Experimental learning: fox baiting for malleefowl conservation

[This post is an expanded version of the article in the latest edition of ‘Around the Mounds‘, the newsletter of the National Malleefowl Recovery Team]

It is not every day that one adds a new megapode species to the life list! I’m spending a few days on Tetepare Island, a great conservation initiative to preserve the largest uninhabited island in the Solomons, and it has Melanesian megapodes (Megapodius eremita). Brownish, inconspicuous and rather small, they roam around the traditional palm leaf house that serves as lodge. The species resorts to external sources of heat to incubate the eggs, but unlike its cousin the malleefowl it tends to use piles of decomposing vegetation or volcanically-heated sandy soils.

Melanesian megapode

Clockwise from upper left: a) A shy Melanesian megapode (Megapodius eremita), a relative to the orange-footed scrubfowl (Megapodius reinwardt) that inhabits Australia. b) Mangrove monitor (Varanus indicus) digging eggs from a mound in Tetepare island. c) Megapode eggs on sale at Gizo market (S$7 = A$1). d) Man-made megapode hatcheries in volcanic soil on Simbo island, used to stimulate egg production for harvesting.

This morning, I found a monitor lizard digging megapode eggs from a mound. This reminded me of foxes and malleefowl, and I started reflecting on the progress that the Adaptive Management team has made over this year. Like the monitor lizard I was observing, we know for a fact that foxes prey on malleefowl. There is even very graphic evidence from camera traps of foxes digging eggs from mounds. From the conservation management point of view, however, the real question is whether such predation compromises the long term survival of malleefowl, and whether fox baiting – one of the main conservation actions currently undertaken – actually contributes to the recovery of the species. The evidence in the scientific literature regarding these questions is rather mixed.

Adaptive management is about using management actions to learn how we can improve future management decisions. In the case of foxes and malleefowl, a carefully planned “experiment” that compares malleefowl breeding activity at sites where baiting is carried out with similar sites where it isn’t, could help settle whether fox baiting is an effective way of protecting malleefowl. Indeed, the ideal way of conducting such an experiment would be to find pairs of sites where: a) one site can be baited and the other left unbaited as a reference; b) malleefowl mound activity (the ‘response’ of the species) is being monitored or monitoring can be started; and c) paired sites are close enough to share the same environmental variability (such as how much rain falls in a given year), but not so close that baiting can affect what happens at the non-baited site.

The adaptive management team set out to plan and organize such an experiment across the species’ range, with the help and support of Tim Burnard and Joe Benshemesh. Finding pairs of sites that fulfil these four conditions is a real challenge. Tim initially gathered information and contacts of potential malleefowl monitoring sites. They then both travelled across the country gathering support from land managers.

One of the keys to a successful statistical experiment is “replication”: if the same reaction to baiting is observed again and again at many experimental sites, we can draw conclusions with more confidence. So, how many pairs of sites are needed to be able to make some useful inference about fox baiting? The adaptive management team conducted a “power analysis” to help answer that question. This statistical procedure requires lengthy computer simulations and analyses, but the basic principle behind it is quite simple. If the increase in malleefowl breeding activity due to fox baiting is very large, then observing it at a few sites would be enough. At the other end of the spectrum, if the effect of baiting is very small, it would take many sites to detect it in a statistical analysis, but then again it wouldn’t matter because a small effect would mean that fox baiting would not be an efficient way of increasing malleefowl breeding activity. Our power analysis indicated that about 20 pairs of sites monitored for at least 5 years could be needed to have a reasonable chance to detect an effect of fox baiting that we believe would have a real impact at a population level.

The devil is in the details though, and conducting experiments with natural systems is often trickier that in a laboratory. Ecosystems are always messier and more nuanced than the simplified model used in our power analysis. We also know that we will never get pairs of treatment and reference sites that experience exactly the same environmental conditions over time. Furthermore baiting can be conducted at different regimes of intensity and timing, and the experimental sites will be located within a broader geographic landscape in which other landowners may conduct baiting. As we begin incorporating real sites into this plan, we can collect this more detailed information and account for at least some of it in our analysis.

Despite these challenges, the experiment approach we propose is still our best shot to obtain a robust answer to the question of fox baiting as a management tool for malleefowl conservation. A growing network of experimental sites will establish a solid base to provide the learning we need to improve management practices. And the methodology can be used more broadly, as it will serve as a blueprint to tackle other management uncertainties in malleefowl conservation, such as the effect of fire regimes on malleefowl populations.

It will take several years to gather enough data, so the sooner we start the better. In the adaptive management team, we are really excited to see the progress so far, with several potential sites in WA and SA already under consideration. Stayed tuned!


Posted in Uncategorized | 2 Comments


Congratulations to Phil and his team! They got the second prize at the EDG video competition 2014 !!!

Phil and his team If you haven’t watched the video yet, you can find it in my previous blog or in Gurutzeta’s blog.

Some people have asked about how the video was created, so we list here the main technical details:

• The technique used is called ‘stop motion’, where a sequence of still pictures is put together in a video to create the impression of movement.

• High resolution pictures with a digital camera mounted on a tripod, using a remote infrared trigger. Tripod (or a stable surface) and trigger were essential to ensure stability of the frame throughout the process of taking hundreds of pictures.

• The pictures were combined into a video with program iMovie (the default video editing tool in Macs) with a frame rate of 10 frames per second.

• Floating text and figures were incorporated manually into each frame using Photoshop.

• The floating text itself was created by hand with one of the cheap animation tools (i.e. Loop) for iPad.

• Audio: captured with an external microphone and incorporated into the final video in the video editing tool.

• Sound effects and the music clip in the Bloopers video come from the free library within iMovie (and are royalty-free under license Apple inc. iLife software).

And, by the way, happy 40th anniversary to the Playmobil, who were created in 1974!


Posted in Uncategorized | 1 Comment

Measuring the impact of herbivory on regeneration of pine-buloke woodlands

Last week, Cindy and I travelled again to Wyperfeld, a national park in the semi-arid mallee of northern Victoria. The reason this time was a pilot study of some vegetation monitoring methods for our kangaroo adaptive management project. Without getting into details of the project itself (I leave that for another blog entry), the basic idea is that if you are trying to manage the impact of overabundant herbivores on vegetation regeneration, you’ll need to have some way of monitoring whether management actions taken are actually working or not! This is important in any kind of management, but necessary by definition for ‘adaptive management’ since otherwise we cannot learn and improve our management over time.

The focus of our project are mixed woodlands of slender cypress-pine Callitris gracilis and buloke Allocasuarina luehmannii, located in some areas of Wyperfeld. The seedlings of these species are being eaten (mostly) by Western grey kangaroos and rabbits when other food sources are scarce, before they are able to recruit as adult trees. These species are not the favourite food for kangaroos, but they’ll resort to them when other food sources have already been depleted.


Slender cypress-pine (left) and buloke (right)

With this trip, we wanted to get a better understanding of the distribution of pine-buloke regeneration within the park, measure seedling density around mature trees and trial a range of vegetation survey methods to assess their feasibility for ongoing monitoring.

We were joined by a bunch of researchers: plant qaecologists Chris Jones and Dave Duncan directed the surveying, while some volunteers (qaecologists Gurutzeta Guillera-Arroita and Kate Cranney, and Simon West who was visiting our lab) provided the much needed manual labour. Many thanks to all involved!!

Point surveys in a 20x20m plot around a mature pine

Counting seedlings in a 15m radius circle around a pine

Exploring an area looking for seedlings

After the two first days driving around the park, it was clear that pine seedlings were not abundant and many showed signs of herbivory. After summer, when herbivores are at their hungriest levels, the impact on vegetation is most visible.

A couple of sites stood out, with many small seedlings. But we had only found one solitary buloke seedling despite searching several woodland areas.


An area of nice pine regeneration (left), although showing some signs of herbovory (right)

Only on the third day, we stumbled upon a large area with many buloke saplings. Sadly, most of this area had been affected by the devastating fire that consumed parts of Wyperfeld this summer. Lightly burn and often chewed on, many of these saplings were still alive… there is hope for a new generation of bulokes in this area!

buloke seedling

Buloke seedling after a recent fire

The trip was very productive, and provided a lot of insight about the situation on the ground. This will help us produce recommendation regarding location and size of grazing exclosures, which will help track the progress towards the long-time objective of pine-buloke woodland regeneration, as well as providing the means of disentangling the effects of herbivory and environmental variation on an annual basis.


Any guess about what is inside and outside the kangaroo-proof fence?


PS: For the ornithologically-minded: it was a pleasure to get some nice views of Major Mitchell’s cockatoos and blue bonnets, two beautiful species of parrots from the semi-arid interior!

Posted in Uncategorized | 3 Comments

From paper to movie

A short while ago, I published a paper with Gurutzeta Guillera-Arroita and Brendan Wintle about the impact of disregarding imperfect detection when modelling the distribution of species:

Lahoz-Monfort, J.J., Guillera-Arroita, G., Wintle, B. A. (2014) Imperfect detection impacts the performance of species distribution models. Global Ecology and Biogeography, 23(4), 504-515.

I won’t get here into the details of our study (you can read a Decision Point article about it in issue #77 here or as a blog entry in the Qaeco website here). What matters is that when we heard about the ‘Great EDG Video Competition’, Guru and I thought this study provided a great excuse to have some fun while experimenting with new ways (for us!) of communicating science to the broader audience.

We set out to produce our video having ‘managers’ as a target audience, or in a broader sense, people with direct responsibility of making decisions about wildlife and habitat management that require some knowledge about where a species is or is not. When this knowledge comes in the form of a ‘species distribution model’ that is based on survey data, decisions may unknowingly be impacted by the effects that imperfect detection of the focal species can have on the distribution modelling. And that’s exactly what our paper is about!

Well, without further ado, here’s our entry for the EDG competition! (for better image quality, please change the resolution to 1080p HD using the settings button – the little cogwheel under the video)

And since things don’t always go right when you try new techniques, here are also some outtakes:

We chose to use the stop motion technique, terra incognita for us. And some of my old Playmobil figures got to see the light of the day again (including ‘first generation’ klickies without articulated wrists from the early-80s!).

outback at home

Kitchen-turned-into-studio: the outback at home!

Although we underestimated the amount of time it takes to get the studio setting right, shoot the thousands of pictures required to fill 3.5 minutes (the video was made with 10 frames per second), record and synchronise the sound (thanks David for the nice mic!) and produce the final file, we thoroughly enjoyed it!

…only 1082 shots left…

…only 1082 shots left…

Golf players can be an occupational hazard when filming at Royal Park!

Golf players can be an occupational hazard when filming at Royal Park!

It has been fun producing the video. I hope you’ll enjoy it!

Posted in Uncategorized | 8 Comments

Streamlining existing long-term monitoring programs

How much can we reduce survey effort in a long-term environmental monitoring program based on mark-recapture-recovery (MRR) data, and still get what we want out of it?

In my latest paper (Lahoz-Monfort et al. 2014), which just appeared in early view in Journal of Applied Ecology, we explore this question for the monitoring of juvenile survival of a northern hemisphere seabird, the common guillemot (Uria aalge), at the Isle of May in Scotland. This small island, not far from Edinburgh, is one of the four ‘Key Site’ seabird colonies in the UK’s Seabird Monitoring Programme, and the colonies that cover most of its cliffs and grassy slopes have been intensively monitored for several decades.

Some of the seabird colonies cover very large cliffs and are truly impressive! Most of the dots in the picture are guillemots or razorbills, another auk species.

We perform data thinning of an existing MRR data set to identify potential field effort reductions, both in terms of ringing and resighting efforts. This way we can explore what would be the effect of spending less time ringing guillemot chicks or looking for ringed adults sitting by the cliffs with a fieldscope, on our ability to estimate environmental influences on juvenile survival.

We also highlight how more aggressive alternative monitoring scenarios (i.e. not ringing chicks at all) can be evaluated with Integrated Population Models (Besbeas et al. 2002), a relatively new modelling approach that combines data collected on different aspects of demography and abundance.

Guillemots do not construct nests, but lay the eggs directly on the narrow cliff ledges. The eggs are slightly conical so they don’t roll out of the ledge. And when the chicks are just a few weeks old, they jump to the water many meters below them to follow their dads to the open sea. It’s an impressive show that I’ve been lucky enough to witness while at the Isle of May.

This type of post-study evaluation can help streamline existing long-term environmental monitoring programs, but we’re not advocating such streamlining ourselves: obviously, by reducing effort we always lose something, and some of the most aggressive strategies would deliver much poorer ecological data sets (e.g. we wouldn’t be able to investigate individual-level effects on juvenile survival if no chicks were ringed). But in the current public funding climate, many wildlife monitoring programs are coming under pressure to reduce costs; our approach explores how to adjust field protocols to collect demographic data when the effort reduction is a mandate, incurring the least impact on our ability to achieve monitoring and research objectives.

We used these tools in an exploratory way, but the next step would be to make an explicit link to a decision context (i.e. a fixed budget, explicit objectives for probabilities of detecting effects). Food for thought! (and future papers?)

Looking for ringed guillemots at one of the cliffs.


Lahoz-Monfort, J.J., Harris, M.P., Morgan, B.J.T., Freeman, S.N., Wanless, S. (2014) Exploring the consequences of reducing survey effort for detecting individual and temporal variability in survival. Journal of Applied Ecology. DOI: 10.1111/1365-2664.12214

Besbeas P., Freeman S.N., Morgan B.J.T. & Catchpole E.A. (2002). Integrating mark-recapture-recovery and census data to estimate animal abundance and demographic parameters. Biometrics, 58, 540-547.

Posted in Uncategorized | 1 Comment

Malleefowl and kangaroos: a weekend at Wyperfeld National Park

Although I’ve been trying to start blogging for a while now, the conference season and some busy weeks have kept me away from the growing backlog of ideas for posts. But the last weekend provides the perfect excuse for a first entry related to two of my projects, so here we go!

The Victorian Malleefowl Recovery Group (VMRG) is the organisation of volunteers that conducts the monitoring of malleefowl mound activity in the state of Victoria. Malleefowls (Leipoa ocellata) are really cool birds with a fascinating breeding ecology, but I’ll dedicate a future blog entry to introduce them properly. For the time being, I’ll just tell they construct very large mounds in the sand in arid and semi-arid regions of Australia, in which they incubate their eggs.

The VMRG is a key partner in one of our Adaptive Management projects, providing a truly impressive data set of mound activity observations at many survey sites scattered along the range of the species in Victoria. Every year, they gather at Wyperfeld National Park for a weekend in which new volunteers are trained in bush safety, use of equipment and learn the monitoring protocol, an essential step to ensure data consistency!

Learning how to measure mound height [left]. Next to an inactive (but not too old) mound [right]

Learning how to measure mound height [left]. Next to an inactive (but not too old) mound [right]

This weekend provides us (Cindy and myself) with an opportunity to catch up, learn from their experiences in the field and update them with what we’ve been up to with the Adaptive Management project (food for another blog entry, by the way). Oh well, and an excuse for us to get away from the computer monitor for a while!


Update on the malleefowl Adaptive Management project during the VMRG annual general meeting

The rest of the weekend was dedicated to our other Adaptive Management project, which is in Wyperfeld National Park, in collaboration with Parks Victoria (again, I’ll add more detail about this interesting project in a future post!). We visited different areas of the park that are of particular conservation concern due to overgrazing by native kangaroos and introduced herbivores (rabbits and goats). Acting ranger-in-charge David Christian gave us a comprehensive tour of parts of the park that include mature stands of buloke (Allocasuarina luehmannii) and slender cypress-pine (Callitris gracilis).

Young slender cypress-pine at Pine Plains

A growing slender cypress-pine at Pine Plains

Western grey kangaroos (Macropus fuliginosus) are abundant in parts of the park

Western grey kangaroos (Macropus fuliginosus) are abundant in parts of the park

As a side note, I had a first go at driving a 4WD on sand! (often remarkably similar feeling to driving on compacted snow, to which I’m used after living 8 years in Finland).

And finally, the 5 hours’ drive back to Melbourne… did I say already that Australia is indeed a BIG country?


PS: for the birders around, we had some exciting sightings that provided the icing on the cake: some of the parrot specialities of the mallee (Major Mitchell’s cockatoos, blue bonnets, regent and mulga parrots) and an inquisitive striped honeyeater! We also saw two malleefowls in what I’d call the “malleefowl highway”, close to Ouyen.

Posted in Uncategorized | 2 Comments