Interpreting Evidence of Depredation of Duck Nests in the Prairie Pothole Region

1. What they did

Ducks nesting in the Prairie Pothole Region lose many nests to predators, and managers need to know which predators are responsible. Yet nearly all past identifications rested on subjective reading of the signs left at destroyed nests, and the published descriptions were thin, vague and sometimes contradictory. This report sets out to help investigators collect and interpret that evidence more accurately and objectively.

The authors drew on the literature and on their own work from 1972 to 1992 with nine predator species and two species groups. They gathered “verification data” (depredations where the predator was known) in eight ways:

• Offering duck eggs to captive predators in cages and pens.
• Offering artificial nests (six eggs, usually unincubated mallard eggs, in nest material salvaged from real duck nests) to predators in a compound and a large enclosure, watched from an observation booth. Some coyote and fox nests held a restrained live mallard to imitate an incubating hen.
• Photographing free-ranging predators at artificial nests with time-lapse cameras (one frame every 15 to 30 seconds).
• Watching free-ranging predators eat loose eggs.
• Monitoring natural nests raided by radio-collared predators.
• Examining their file of nest depredation records for natural nests.
• Examining natural nests destroyed inside a fenced exclosure where mink were the only predator able to get in.
• Examining eggshells from natural nests where unique evidence (such as eggs buried by badgers) pointed to one species.

The report has three parts: (I) what affects the evidence left at a nest, (II) the pattern each predator leaves, and (III) how to interpret the evidence, with worked examples from 389 destroyed nests on waterfowl production areas in Minnesota, North Dakota and South Dakota. It ends with a glossary, photographs of eggshells opened by each predator, and a standard form, the nest depredation record.

2. What they found

What shapes the evidence (Part I)

• Eggs late in incubation have brittle shells and break more; bigger, thicker eggs are harder for small predators to carry or open.
• Hens may carry away eggshells, return eggs or nest material, or keep sitting after part of a clutch is taken.
• Individuals of one species can behave differently, and young predators handle eggs differently from adults.
• Body and jaws matter: large carnivores can carry a whole egg in the mouth; small carnivores rely on repeated biting and leave paired canine puncture marks; Franklin’s ground squirrels gnaw, leaving finely serrated edges; birds peck. The raccoon, with handlike front feet, is the only predator in the region adept at holding an egg upright to eat. Badgers, and to a lesser degree skunks, have long claws and are the most likely to dig.
• The distance between paired canine punctures helps separate predators: male upper canine tips averaged about 31 mm apart in coyotes, 23 mm in red foxes and raccoons, 17 mm in striped skunks, 30 mm in badgers, 12 mm in minks, 8 mm in long-tailed weasels and 5.5 mm in ermines.
• Later visitors (other predators, birds, mice, beetles) can change the evidence.

Each predator’s pattern (Part II)

• Coyote: the most variable pattern and easily confused with others. They took eggs one at a time; of 141 eggs removed, 64% were eaten, 24% cached singly 6 to 80 m from the nest (average about 27 m) and 12% dropped. Most eggshells (93%) had small, usually round holes, almost all in the side. A whole hen carcass without bite marks in the skull is a strong but rare sign.
• Red fox: the most distinctive pattern. Foxes carried off and cached nearly every egg (99.8% of 454), one at a time, 4 to 163 m from the nest (average about 44 m), barely covered with soil. A nest destroyed by a fox typically has all eggs missing, no disturbance of the ground or vegetation, and under 10% of the nest material moved. They often kill the hen but leave little sign of it at the nest.
• Raccoon: eats eggs at or near the nest, leaving shells of nearly all eggs (91% accounted for). Most openings (59%) were in an end, presumably because raccoons hold eggs upright, with broken pieces caved inward. They moved little nest material and often trampled shells.
• Striped skunk: also leaves shells of nearly all eggs close to the nest, but with large, often elliptical holes mainly in the side or side-end. Skunks often pulled nest material out to one side and left a short trail of trampled vegetation. One skunk with a den about 10 m away carried many eggs along the trail to it.
• American badger: causes the most disturbance. The authors treated eggs buried under 2 cm or more of soil or debris at the nest as unique to badgers; in 61 such nests, eggs were usually cached singly in freshly dug holes, up to 14 cm deep. Shells were often crushed or severely damaged.
• Mink: little is known. A hen carcass with bites to the skull or neck suggests mink; other signs are eggs missing without trace, shells with many puncture marks and yolk residue, no displaced nest material, and well-defined narrow trails leading away from the nest.
• Weasels: remove eggs gradually over several days, so hens often keep sitting on partly raided nests. Openings are narrow slots with finely chipped edges, sometimes with paired canine punctures.
• Franklin’s ground squirrel: took on average 2.2 days to raid a six-egg nest. Openings are broad and irregular with fine serrations; shells are scattered; sometimes the squirrels covered remaining eggs with nest material, making the nest look like one the hen had covered.
• Black-billed magpie: seemed only moderately interested; often carried eggs away, scattered tufts of nest material that caught in standing vegetation, and left yolk residue in shells.
• American crow: usually flew off with eggs; at 59% of nests raided on the day of discovery, all eggs vanished without a trace of shell. Crows tossed nest material, and shells sometimes had two openings on opposite sides.
• Gulls: responded weakly; only 27% of 63 artificial nests were visited. Openings were typically narrow, elongated slits with pointed ends on the side, and no shell was opened on an end.

Interpreting evidence (Part III)

• The most useful clues for telling predators apart were how eggs were treated, the type of damage to the shell, and where the opening was. Other clues (cached egg, dug area, dead hen, yolk residue) rule species in or out, and others (where the shells lie, moved nest material) back up the rest.
• Shell evidence held up fairly well over time: at 81 destroyed nests revisited after an average of 8.9 days, 89% of the eggshells remained. But at 15 of 18 nests that still held whole eggs, more eggs were taken before the revisit. The authors recommend using only nests found destroyed within two weeks of the last visit.
• In the worked examples (389 nests), they assigned about 40% of destroyed nests: red foxes about 19 to 21%, striped skunks a conservative 18%, raccoons a conservative 4%. Eleven nests fitted both skunk and raccoon. About 60% could not be assigned to these three species.

3. What they concluded

• Identifying nest predators always means interpreting circumstantial evidence, so results are subjective, not definitive. The authors question the accuracy of most earlier identifications in the literature.
• Some predators are easy to recognise (red fox, badger), others vary greatly (coyote), and some overlap a great deal (Franklin’s ground squirrel and weasels; certain skunks and raccoons).
• The report can be used as a field guide, but for research or management, investigators should not rely on identifications made at the nest. They should record the evidence on a standard nest depredation record and assign predators later.
• Estimating the share of nests destroyed by one species takes three steps: define a precise question, delimit the data (for example, nests with six or more eggs, found destroyed within two weeks), and apply a hierarchy of criteria from most to least definitive.
• Only predators actually present should be considered, so investigators are urged to collect predator population data in nest success studies.
• Evidence from artificial nests probably exaggerates how much nest material predators move, because real nests are packed more tightly.

4. How tracking was or was not used

The whole report is about reading feeding signs at destroyed nests: eggshells and how they were opened, tooth puncture marks, cached and buried eggs, dug areas, displaced nest material, trampled and matted vegetation, trails and runways leading away from nests, hen carcasses and bite marks, and urine or scat left by predators (red foxes often mark eggshells this way, which shows a visit but does not prove the fox took the eggs). Track plots were used to confirm that mink were the only predator inside the exclosure. No trailing was reported.

The recommended nest depredation record deliberately leaves out tracks, hair, odours and tooth marks, and the authors gave up trying to record trails and matted areas, because investigators could not reliably tell predator disturbance from that of people and hens. They encourage experienced investigators to record such extra evidence anyway.

5. Tracker or observer certification

Not reported. The verification data were gathered by the authors and many named assistants, and the nest records came from investigators “with varied backgrounds”. The recording form was designed so that no special expertise would be needed. No observer skill was tested.

6. Other technologies, skills or innovations

• Time-lapse cameras photographing artificial nests every 15 to 30 seconds.
• Radio-collared free-ranging striped skunks, found while raiding natural nests.
• A large enclosure with an elevated observation booth; behaviour described onto cassette tapes and cache sites plotted on maps.
• A standard nest depredation record (Appendix D) with instructions, covering eggs and eggshells, ground surface and nest material within 3 m of the nest, and guidance on telling hatched eggs (end cap, detached membrane) from depredated ones.
• Photographs of eggshells opened by each predator (Appendix C) and a table rating each type of evidence as positive, negative or inconclusive for each predator (Appendix B).

7. Implications for conservation (from the paper)

• Waterfowl managers need to know which predators destroy duck nests to design ways of reducing predation.
• More careful, objective recording of evidence should give results that can be compared between studies, and the ability to assign nests to predators should improve as more is learned.
• An earlier study using this approach (Sovada et al. 1995) linked higher duck nest success in coyote areas (32%) than in red fox areas (17%) to the difference in the canid community.