Verifying bilby presence and the systematic sampling of wild populations using sign-based protocols – with notes on aerial and ground survey techniques and asserting absence

1. What they did

The greater bilby is a burrowing marsupial of arid and semiarid Australia, listed as vulnerable and still declining. Bilbies are sparse, mostly active at night and hard to see or catch, so surveys rely largely on their tracks and signs. But tracks and signs can be misidentified or wrongly aged, which can make a site look occupied when it is not (false presence), and sign can be missed where bilbies do live (false absence).

The authors, who have long worked on bilby surveys, reviewed bilby tracks and signs and set out:

• Which sign confirms a bilby and which sign should only flag possible bilby activity.
• A verification protocol (Table 2) saying what to record and what to search for next when each kind of sign is found.
• A list of plants (Table 1) whose roots hold insect larvae that bilbies dig out, with where each was recorded and by whom.
• Advice on systematic sampling with the 2-hectare (2-ha) plot method to estimate how much of a region bilbies occupy.
• How much effort is needed to say bilbies are absent from a site, for example before land is cleared for development, and how aerial surveys can be used.

2. What they found

Tracks

• Fresh bilby tracks are distinctive. The front foot leaves three distinct, parallel toe and claw marks of similar length. The hind foot print is narrow and longer than the front, made mostly by the fourth toe, with little sign of the short fifth toe.
• Rabbit tracks are rounded, with front and hind feet of similar shape and size. Bandicoots leave clear prints of both the fourth and fifth toes on the hind foot, and front toe marks of uneven length. Dasyurids and rodents show more toes.
• Bilbies move in a bounding gait, with front prints staggered and hind prints mostly side by side. Quolls, mulgara, bandicoots, rabbits and rats make the same pattern, and brushtail possums sometimes do, so gait pattern alone is not enough.

Diggings

• Bilbies eat insects (beetles, termites, root-dwelling larvae) and plant material (seeds, bulbs), and dig a lot. But goannas, echidnas, rabbits, wallabies, bandicoots, mulgara and mice also make bilby-like diggings.
• The only diggings that can be put down to a bilby alone are at the base of shrubs or forbs where the roots have been ripped open for larvae. No other living species in arid and semiarid Australia is known to do this. Most of these plants are Acacia species. These diggings are usually obvious and numerous and can last months or even years, especially on stony ground.

Scats

• Bilby scats are usually found on top of or inside the sand spoil of diggings, rarely away from digging activity. They come in groups of 2 to 5 pellets: oblong, almost round in cross-section, smooth-coated with rounded ends, and usually a mix of sand, plant and insect material that can be seen under 10 times magnification. Rabbit scats are more spherical.
• No other living species in arid and semiarid Australia makes scats like these. Echidna scats have jagged ends and are mostly ant and termite fragments. Some lizards (including the thorny devil) and frogs make similar-looking pellets, but these are rarely at diggings, hold little sand, are light, crumble easily and sometimes show white uric acid.
• Scats can last several months, especially when buried in spoil with little rain. Pellet size can show juveniles.

Burrows

• Bilby burrows can be up to 4.5 m long and 2 m deep, with round entrances and usually an apron of dug sand. Goanna and other reptile burrows are often crescent-shaped, wider than high.
• Burrows can be hidden under logs or in termite mounds, and old burrows with flattened aprons may be reused. Cats, foxes, goannas, echidnas, hopping mice and mulgara can use or rework old bilby burrows and make them look active. One bilby can use up to 18 burrows over several months, and burrow density ranges from 22 per square kilometre to hundreds.
• So burrows alone should not confirm a bilby: they can be missed, and other species’ activity can be mistaken for bilby activity.

The verification protocol (Table 2)

• Confirms presence: clear tracks (three parallel front toe marks, narrow hind print with indistinct side toes); scats; multiple diggings into the roots of larva-bearing plants.
• Potential activity only: burrows, diggings in the open, unclear tracks (gait pattern only). For each, keep searching nearby for confirming sign.
• Recommended records include measuring and photographing tracks with a scale, estimating the age of sign, checking for juveniles, collecting scats dry in paper bags or vials with silica gel, and collecting a plant specimen if the host plant is uncertain.

Sampling and absence

• Past plot surveys sampled each plot only once and ignored imperfect detection, which measures how good surveyors are at finding bilbies rather than where bilbies are. The authors recommend resampling plots within 1 to 3 months, because bilby numbers can grow quickly through breeding.
• Accurate ageing of sign is needed, for bilbies and also for predators and competitors. The order in which small animals’ tracks overlie the focal animal’s tracks can help judge the age of sign.
• Sampling should be stratified across habitats: residual landforms and places with larva-bearing shrubs, drainage lines, and sand plains and dune fields, plus fire age (1 to 2, 3 to 10 and over 10 years) where fire is common. Rainfall for up to 12 months before, and occupancy by rabbits, stock and predators (dingo or dog, red fox, feral cat), would ideally also be recorded.
• The more likely a habitat is to hold bilbies, the more non-detections are needed before calling it empty. Favoured habitats, areas with larva-bearing plants and areas burnt in the last 1 to 2 years must be sampled.
• Small development sites should be surveyed with a buffer of 8 to 10 km (twice the foraging range) around them, because bilbies emigrate and colonise new habitat.

Aerial surveys

• Bilby digging and burrow activity can be seen from a helicopter flying about 20 m high at 40 knots. Drones may make this cheaper as they improve.
• Ground checks are essential. Southgate et al. (2005) found 42% false-presence error and 3.2% false-absence error when they checked 55 and 92 plots under a 1084 km flight path. On a 490 km pipeline transect (EcOz 2016), potential bilby sign at 83 sites turned out not to be bilby sign in 18 ground and 65 hover checks, mostly goanna burrowing, and 44 2-ha plots in favoured habitat found no bilby sign: compelling evidence that bilbies were absent.

3. What they concluded

• Clear tracks, scats and diggings into roots for larvae can each confirm a bilby, alone or together. Gait pattern, diggings in the open and burrows should only flag possible bilby activity.
• The protocol is a small but vital part of monitoring bilbies across whole landscapes and of deciding how much survey effort is needed to say they are absent. Repeat sampling of plots and accurate ageing of sign are needed to estimate detection.
• The other pieces are now largely in place: a model of key bilby habitats, a standardised, widely used method for presence and absence data, and occupancy analysis methods.
• The federal survey guidelines (2 hours of searching per 1 ha site, then spotlighting at burrows) put more search effort per hectare than is warranted, and their cues for verifying presence from sign are not explicit enough. This leads to over-reliance on seeing the animals and to oversampling small areas instead of understanding bilbies across the region.
• Accurate bilby sign detection is challenging. Well-trained, experienced observers with good knowledge of the tracks and signs of different species are needed. The authors hope the protocol will reduce guesswork and false presence and absence errors, and make decisions more defensible.

4. How tracking was or was not used

The whole paper is about tracks and signs identification: telling bilby tracks, gait, diggings, scats and burrows from those of similar species, and judging the age of sign (including from small animals’ tracks overlying the bilby’s). The protocol asks observers to measure and photograph tracks and to keep searching around uncertain sign for confirming sign. No trailing is described. The 2-ha plot method records species mainly from tracks, and has been used by many Indigenous Ranger Groups. The authors thank Traditional Owners who shared their knowledge of interpreting animal sign, bilby populations and the plants that host root-dwelling larvae.

5. Tracker or observer certification

Not reported. The paper does not say who identified the sign shown in its photographs or recorded in Table 1 (mostly the authors’ personal observations), and no tracker certification or skill test is described. The authors stress that well-trained, experienced observers are needed, and that observer skill affects whether a species is detected on a plot.

6. Other technologies, skills or innovations

• Helicopter aerial surveys for digging and burrow activity, with ground visits or “hover checks” to verify flagged sign; drones suggested for the future.
• Low-power (10 times) magnification to see the contents of scats.
• Scat collection dry in paper bags or vials with silica gel beads and cotton wool.

7. Implications for conservation (from the paper)

• Unbiased data are needed to set the bilby’s threatened status correctly and to judge how populations respond to management.
• Misidentified or wrongly aged sign inflates estimates of where bilbies live; missed sign hides them.
• Assessments for development sites need enough well-designed effort, including buffer areas and favoured habitats, to support a claim that bilbies are absent. Ineffective sampling has been used to inflate the survey effort claimed to assert absence.
• The need for robust, defensible data will grow if the bilby stays vulnerable and development pressure increases.