Predator Guards Carry Their Weight

Photo © Glenda Simmons

Humans have been providing nest boxes for cavity-nesting birds for hundreds of years. In the last 50 years, a lot of creative ingenuity has gone into trying to solve the problem of nest predation. Enter the predator guard, a device that is installed on or below the nest box to keep predators away from vulnerable eggs, nestlings, and even incubating females. However, almost no studies have tested their performance. In 1969, L. Kibler hypothesized that metal cone-shaped guards on nest-box mounting structures are ‘probably’ the most reliable device against ground predators, yet five decades later no conclusive research has been published.

With the rise of citizen scientist nest monitors, it’s now possible to test the effectiveness of predator guards in promoting the nesting success of cavity-nestings birds at a large spatial scale (United States and Canada). Using NestWatch data from 24,114 nest records submitted from 2014–2016 to NestWatch, the Cornell Lab of Ornithology tested whether installing predator guards on nest boxes is an effective management technique. We also tested how different guard types compared to each other, and whether or not multiple guards is any better than a single guard.

Types Of Guards
Types Of Guards

These four common types of predator guards were compared in our national study. Types A (cone baffle), B (stovepipe baffle), and C (entrance hole extender) were equally effective, while D (Noel guard) was somewhat less effective.

Guarding Your Interests

When we looked at all species combined, the nest survival data suggested a 6.7% increase in nest success for attempts in boxes with guards versus attempts in boxes without guards. That may not be the 100% protection that many people believe they’re providing, but 7% is actually a large increase at the national level. There are few other actions a homeowner can take that would improve nest success to that extent. However, all species did not benefit equally. Western Bluebird stood out as a species for which a predator guard did not seem to make a difference. We’re not sure why, but it could be that other factors (e.g., drought, insect supply) have more impact on their nesting success than do predators. Other species, such as the Carolina Wren, showed a 15.7% increase in nest survival when guarded!

Although all types of guards were correlated with improved nesting success, birds nesting in boxes with cone-type baffles, stovepipe baffles, or entrance hole extenders (also called “wooden block hole guards”) were most likely to result in successful nesting. The Noel guard did not rise to the top as a clear leader, although it is surely better than nothing (see guard types in photo above). Additionally, birds nesting in boxes with multiple predator guards (such as a cone baffle and a hole extender) were more successful, on average, than birds nesting in boxes with only a single guard.

Eastern Bluebird Box With Hole Guard
Eastern Bluebird Box With Hole Guard

A female Eastern Bluebird feeds her youngster through an extended nest hole entrance. Not all extenders are made of wood; some designs are made of metal or plastic. These keep predators from easily reaching into the box.

We Are A Force For Birds

At NestWatch, we know that caring for wild nesting birds is a top motivation for providing nest boxes, and most people want to maximize nesting success while minimizing human effort. The predator guard is therefore an inexpensive, passive, and effective way to increase the survival of nests, especially with other factors being less under our control (e.g., weather, food supply). Download predator guard plans here. However, there is still no such thing as a “predator-proof nest box” because it is hard to control for predators such as bears and House Sparrows, which are not so easily deterred.

We would like to thank the legions of NestWatchers who monitored 12,274 nest boxes (both guarded and unguarded), enabling this comparative study. Without you, large-scale studies like this would not be possible!


  • Bailey, R. L., and D. N. Bonter. 2017. Predator guards on nest boxes improve nesting success of birds. Wildlife Society Bulletin 41(3):434–441.
  • Kibler, L. F. 1969. The establishment and maintenance of a bluebird nest-box project: a review and commentary. Bird-Banding 40:114–129.

Filed under:


  1. SDBluebirds says:

    The Western Bluebird in California does well in hanging boxes as well as in pole-mounted boxes without predator guards. In San Diego County, we don’t have a lot of problems with climbing predators like raccoons or opossum. That’s because the riparian habitat that supports these animals is disappearing, and also because we still have large predators (mountain lions, coyotes, bobcats and foxes) which help keep their populations in check. Nor do we have climbing black rat snakes, which are so prevalent in the south-east. So hanging a nestbox in a tree here is not the “invitation to disaster” that it might be in other parts of the country. Nestboxes that have been hung in trees for years here might only face problems with ants or earwigs (which can be controlled with a Tanglefoot guard) or from strong Santa Ana winds. Trees also provide needed shade during our hot summer months. However, mounting the box on a pole under the shade of a tree at a height of 5 to 6 feet (with no predator guard) also works very well.

    • Robyn Bailey says:

      Hi, thanks for reading. Our study unfortunately did not differentiate between hanging boxes, and boxes mounted on poles. Therefore, the results you’re reading about only apply to the specific types of guards mounted (which in general tend to be used on poles).

  2. Robert Spahn says:

    The obvious problem in the analysis here is that there is no assessment of the percentage of nests lost to various causes. Without that the number resulting from adding nest guards is almost just a number hanging in space with interpretation very difficult.
    Going to your WEBL example, maybe the observation of little effect from adding guards just means there was very little predator impact to begin with.

    • Robyn Bailey says:

      Hi Robert,
      We removed other sources of loss, such as eggs not hatching, or all young found dead in the nest (starvation). Therefore, we believe we have isolated the effect of predation as much as possible. Let me know if you’d like a PDF of the original paper, and I’d be happy to send it to you. Thanks for reading!

    • SDBluebirds says:

      “Going to your WEBL example, maybe the observation of little effect from adding guards just means there was very little predator impact to begin with.”

      Yes, I think your statement is correct. Competition from other birds and insects are a much bigger threat to WEBLs than climbing predators.

  3. Mason Sexton says:

    I would think that a similar study on house sparrow deterrence is even more important given the recognition that wherever house sparrows are nesting eastern bluebird/tree swallow populations are statistically lower by as much as 50%. The truth is that because of local and state regulations in many parts of the country trapp and release is not a viable option.

    • Robyn Bailey says:

      Hi Mason, NestWatch is collaborating with some scientists at North Carolina Museum of Natural Sciences to work on House Sparrow deterrent methods. We have had some promising results, but this is a long-term study that will take a bit more time to report on. Thanks for reading!

    • I am uncertain that I understand what is meant by the statement, “that because of local and state regulations in many parts of the country trapp [sic] and release is not a viable option.” Perhaps I am simply missing the point, yet would you kindly clarify this for us?

      As a rule, we (MBS, as an organization) do not trap and release nor to we recommend the trapping and releasing of House Sparrows on our trails. We trap and/or euthanize HOSP outright. There are no legal restrictions or regulations for the trapping, releasing, euthanizing or even the transport of House Sparrows across county boundaries inasfar as I am aware (according to state, federal or MiDNR regulations), at least in Michigan (where we are located).

    • Paula says:

      Trap and release is no-win, anyway. Besides depositing the problem in someone else’s backyard, populations of HOSP allowed to reproduce unchecked will eventually fan out to re-colonize the areas from which they were “evacuated”.

  4. Rick Barber says:

    Conical baffles are not only useful for nesting boxes, but also for bird feeders on poles. For a while, my feeders were being regularly emptied overnight. I suspected racoons were to blame. Installed a baffle on the pole, end of problem.

  5. Marcia says:

    I monitor a few trails totaling over 200 boxes with a large number of them deliberately placed in high snake and raccoon locales. I do not report my nesting results, not because I don’t know whether I have lost nesting attempts to predators, but because, unfortunately, I do not believe all of the results reported by others. For example, snakes quite often will take birds even one day before fledging, giving the impression to the monitor that the birds fledged and reported as such, when the snake , in fact, consumed the young birds, with little evidence left of their work. I, personally, always use a method to determine predation whereas most others do not. The reason I place my boxes where I do is that I am studying a few low cost methods to hopefully end all predation by snakes (a major problem). I had tremendous success this year and will need one more year before reporting my results. It’s the most excited I’ve been in the 22 years I have been monitoring. (I will say, after many years of using the Kingston baffle, made from 8″ diameter stovepipe, I can most definitely say that, alone, they do not work against black or rat snakes. It’s not to say I don’t use them, but never, ever alone.)

  6. Dick Malnory says:

    There was no discussion concerning snakes. How effective are guards A and B at deterring snakes? I wouldn’t think type C and D would be of any deterrent.

    • Robyn Bailey says:

      Hi, thanks for reading. We agree that types A and B are probably most effective against snakes and other climbing predators. However, when we look at all types of predators, and all over the country, types A, B, and C were equally effective.

    • Marcia says:

      In my experience of hundreds of unsuccessful attempts to stop snakes, type B alone definitely does not work. One method of detection is to use vaseline with added food coloring placed with a syringe around the cylinder. A snake gives a telltale sign of predation. In the photo included,

      you can see how the snake dragged the vaseline up the guard to the box. The snake holds the pipe with the bottom of its body then moves upward with the top part to move over the guard to the box. Therefore, we must stop them either below the cylinder and/or above it. I am hard at work on that and will report after another year of success.

  7. Julia G. says:

    Can the cone-type guard be placed directly on a tree trunk? Squirrels and roof rats climb the trees, then go from branch to branch looking for bird food, nests, eggs, and babies.

    • Robyn Bailey says:

      Julia, I have seen some folks wrap sheet metal directly around a tree trunk to protect nest boxes mounted on trees. You might also see something like that mounted on fruit/nut orchard trees to keep pests out. We did not look at the effectiveness of those because sample sizes were too low, but I have seen at least one study on stork nest trees being wrapped in plastic, and it was effective.

  8. John Finnell says:

    I was experiencing predation on 2 or more attempts per season which I suspected were due to snakes since I saw king snakes and racers occasionally. After I installed a stovepipe baffle I have had no predation problems at all. I am only surprised that the increase is only 7%. My experience would be more like 20%.

    • Robyn Bailey says:

      Hi John, thanks for commenting. The 7% increase is all species pooled, across the country. When you break it down by individual species, the results are varied, with some species seeming to benefit more than others. However, I do think that people tend to overestimate how much of an improvement the guards make, and this is likely because they are confounding other variables, such as natural food supply, weather, and nonviable eggs. In this study, we removed those nests that failed due to nonviable eggs, or where all young were found dead in the nest (not predation-related). Therefore, the effect of excluding some predators is not as high as you might think. I have never seen a study that looks at what proportion of attempted predation events are thwarted by these devices (if you know of one, please share it with us!), so it would be great to know how effective these guards are…20%? 30%? 50%? I don’t think we know.

  9. Leda Beth Gray says:

    What are the specs for a cone baffle that is effective against raccoons? We had something get past our baffles in Sedgwick, Maine and I am guessing it was raccoons and maybe the baffles weren’t large enough.

  10. Linda Orkin says:

    Great study with such a large data set. Is there any data to compare to natural nest sites, which are, of course, in short supply these days

    • Robyn Bailey says:

      Linda, there are lots of studies on natural cavities versus nest boxes. This paper mentions a few in the introduction (although we did not study natural cavities). I can send you the PDF if you like.

  11. As both a long-time nestbox trail steward and the Kalamazoo & Barry County Coordinator of the Michigan Bluebird Society, and also the owner of Friend of Birds Wildlife Products, I have always stressed the importance and value to trail monitors and to clients regarding the attachment of multiple nestbox guards to all nestboxes if at all possible. In my experienced opinion, it is better to not install a box than to erect an un-protected box. The risk is simply too great and the odds are too stacked for the potential of unwanted incursion and/or predation.

    In my 20+years of anecdotal evidence, the use of multiple guards increases the nesting and fledging success rate consistently at or above the levels indicated in the above article. We attach deep (up to 3″ long) wooden entrance hole extenders to all of our production Bluebird nestboxes, and to most of the boxes on the trails that I manage. Nesthole extenders are very simple and inexpensive to create and quite easy to attach.

    An extenuating and highly changeable variable is the siting/location of boxes, which by nature may be highly subjective in actual practice- while also being an inexact science subject to the seasonal and daily iterations and fluctuations of any intrusive species, any alterations in the habitat and to the breeding intensity and/or density of any target species.

    Many nestbox monitors are hesitant to attach entrance hole extenders during the breeding season when boxes may be occupied. Another issue is that boxes with upward-front-opening doors do not accommodate an extender since it makes the door incapable of being opened when an extender is attached. The consumer-ready cylindrical plastic or ceramic extenders have never worked well on my trails, with typically only House Wrens appropriating boxes with that type of guard attached. The hesitancy or fear of monitors in attaching an extender block (or other guard) even after a box is claimed, a nest constructed, eggs have been laid, eggs have hatched, and even into the juvenile stages can be assuaged by my personal experience in successfully attaching guards (of all types) with the guards being accepted (in all cases, almost immediately, and at an almost 100% acceptance rate) during all such crucial phases or junctures. One simply has to attempt to initially ascertain the overall “temperament” of the individual adult (parental) birds or the strength of their pair bond/site fidelity (this may take years of working around wild birds), and then most importantly also to observe the nestbox until and after the new object (a guard) is accepted, optimally by both adult (parental) birds, and the nest, eggs or young are subsequently observed being tended. Conversely, if the guard is not accepted within an appropriate period of time, it must be removed immediately to prevent potential nesting failure.

    Of course, attaching guards before the breeding period begins is optimal, yet this doesn’t always occur, especially with large trails consisting of many multiple boxes.

    I hope this information might assist all nestbox trail stewards in their service and tasks.

  12. Vaseline on a birdbox pole is a horrible idea and people should stop using it to try to prevent predation of their birds. Vaseline is difficult for animals to remove from their bodies and this could have detrimental effects upon their lives. Additionally, this sticky substance could trap insects and spiders, the numbers of which are declining. They should not be needlessly killed.

    Also, netting that entraps and kills snakes, is equally horrendous and wrong-headed.

    Remember, predation is SUPPOSED to help limit bird numbers, so losing SOME nestlings is natural and should not be seen as something to be totally prevented.

    • That is an interesting comment regarding “natural” predation and the idea of acceptability regarding the allowance of “natural” predation on nestbox trails.

      My opinion differs for a couple of reasons. I disagree that all “natural” predation be allowed or viewed as acceptable.

      One reason is based upon the fact that as nestbox trail stewards who erect nestboxes and subsequently establish and monitor trails, we are creating artificial cavities for the target or focal species. Within this construct, we are thus also tasked by common rule of our being trail managers to do all within our power to secure those boxes for the overall safety and success of those individual birds. Thus, I disagree that in these scenarios of focused, targeted hosting that any predation is “acceptable”.

      There is a clear differentiation or disparity of thinking, terms or definition between “natural” predation, “unnatural” predation and “acceptable” predation, in my mind. I would add the terms “unnatural” and/or “preventable” predation to the scope of predation scenarios that monitors should combat as a regular part of their duties as nestbox stewards.

      For example, although it is nigh-impossible to predict House Sparrow incursion, HOSP must be rebuffed at all costs and/or eradicated by monitors as they are neither a native, “natural” predator nor are they afforded protection, yet it is difficult to prevent them from accessing nestboxes. On the other hand, a Cooper’s Hawk predating the hen while she is on the nestbox is “natural” predation and should objectively be viewed as “acceptable”, while it also in all likelihood cannot reasonably be curtailed nor can the raptor be lawfully harrassed or taken.

      Conversely, a roaming domestic cat or non-native snake species climbing up a post to a box and predating the hen is unnatural, unacceptable and preventable and should by any reasonable standards be prevented by any means available, such as guards, a live trap, or outright euthanization of the predator.

      A raccoon is a native predator, yet it should also be prevented by all means available from accessing nestboxes. The same is true for House Wrens, whose nature and definition is on par with the Cooper’s Hawk in the acceptability of “natural” predation and loss and the resources lawfully allowed to prevent predation. The White-footed Deermouse, on the other hand, is a native, “natural” incursor yet whose presence should be rebuffed, and whom like most small game or non-game furred mammals either have seasonal takes or hold no legal protections. They should be prevented from entering and predating the nestbox (I only remove them if extant, and do not harm them).

      Snakes (and also other species on the decline) typically are afforded special protections and should only be prevented from accessing nestboxes. I agree that using Vaseline and netting fro prevention is unnecessary and harmful.

      I understand the underlying caveat or fact that native, “natural” predators also need to eat, feed their offspring and procreate. When an artificial cavity is erected, an invitation is likewise extended to a target species to risk nesting in it, exposing it the threat of predation. We accept this when we establish a nestbox trail. We likewise rarely ever know what occurs with natural cavities, although research shows that most are predated in much higher numbers than artificial cavities (and there lies an important piece of the puzzle, too). In my opinion, native predators will locate prey items elsewhere and anywhere that they are able to and they will thrive; simply giving them a trail to buffet from is not going to tip the scales in favor of their overall populations- yet will undoubtedly decrease the populations of the hosted species on those same trails.

      We as nestbox stewards must prevent any unacceptable and preventable incursion, predation and loss from occurring. We are trying to produce breeding birds from our trails, either to keep population numbers static/stable, to reduce decreasing population numbers in some species, or to increase population numbers overall. Allowing “preventable” predation is not a logical methodology to use within the artifices we have established. I simply wish to highlight the difference, and underscore the responsibility of the steward toward this goal. Anything one can lawfully do within these parameters to prevent predation is a positive thing.

      I’m certain that the moderators/authors may have some valid input on this very important topic.

      • Paula Ziebarth says:

        All great insights, John. Simply put, IMHO when we install an artificial nestbox, we invite native cavity nesters to nest there. It makes NO sense to invite them to be eaten. There is personal responsibility associated with installing nesting cavities for native birds. I would agree with earlier comments that other native wildlife should not be killed if they attempt to enter that nest cavity. They must, however, be stopped if possible.

  13. Lacy Thomas says:

    I make ‘healthy’ bird houses (ventilation, drainage, easy access, etc.) for cavity nesters; predator guards are VERY important to me. I have found that the ‘standard’ hole extension of 3/4″ is not very effective, so I have started adding a 1 1/2″ guard. Also, I have found that the extra thickness PLUS a 2-3″ extension using PVC conduit is VERY effective against HOSPs – the HOSPs hate the “tunnel” effect and are reluctant to even try to enter. This is 1.5″ inside diameter conduit and I use it for Bluebird homes.
    Also, the Woolwine Bluebird Society has conducted a 5-year study for Bluebird houses using TWO entrance holes and have found that this style is very effective a.gainst HOSP. You can easily review their study by googling. I do not have a website but I do have a facebook page (Coops D’ Villas) where I have photos of my PVC guard boxes – they are SW Kachina style; if anyone needs more info I will be glad to share.

  14. Hi Jonathan and Paula,

    Please note that I said SOME predation, not “that all ‘natural’ predation be allowed or viewed as acceptable.”

    Also, I agree with Paula that “It makes NO sense to invite [birds] to be eaten.”

    My point is that you do your best to minimize predation, but people need to understand that no predation whatsoever leads to OVERpopulations of particular species of birds, and that is not the result anyone should want. Please keep in mind that a properly functioning environment is a continuous balancing act in which no species is supposed to take too much from the environment.


Post A Comment

Your email address will not be published. Required fields are marked *