MANAGEMENT doi: 10.1111/emr.12422 REPORT Evaluation of intervention aimed at improving reproductive success in Orange-bellied Parrots Neophema chrysogaster: Lessons, barriers and successes

By Dejan Stojanovic, Catherine Young, Shannon Troy and Robert Heinsohn

Summary Assessing feasibility and identifying constraints that affect project imple- mentation is a crucial step for planning long-term species recovery actions for field-based programs. We report on the outcomes of a conservation intervention on the most endan- gered parrot in the world, the Orange-bellied Parrot (Neophema chrysogaster). We aimed to trial new techniques to increase reproductive success of wild nests and address key knowledge gaps. We aimed to achieve higher reproductive success using (i) intervention – where fertile eggs or nestlings would be fostered from captivity to wild nests that suffered infertility or had small brood sizes and (ii) rescue – where wild-born nestlings would be removed from nests if they were ailing and either fostered to another nest or hand reared to improve their survival. Our project provided proof of principle that it is possible to imple- Dejan Stojanovic: Fenner School, Australian ment intensive, individual-level monitoring and intervention (via fostering of nestlings to National University, Acton, ACT, Australia infertile nests) to address reproductive problems for the Orange-bellied Parrot. However, (Email: [email protected] ). we also found important factors that hindered our ability to achieve project aims (manage- ment of biosecurity), and identified unexpected factors that have important implications for Catherine Young: Fenner School, Australian future application of these techniques (nest abandonment from video camera deployment, National University, Acton, ACT, Australia. rapid death of unhealthy nestlings hindering rescue attempts). Our project tested tech- Shannon Troy: Department of Primary niques and tools to provide new approaches for fighting extinction of the Orange-bellied Industries, Parks, Water and Environment, Parrot, and yielded important new information about the species ecology and management Hobart, Tas, Australia. Robert Heinsohn: options. Fenner School, Australian National University, Key words: Conservation intervention and management, disease management, fostering, Acton, ACT, Australia.. hand-rearing and captive management, nest monitoring, reproductive management.

2010; Doherty & Ritchie, 2017). Part of this captivity to wild nests that suffered Introduction process includes assessing feasibility and infertility or had small brood sizes, and onservation interventions to manage identifying constraints that affect project 2 rescue – where wild-born nestlings Cthreatened species can be critical to implementation, which is a crucial step would be removed from nests if they population recovery. The most effective for planning long-term species recovery were ailing and either fostered to species recovery projects identify clear fac- actions (Walls 2018). another nest or hand reared to improve tors that are driving decline and implement We report on the outcomes of a conser- their survival. targeted conservation action to remedy vation intervention on the most endan- these threats and alleviate pressure on pop- gered parrot in the world, the Orange- These approaches are not in them- ulations (Scheele, et al. 2018). However, bellied Parrot (Neophema chrysogaster). selves novel because they have been suc- clear diagnosis of threats is not always pos- The decline of the wild population of this cessfully trialled and implemented on sible for species lacking detailed data on species to only 3 wild-born females and 13 other species (for a summary of tech- ecology and demographic processes males in 2016 triggered this project (Sto- niques and case studies see Jones 2004). (Bland, et al. 2015). In such cases, by neces- janovic, et al. 2018). We aimed to trial However, past applications of intensive sity managers might implement conserva- new techniques to increase reproductive conservation management have most tion actions iteratively by means of trial success of wild nests and address key often been applied to sedentary or and error in an adaptive management knowledge gaps. We aimed to achieve island-dwelling birds. Migratory species framework (Gerber & Kendall, 2018). Clear higher reproductive success using two like the Orange-bellied Parrot pose sub- performance metrics that relate to the focal approaches: stantial conservation challenges because ecological process targeted by the interven- they live in very remote, difficult to access 1 intervention – where fertile eggs or tion are crucial for effective evaluation of locations across very large geographical nestlings would be fostered from conservation interventions (Wintle, et al. ranges. In addition, the project sought to

ª 2020 Ecological Society of Australia and John Wiley & Sons Australia, Ltd ECOLOGICAL MANAGEMENT & RESTORATION 1

Ecological Society of Australia MANAGEMENT REPORT trial video monitoring as an approach to https://fieldcapture.ala.org.au/;jsessionid= evaluated the efficacy of nest competitor increase the resolution of nest monitoring FE76A594F6E3D265018D76D295501F88) control (Stojanovic, et al. 2019). and improve capacity to achieve interven- because this approach is widely utilized On average per year, the project cost tion and rescue. These aims were under- for evaluating on-ground environmental approximately $20,000 AUD for travel to taken in context of ongoing projects in Australia. Project outcomes and from the field site, $50,000 for person- complementary work on Orange-bellied and monitoring indicators are presented nel, $5,600 for disease screening of cap- Parrots by the Tasmanian Government in Table 1. Nest boxes were checked tive birds for BFDV (comprising ~$280 in and their collaborators in the species either manually by climbing trees or by tests per bird), plus capital expenditure recovery team (Department of Environ- reviewing recordings from video cameras ($15,000 for video cameras, $3,000 for ment Land Water & Planning 2016; Troy (model HK101182w www.handycam.c purchase of additional nest boxes). & Hehn, 2019). Here, we summarize the om) mounted inside or on the outside of aims, rationale, methods and results of boxes (external cameras were Reconyx Discussion the conservation interventions we trialled. Hyperfire HC600). We checked egg fertil- We also explain barriers to success and ity by candling eggs using a small flash- Our project provided proof of principal limitations of our approach, in the hope light. We monitored nestling condition that it is possible to implement intensive, that these factors can help inform other (for the rescue component of the work) individual-level monitoring and interven- similar projects on threatened species. using the approach described by Sto- tion to address reproductive problems janovic et al. (2020) for estimating body for the Orange-bellied Parrot. Other pro- condition. We tested all captive animals jects have applied similar efforts to other Methods selected for use as donors of eggs or nest- species (Jones 2004), but this is the first The project ran between September 2016 lings for Psittacine beak and feather dis- time these approaches have been and June 2019, spanning three field sea- ease virus (BFDV) from blood samples attempted for the Orange-bellied Parrot. sons overlapping the Orange-bellied Par- (Troy & Kuechler 2018). Our results are important because lessons rot breeding season (Sep – Mar). We around ways to improve management of present our aims, monitoring approaches threatened species typically go unpub- Results and evaluation criteria based on the Aus- lished, and we hope to make ‘reinventing tralian Government Monitoring Evaluation Table 1 summarizes our findings against the wheel’ unnecessary for other projects Reporting and Improvement Tool (MERIT: project objectives. We also identify key seeking to deploy similar actions in barriers, lessons, costs and benefits for remote field sites for difficult species like each method used in the project in the Orange-bellied Parrot. We found and Table 1. Below we summarize our evalua- report important factors that hindered tion of success against the project out- our ability to achieve project aims and comes and monitoring indicators. In identified unexpected factors (e.g. poten- total, we monitored 53 nests (i.e. a census tial sensitivity of parrots to deployment of all breeding events in the contemporary of video cameras after nesting has begun) population), deployed video recorders that have important implications for inside 15 nest boxes, attempted interven- future application of these techniques. tion to correct egg infertility on 4 nests, Based on the severity of the barriers, ben- rescued 5 nestlings either by fostering or efits and costs of each method, we discuss hand-rearing (Fig. 1), and found ways to which elements of the project may be work around the challenges posed by worth incorporating or excluding from the remoteness of the field sites (Fig. 2). future management efforts. Unfortunately, most nestlings fostered in Useful components the intervention component of the project (including caveats) died (only one of five nestlings survived to migrate), but these results were skewed We provide proof of principle that nest- by a disease outbreak unrelated to this ling fostering to control brood sizes and project (Stojanovic, et al. 2018). Disease correct infertility are methods that could risk management was a major challenge improve reproductive success of the Figure 1. This Orange-bellied Parrot nestling was identified as underweight using during implementation of the project, Orange-bellied Parrot. However, interven- a body condition index developed during this and outbreaks of disease directly hindered tion by fostering is dependent on having project and was rescued by fostering to our project objectives in 2 of 3 years. donor captive nests that meet disease another nest. This intervention resulted in a Based on our index of nestling body condi- screening requirements (i.e. no evidence 12.1 g improvement of body condition, and tion (Stojanovic, et al. 2020), we identified of BFDV) and logistic and timing con- the nestling survived to fledge. candidate nestlings for rescue. We also straints on the availability of donors of

2 ECOLOGICAL MANAGEMENT & RESTORATION ª 2020 Ecological Society of Australia and John Wiley & Sons Australia, Ltd MANAGEMENT REPORT

nests and individual animals. However, we suggest that to derive the maximum benefit from this method, personnel should review camera footage daily (preferably twice daily) so that rescue hap- pens before nestlings die. Because of the 3-5 day intervals between manual checks of videos in our project, recordings mostly served to confirm when and (sometimes) how mortalities occurred. Had more regu- lar checks occurred, some of these mortal- ities may have been prevented. Video deployment midway through incubation may have caused failure of two nests due to abandonment. In 2018 (in response to Figure 2. Orange-bellied Parrots nest in remote field locations. This photograph depicts the the first nest abandonment), we deployed ‘Logistics Hub’ at Pandora’s Hill, where critical field equipment is stored in situ to reduce the need dummy cameras in all boxes to habituate to carry heavy field equipment long distances over rough terrain. parrots to the hardware, but one of 10 mothers abandoned her nest despite this modified approach. We note though that eggs/nestlings. The presence of disease in was not tested in this project). Fostering three other nests (without cameras) were the captive population may have compro- fertile eggs from captive to wild nests abandoned in that same week (a period mised the results of fostering in 2016 (at mid incubation may be a more effective of heavy rainfall), so we are uncertain of least two foster nestlings may have died method of increasing nest productivity the true cause of nest failure. Neverthe- from Pseudomonas infection) (Stojanovic, than using nestlings. However, further less, if the potential risk of nest abandon- et al. 2018), and prevented it altogether in evaluation of biosecurity risks of moving ment from deploying cameras during 2018 (BFDV detected in donor parents, eggs between nests must be evaluated. incubation is unacceptable, cameras could halting captive to wild transfer of nest- During this project, we developed a be deployed in all boxes before breeding lings) (Troy & Kuechler 2018). In both body condition index as a way to evaluate begins. Those boxes eventually occupied of those years, there were opportunities nestling condition of Orange-bellied Par- by parrots could have the necessary addi- for fostering in the wild due to infertility rots (Stojanovic, et al. 2020). This was a tional hardware for functionality (solar of eggs, but this was not attempted due useful, empirical way of assessing which panels and battery) assembled from the to biosecurity precautions. nestlings might benefit from intervention ground once nest box occupancy is con- In the long term, further research to and rescue. It also provided a useful firmed, with no disturbance to incubating understanding the demographic impact means of evaluating the impacts of this parrots. This approach is more expensive of disease should be a high priority in intervention. The method is fast and rela- (because of the need to purchase a real the Orange-bellied Parrot so that disease tively repeatable among observers and camera for every box), but will greatly mitigation is managed to reduce impacts reduced uncertainty about how to identify reduce the risk of abandonment (because on other aspects of management. Similar when a nestling is underperforming. We dummy cameras were ignored by all par- evaluations for other species have been suggest that nestlings that fall below one rots). valuable in weighing the impact of disease standard deviation for first or middle- Components that could be on management actions (Tollington, et al. hatched nestlings (Stojanovic, et al. discontinued 2015; Fogell, et al. 2019). We suggest that 2020) could be considered for interven- nestling fostering may be a useful way to tion/rescue. If such nestlings are identified Unless more frequent manual checking maximize reproductive success of the before they die, and exhibit no symptoms can be achieved without nest disturbance, wild population if disease risk and biose- of infectious disease, rescue could be the rescue component of our project curity protocols are modified to: (i) implemented via fostering to other wild could be discontinued, or implemented address the likelihood of exposure nests with small broods (if available). In opportunistically (e.g. when underper- through both vertical and horizontal trans- cases where suitable host nests are not forming nestlings are discovered early mission to eggs and nestlings, (ii) identify available, rescuing nestlings by hand-rear- and can be fostered to other nests). The methods to mitigate risk while maximizing ing them has major limitations (see logistic, financial and personnel con- available management options. Fostering below). straints that currently prevent permanent of nestlings was achieved, but further eval- Video monitoring of nests has the deployment of staff at the field site are uation of whether fostering eggs could potential to yield large volumes of high- unlikely to be overcome without substan- improve outcomes is worthwhile (this resolution data on the performance of tial new funding. After rescue is

ª 2020 Ecological Society of Australia and John Wiley & Sons Australia, Ltd ECOLOGICAL MANAGEMENT & RESTORATION 3 4 REPORT MANAGEMENT

Table 1. Summary of the approaches we took, their efficacy and lessons learned during the Orange-bellied Parrot project. CLGCLMNGMN RESTORATION & MANAGEMENT ECOLOGICAL Desired years Monitoring Evaluation Barriers Key Lessons Benefits of the Costs of the outcome implemented approach approach approach Nest boxes 2016–2018 1) record the ID of All nest boxes were Video camera Dummy cameras Improved resolution of Risk of nest cleaned and boxes cleaned and cleaned annually; deployment could not (match boxes nest monitoring at abandonment (2/15 video cameras fitted with cameras video cameras be ruled out as a wrapped in black reduced effort; new nests); high per unit installed installed on active cause of two nest tape) were installed in information about cost of video cameras nests (206 n = 0, abandonments (one all nest boxes in 2018 breeding behaviour 2017 n = 5, 2018 each in 2017, and to enable derived from high- n = 10) 2018) acclimatization of quality video footage birds to cameras (Video S1) Early 2016–2018 1) Determine parent All nests monitored Personnel costs are Parent provenance High quality, detailed High workload for identification of provenance (captive- (2016 n = 17, 2017 high for the frequent difficult to ascertain data on nest timing, frequent nest nest or wild-born) to n = 14, 2018 n = 22); monitoring required to with poor camera early identification of checking, requiring productivity and ensure only nests of egg fertility determine nest timing focus, easier with infertility provides extensive climbing identification of captive-born mothers monitored; nestling and fertility, but costs direct observations. opportunity to effort (minimum four foster subject to fostering; survival and body alleviated by video Identification of improve productivity climbs per nest to candidates 2) Timing of important condition monitored cameras because the mothers inside the in low fertility nests. clean box, confirm nest phases deter- interval between nest box is disruptive. Individual nest initiation, check mined; 3) Identify manual checking is Body condition index identification of egg fertility and ring infertile eggs within longer without loss of is costly to develop mothers very nestlings). 14 days of egg laying; information due to but removes accurate. Reduced Development of body 4) develop an index of video recordings. ambiguity of uncertainty for condition index is body condition for Personnel costs high subjective health identifying when labour intensive and nestlings to empiri- to identify fathers. assessments. nestlings are unwell requires very frequent cally evaluate nestling and need to be handling of animals quality and identify fostered or rescued. for model fostering candidates development Stojanovic, et al .

ª (2020) 00Eooia oit fAsrlaadJh ie osAsrla Ltd Australia, Sons & Wiley John and Australia of Society Ecological 2020 Reproductive 2016–2018 1) implement Intervention had Intervention was (1) Biosecurity is Intervention showed The feasibility of success intervention (fostering mixed results (2016 – dependent on the important but the (limited) proof of intervention is subject improved by of nestlings to correct 1/5 nestlings availability of donor impacts of disease on concept that fostering to several unrelated implementing egg infertility or successfully fostered, nests in captivity to demographic is a potential tool for limiting factors (need, intervention nestling mortality); 2) 2017 - no fostering supplement wild parameters must be increasing timing, biosecurity, and rescue implement rescue necessary due to high nests. Donor nest weighed against the reproductive output. funding) and in two of actions (ailing nestlings fertility in the wild, availability limited by: impact of biosecurity Rescues showed that three years, these identified by body 2018 - fostering (1) biosecurity, for protocols reducing nestlings benefit from factors prevented condition and either abandoned due to example disease scope for on-ground being removed from implementation of removed for hand- BFDV in captive outbreaks in 2016 and intervention; (2) high nests where they are interventions. Rescue rearing or fostered to donor parents). 2018 stopped transfer personnel costs may underperforming if had limited success other wild nests) Rescue was of birds from captivity be unavoidable if they can be detected when the interval attempted in 2016 to wild, affecting rescue is to improve early enough before between checks was (one nestling) and project survival of nestlings, they die. long, but smaller 2018 (four nestlings) implementation and with at least twice intervals between but in all years most outcomes; (2) timing, daily nest checks (via checks will equate to candidates for rescue e.g. captive and wild video or manually) a substantial increase died before they were nests must be necessary to improve in personnel costs and found (2016 - six dead initiated in close odds of detecting disturbance to nestlings, 2017 - three synchrony with one underperforming animals. MANAGEMENT REPORT

completed, hand-rearing nestlings is labour intensive if suitable foster nests are not available. In one case during 2018, a nestling that was successfully res- cued and fostered to an available wild nest had to be removed along with its foster nest mate because their (captive-born) Costs of the approach mother plucked their feathers (the first time this has been recorded in the wild). This example shows that irrespective of whether suitable host nests are available, careful monitoring of rescued nestlings is crucial to ensure the success of interven- tion. Unless suitably skilled personnel are

approach deployed permanently in the field during the breeding season, even checking inter- vals of 3 days can be too long to intervene if something goes wrong in a nest. How- ever, if skilled staff are always present, video monitoring and more regular check- ing may make rescue worth trialling again.

nestlings before they die; (3) Fostering eggs worth considering as an alternative approach to nestlings, which had mixed success. Consideration should be given to evaluat- ing the cost versus benefit of rescuing a given nestling relative to its importance to the population. If the nestling is from a genetically valuable lineage, then more intensive interventions (e.g. hand-rearing) may be justifiable, compared to another nestling whose lineage may be over-repre- another to be potential candidates for fostering. Rescue was highly dependent on the frequency of monitoring, with most ailing nestlings dying in the 4-day nest check intervals. Cameras improved monitoring detail but still required personnel to manually review video and detect problems. Logistic issues with site access made regular manual checking challenging (Figure 2). sented in the population. 1,

– Conclusion Our project tested techniques and tools to 2) but two

– provide new approaches for fighting extinction of the Orange-bellied Parrot. dead, 2018 - sixteen dead). One nestling rescued in 2018 showed a 12.1 g improvement in mass over 14 days after fostering to a new nest (Figure 1). Another fostered nestling temporarily gained weight but was later feather plucked by the host mother and so was takenhand-rearing. for Three nestlings were hand reared, over the project (2016 2018 were ultimately euth- anized. Evaluation Barriers Key Lessons Benefits of the Our study has yielded important new information about the species ecology (Stojanovic, et al. 2018; Stojanovic, et al. 2019; Stojanovic, et al. 2020) and pro- vided managers with new options for data collection and intervention to address reproductive problems facing the species. We hope our study provides a useful tem- Monitoring approach plate for practitioners to trial these tech- niques and evaluate their efficacy on other species.

years implemented Acknowledgements This project was funded by 1609 support-

(Continued) ers of the crowd-funding campaign ‘Oper- ation OBP’. Special thanks to Henry Cook for creating and running the funding cam-

Table 1. Desired outcome paign. Thanks also to Fernanda Alves,

ª 2020 Ecological Society of Australia and John Wiley & Sons Australia, Ltd ECOLOGICAL MANAGEMENT & RESTORATION 5 MANAGEMENT REPORT

Laura Rayner, Ross Crates, Matt Webb, of data-deficient species. Conservation Biol- critically endangered orange-bellied parrots – Matt Eyles and Ben Oliver. Thanks to the ogy 29, 250 259. Neophema chrysogaster. Emu - Austral Department of Environment Land Water and Plan- Ornithology 120, 135–141. others that assisted in the field work, and ning (2016) National Recovery Plan for the Tollington S., Greenwood A., Jones C. G. et al. the species recovery team for their sup- Orange-bellied Parrot Neophema chryso- (2015) Detailed monitoring of a small but port. We are grateful to Multiple artists gaster. Australian Government, Canberra. recovering population reveals sublethal Doherty T. S. and Ritchie E. G. (2017) Stop jump- effects of disease and unexpected interac- and businesses contributed to our fund- ing the gun: a call for evidence-based invasive tions with supplemental feeding. Journal of raising efforts. Airlines of Tasmania (Par predator management. Conservation Letters Animal Ecology 84, 969–977. – Avion), Bob Brown, Bruny Island Cheese, 10,15 22. Troy S. and Hehn K. (2019) Report on the Mela- Fogell D. J., Groombridge J. J., Tollington S. et al. leuca wild population 2018/19. In: Tasmanian Bruny Island Honey, Bruny Hotel, Eggpic- (2019) Hygiene and biosecurity protocols Orange-bellied Parrot Program, Tasmanian nic, First Dog on the Moon, Paul Fleming, reduce infection prevalence but do not Department of Primary Industries, Parks, Leila Jeffries, Inala Nature Tours, Jon improve fledging success in an endangered Water and the Environment. parrot. Scientific Reports 9, 4779. Troy S. and Kuechler A. (2018) Report on the Tas- Kudelka, Get Shucked Oysters and David Gerber B. D. and Kendall W. L. (2018) Adaptive manian Wild Population 2017/18. In: Depart- Pope provided free artworks and/or gifts management of animal populations with sig- ment of Primary Industries, Parks, Water and to support the crowd funding campaign. nificant unknowns and uncertainties: a case Environment. study. Ecological Applications 28, 1325– Walls S. C. (2018) Coping with constraints: Par Avion also provided free transport 1341. achieving effective conservation with limited for 2 years. The Tasmanian Government Jones C. (2004) Conservation management of resources. Frontiers in Ecology and Evolution Department of Primary Industries, Parks, endangered birds. In: Bird Ecology and Con- 6, 24. servation. (eds W. Sutherland, I. Newton Wintle B. A., Runge M. C. and Bekessy S. A. Water and Environment and the Australian and R. Green) pp.269–301. Oxford University (2010) Allocating monitoring effort in the face Government’s National Environmental Press, Oxford. of unknown unknowns. Ecology Letters 13, – Science Program through the Threatened Scheele B. C., Legge S., Armstrong D. P. et al. 1325 1337. (2018) How to improve threatened species Species Recovery Hub. The research was management: An Australian perspective. conducted with approval from the Aus- Journal of Environmental Management 223, Supporting Information – tralian National University Animal Ethics 668 675. Stojanovic D., Alves F., Cook H. et al. (2018) Fur- Additional supporting information may be Committee (A2016/48) and the Tasmanian ther knowledge and urgent action required to found online in the Supporting Informa- Department of Primary Industries, Parks save Orange-bellied Parrots from extinction. – tion section at the end of the article. Water and Environment (TFA17037). Emu - Austral Ornithology 118, 126 134. Stojanovic D., Young C. M. and Troy S. (2019) Video S1 Example video from cameras Efficacy of intervention to relieve nest box competition for Orange-bellied Parrot Neo- mounted inside an Orange-bellied Parrot References phema chrysogaster. Ecological Management nest box. Here a mother parrot can be & Restoration 21,66–68. Bland L. M., Collen B., Orme C. D. L. and Bielby seen feeding a young nestling. Stojanovic D., Alves F., Webb M. et al. (2020) J. (2015) Predicting the conservation status Nestling growth and body condition of

6 ECOLOGICAL MANAGEMENT & RESTORATION ª 2020 Ecological Society of Australia and John Wiley & Sons Australia, Ltd