in Highland, UK, was treated with PyBlast (a biocide derived from natural pyrethrin) to eradicate a population of invasive non-native signal crayfish Pacifasticus leniusculus.  Although it was anticipated that pyrethrin application would lead to the death of all poikilothermic animals present in the quarry pond, its use was sanctioned as surveys did not reveal the presence of any protected or other scarce species. It was assumed that native fauna, including amphibians, would re-colonise from an adjacent pond which was not treated. PyBlast (0.4 mg/l) was applied from 12 to 13 June 2012. Follow-up surveys later in June, and in August and September, found no live crayfish, but established the presence of common toad Bufo bufo tadpoles, and both larval and adult palmate newt Lissotriton helveticus.  All appeared developmentally and behaviourally normal.  These observations suggest that common toad and palmate newt larvae are able to survive levels of Pyblast generally lethal to crustaceans, indicating that amphibian presence at a site should not necessarily halt crayfish eradication programmes.

 

 

A long-running pest management programme was tested to determine to what extent rats were removed from a 50 ha managed area on the shore of Lake Taupo, North Island, New Zealand.  It was confirmed that the trapping protocol employed was effective in catching rats by setting a new trap line in a non-managed area where rats were undisturbed; 64 rats were caught over 10 days. Damage to fake nests, a reliable indicator of the presence of rats, peaked after eight days when 31 of 40 nests were destroyed in one night. The same protocol was then applied in the managed area in comparable forest. Here no rats were caught, and fake nests remained untouched. It was also confirmed that the lack of captures in the managed area was due to effective pest control, rather than to widespread trap avoidance, by using three other methods of monitoring rat presence. It was concluded that the community-led programme was effective in removing rats from the managed area during the nesting season.

 

The ratio of captures to unit effort is an important cost/benefit measure for volunteer pest control programmes. We describe an experiment designed to investigate the use of pre-feeding and trap pulsing as possible means of increasing this ratio. In 20 traps locked-open and pre-fed with non-toxic pellets for five days, the same number of black rats was caught over the next 5 days as in 20 non pre-fed traps set for the whole 10 days (32 rats each). Allowing for successful traps being unavailable for an average of half a night each, the capture rate in the pre-fed traps was 47% over five days, more than double that in the non pre-fed traps set for twice as long (total 19% in 10 days).

 

In New Zealand, invasive non-native mammals threaten the survival of native species such as the North Island brown kiwi (Apteryx mantelli). At Whangarei Heads, in northern New Zealand, community groups are working with local and national government agencies to protect kiwi populations. The abundance of kiwi there has been monitored since 2001 using annual counts of calls. Trapping of invasive mammals began in 2002, and their relative abundance is assessed from annual capture rates. Capture rates of stoats (Mustela erminea), weasels (M. nivalis), cats (Felis catus), rats (Rattus spp.) and possums (Trichosurus vulpecula) have declined significantly since trapping began, suggesting their abundance has been suppressed. Ferrets (Mustela furo) were already scarce when trapping began, and have been reduced to undetectable levels in most years. Numbers of hedgehogs (Erinaceus europaeus) have shown little response to trapping. Kiwi populations were apparently in decline before pest control began, but have since increased. Kiwi call rates in 2011 were the highest so far recorded at Whangarei Heads. Stoats are considered one of the main threats to kiwi, and our data suggest that kiwi numbers remain low unless stoat abundance is reduced below a catch per unit effort threshold of ~0.1 stoat per trap per year.

Campaigns to eradicate introduced rats (Rattus spp.) from small islands are very successful; however, reinvasions on rat-free islands continue to be a major concern. In New Zealand, rodent sniffing dogs are employed to detect suspected rat incursions. The ability to detect and catch a known free-ranging rat on a rat-free island has previously been proven only once and never been experimentally tested. This study tested the ability of a rodent sniffing dog to detect a free-ranging Norway rat R. norvegicus and four caged albino laboratory rats (R. norvegicus) on rodent-free Browns Island. A male Norway rat fitted with a GPS/VHF transmitter was released on the island as part of a trial to test the ability to detect its presence using caged 'lure' rats. A failure to detect any signal from the transmitter forced us to bring in a trained rodent sniffing dog to locate the rat. In a systematic search of the island, the dog found three of the four caged rats, and through air sniffing downwind it was able to track and catch the wild rat from a distance of approximately 170 m. This is one of the very few times that a rodent sniffing dog has been tested in a realistic scenario in which there was a confirmed free-ranging rat on an otherwise rodent-free island. The successful detection and capture indicates that trained sniffing dogs can contribute to the detection of rat incursions on island sanctuaries and assist in rat control.

Two types of anticoagulant rodenticides have proven successful at eradicating invasive rats and mice from islands. Brodifacoum is the most commonly used and has a low failure rate both when delivered from the air and from ground-based systems. It does, however, present a risk to non-target animals such as birds. When such risk is not acceptable or cannot be mitigated, diphacinone has been favoured by some managers because it is less toxic to birds and less persistent in rodents. However, unlike brodifacoum, diphacinone requires a rodent to eat several baits over several days to ingest a lethal dose. This increases the risk that not all rodents will be killed. When data on attempts to eradicate rats and mice for both aerial and ground-based methods are combined, brodifacoum has a significantly lower failure rate at 17% (54 of 322 attempts) than diphacinone at 33% (13 of 39 attempts). The difference is more significant when just rats are considered. Ground-based methods show similar failure rates for both rodenticides, but to date the very few attempts using aerially sown diphacinone baits have had a high failure rate compared with that for brodifacoum.

Mass trapping successfully achieved elimination of black rat Rattus rattus on Feno islet (1.6 ha), Terceira island (Azores archipelago), thus enabling roseate terns Sterna dougallii and common terns Sterna hirundo to recolonize the islet. Rats were first detected on Feno in 2003, when tern breeding-numbers had decreased dramatically. During 2005 no terns bred on the islet and in 2006 fewer than five common tern pairs attempted to nest. Rat eradication was initiated in September 2006. The last rats were captured in March 2007. Monitoring conducted in September 2007, and May and September 2008 indicated that rats had not recolonized. Common terns quickly resumed breeding on Feno islet but numbers (c.120 pairs in 2009) are still below peak levels (c.240-280 pairs) recorded before rat infestation. Roseate terns on the other hand were slower to return but recovered faster with around 260 pairs in 2009, representing 22% of the Azores population. The success of the black rat eradication shows that surveillance and timely action are fundamental to conserve tern colonies vulnerable to rat predation in the Azores.

Introduced common mynas Acridotheres tristis have been implicated as a threat to native biodiversity on the oceanic islands of St Helena and Ascension (UK). A rice-based bait treated with Starlicide® was broadcast for consumption by flocks of common mynas at the government rubbish tips on the two islands during investigations of potential myna management techniques. Bait was laid on St Helena during two 3-day periods in July and August 2009, and on Ascension over one 3-day period in November 2009. As a consequence of bait ingestion, dead mynas were found, especially under night roosts and also at the main drinking area on Ascension, following baiting. On St Helena early morning counts at the tip suggested that whilst the number of mynas fell after each treatment, lower numbers were not sustained; no reduction in numbers flying to the main roost used by birds using the tip as a feeding area was detected post-treatment. On Ascension, the number of mynas that fed at the tip and using a drinking site, and the numbers counted flying into night roosts from the direction of the tip, both indicated declines of about 70% (from about 360 to 109 individuals). Most dead birds were found following the first day of bait application, with few apparently dying after baiting on days 2 and 3. Despite the low concentration of Starlicide used, aversion to the bait was apparent during the trials. These results indicate that Starlicide may contribute to myna control programmes but questions remain over the mode of action of the chemical (in terms of individual differences among birds the responses to its toxic properties) and the longer-term susceptibility of birds to baiting.

Aerial application of cereal bait containing the poison brodifacoum was used to eradicate house mice Mus musculus from three islands in the Abel Tasman National Park, New Zealand, during July and August 2007. Poison bait was spread onto the islands by a helicopter using an under slung bait-spreading bucket and applied at a rate of 4 kg per hectare with a 50 % overlap in swath width. This gave an effective application rate of 8 kg per hectare. Bait was applied to the islands in two separate applications, 31 days apart. Monitoring for mice was carried out on all three islands for two years following the eradication attempt to determine the outcome. No mice or any sign of mice was detected. Eradicating mice from Tonga Island, Adele Island and Fisherman Island was the final step towards making the three islands free of introduced mammalian predators.

Attempted eradication of house sparrows Passer domesticus from Round Island (Mauritius), Indian Ocean
Bednarczuk E., Feare C.J., Lovibond S., Tatayah V. & Jones C.G. (2010), 7, 75-86

In 1982, house sparrows Passer domesticus were confirmed as having established a naturalized population on Round Island (Mauritius). A planned pending translocation of an endangered Mauritian endemic bird, Mauritius fody Foudia rubra to Round Island suggested eradication of sparrows to be pertinent as they were potentially a resource competitor and vector of parasites and pathogens. An attempted eradication using a combination of techniques was undertaken from 19 August 2008 to 25 February 2009. Following food preference trials, microwave-sterilized millet seed was used as bait for trapping and for narcotisation with alphachloralose. House sparrows were also shot, caught in mist nets and on glue sticks, and some nests and chicks were removed. In total, 320 house sparrows were killed, with trapping accounting for 87% (277) of birds removed. However, the population was not eradicated. The assumption that the Round Island house sparrow population was derived from one storm-driven event and is closed to further immigrants needs to be investigated in order to determine whether long-term eradication is in fact feasible. Suggestions for improving the prospects for eradication or ongoing management of the population are presented.

The Kermadec red-crowned parakeet Cyanoramphus novaezelandiae was driven to extinction on Raoul Island over 150 years ago by introduced cats Felis catus and rats (Rattus norvegicus and R.exulans). These predators were eradicated from the island (2,938 ha) between 2002-04 during the world's largest multi-species eradication project. In 2008 we documented a unique recolonisation event when parakeets were observed to have returned to Raoul, presumably from a nearby island group, The Herald Islets (51 ha). We captured and aged 100 parakeets, of which 44% were born in 2008, and breeding was observed on Raoul Island. This represents the first evidence of nesting of this species on Raoul Island since 1836. Our findings highlight the global conservation potential for island avifaunas by prioritising eradication areas through consideration of proximity of remnant populations to target management locations, instead of the classical translocation approach alone. The natural recolonization of parakeets on Raoul Island from a satellite source population is to our knowledge, a first for parrot conservation and the first documented population expansion and island recolonization of a parrot species after removal of invasive predators.

As part of a larger project attempting to reduce predation pressure from introduced rats Rattus spp. on native fauna and flora on several islands off Pohnpei main island (Pacific Ocean), a rat eradication program was undertaken on the small island of Nahkapw (1.58 ha). Prior to commencement of eradication, a pre-treatment rat population assessment was undertaken; only a single female ship rat R.rattus was caught (subsequently fitted with a radio-collar and released) and wax indicator success was also very low, suggesting that the island hosted a very low density of rats. Bait-stations were positioned on the ground along two transects that ran the length of the island spaced approximately 20 m apart; the distance between the two transects varied but was never less than 10 m or more than 30 m. After 2-weeks of inactivity, these stations were supplemented with 22 bait stations attached to tree trunks adjacent to each ground-based bait station. These tree bait stations were maintained until the radio-collared rat died (25 days after initial bait placement). There was no evidence of non-target species being adversely affected by the bait.

As part of a larger project attempting to reduce predation pressure from introduced rats Rattus spp. on native fauna and flora on several islands off Pohnpei main island (Pacific Ocean), a rat eradication program was undertaken on the small island of Dekehtik (2.63 ha) where Pacific rats R.exulans were present. The island was systematically hand-broadcast with rodenticide bait at a pre-determined application rate of 50 kg/ha over one day. Radio-tracking revealed that rats spent a significant amount of time in the forest canopy. Therefore, to ensure that rats living in the canopy had a good chance of encountering bait, bait clusters (bolo baits) were fabricated and catapulted into one third of the palm trees on the island. Pre-bait application trap success (rat captures/trap nights) was 52%. Post-bait application trap success (as recorded 12-15 days after bait application) was 0%. Although rat chew marks were found on one wax indicator block 12 days after the bait application, subsequent monitoring detected no rats. No non-target species appeared adversely effected by the rodenticide bait.

As part of a larger project attempting to reduce predation pressure from introduced rats Rattus spp. on native fauna and flora on several islands off Pohnpei main island (Pacific Ocean), a rat eradication program was undertaken on the small island of Pein Mal (2.17 ha) where ship rats R.rattus were present. The island was systematically hand-broadcast with rodenticide bait at a pre-determined application rate of 50 kg/ha over one day. Radio-tracking revealed that rats spent a significant amount of time in the forest canopy. Therefore, to ensure that rats living in the canopy had a good chance of encountering bait, bait clusters (bolo baits) were made and catapulted into one third of the palm trees on the island. Mangrove forest surrounds the shore of Pein Mal where rats were also present. Throughout this habitat to ensure rat access to the rodenticide, a bait station grid was established with bait stations nailed to tree trunks approximately 2 m above the high water mark. Pre-bait application trap success (rat captures/trap nights) was 39%. Post-bait application trap success (as recorded 10-14 days after bait application) was 0%. Pre-bait application wax indicator success was 53%. Post-bait application wax indicator success was 0%; subsequent monitoring 6-months later detected no rats. No non-target species appeared adversely effected by the bait.

 

A field trial showed that a newly designed, post- or tree-mounted 'hockey stick' rat bait dispenser incorporating 20 g fixed bait blocks, was both more practical and more efficient in terms of rodenticide bait use, compared to a traditionally used 'ground pipe' dispenser containing loose, 5 g bait blocks. Bait in the hockey stick dispenser was less affected by mould and slower to breakdown, therefore making it more effective for longer and reducing the bait replacement rate required compared to the old design. It was also considered that there was also less incidental bait take by both introduced A field trial showed that a newly designed, post- or tree-mounted 'hockey stick' rat bait dispenser incorporating 20 g fixed bait blocks, was both more practical and more efficient in terms of rodenticide bait use, compared to a traditionally used 'ground pipe' dispenser containing loose, 5 g bait blocks. Bait in the hockey stick dispenser was less affected by mould and slower to breakdown, therefore making it more effective for longer and reducing the bait replacement rate required compared to the old design. It was also considered that there was also less incidental bait take by both introduced giant land snails Achatina spp., and of greater concern, endemic snails.

On a Fijian island a decision was made to remove feral dogs Canis familiaris as they were predating upon endangered fauna. Rather than trapping or shooting, by offering food and gradually habituating them to people, almost all were captured unharmed and moved to the main island of Fiji.

 

The critically endangered Antiguan racer Alsophis antiguae, used to be abundant throughout the Lesser Antillean islands of Antigua (and its satellite islands) and Barbuda. Non-native black rats Rattus rattus were identified as a serious predator of the snake on Great Bird Island, therefore the decision was made to eradicate the rats. A poison-baiting programme proved successful, with the racer population more than doubling in only 18 months in response. Other fauna, including several species of seabirds and hawksbill turtle Eretmochelys imbricata, also benefited greatly from rat removal.

Following black rat Rattus rattus eradication, the Antiguan racer Alsophis antiguae population on Great Bird Island increased by over 300% over the next nine years.

Mink rafts positioned along stretches of the River Wensum proved very successful in terms of American mink Mustela vison detection and trapping effectiveness. Survey results indicated that there has been an expansion in the range of water voles Arvicola terrestris along the river from 2003 to 2005, perhaps in response to the removal of American mink.

A combination of rat Rattus baiting, trapping and remote fired detonators was used to protect the eggs of the endangered flightless parrot, the kakapo Strigops habroptilus, from rat predation. The detonators were twice fired in the presence of rats, which left on both occasions. The incubating kakapo showed no evidence of reacting to detonations.

For The Chicken Islands group (New Zealand) a pacific rat Rattus exulans eradication programme was undertaken in an attempt to establish a safe offshore haven for many rare native species. Eradication of the introduced rats was successful, and in response, populations of several native reptiles have increased and seabird fledging success has increased dramatically.

 

Following removal of Norway rats Rattus norvegicus and rabbits Oryctolagus cuniculus from Motuhora (Whale Island), 32 adult tuatara Sphenodon punctatus were introduced in 1996. They produced at least two clutches of offspring, and about 50 individuals were present when surveyed in 2005.

Ship rats Ratus rattus were eradicated from the 9.3 ha Goat Island in 1994; however, rats were redetected in 1996. From April to June 2005 using between 35 and 51 traps were deployed. Subsequent to trapping, 49 poison bait stations were established across the island on 23 June 2005 to assess if this eradication had been successful; only one was touched. Tracking tunnels and waxtags were left on the island, but with no signs of use. It seems that eradication was achieved. Gnaw marks were subsequently discovered in a waxtag at the site where reinvasion was most likely but there were no further signs over the next two months.