Click here for a quick summary of what this document is about.

Solicited comments on
Goats in Olympic National Park:
Draft Environmental Impact Statement for Mountain Goat Management
(Published February 1995)

David Lee Winston Miller

This document is dedicated to someone who taught me
the most important things I know about animals:

my father,
Lieut. Col. John Josiah Miller (Ret.)

Table of contents:


The National Park Service (NPS) is considering several actions with respect to the Olympic mountain goats. Two of the alternatives include a population "elimination program" in which goats would be killed. As an alternative, I propose that the National Park Service employ the scientifically-accepted practice of risk analysis in determining proper action concerning the presence of mountain goats within Olympic National Park (ONP), Washington. Few scientists are well-trained in risk analysis, and I do not necessarily expect National Park Service (NPS) biologists to be exceptions. I do not even claim to be an expert in risk analysis myself. However, the issues raised beg to be analyzed in the context of risk--even if informally.

By using risk analysis, I believe that risk to plant taxa, risk to personnel, and risk to goats can all be minimized. Formal risk analysis would require that DEIS documented research contain fact-based probability figures for the researched assumptions (past, present and future). (For instance, the hypothesis that goat populations will increase in the future would be expressed, preferably, in formal terms of probability or, at least, in subjective terms of confidence such as "unlikely", "likely", "very likely", etc. . . .) To the extent that subjective terms are given in the DEIS, I find them in frequent disagreement with the facts presented.

It is important to note that I do not attempt to directly evaluate the reported research. Instead, my comments mainly concern the research as reported and summarized in the Draft Impact Environmental Statement. From an informal analysis, I have concluded that all the reported risks can be minimized: risk to rare/endemic plant taxa (and to archeological resources), risk to personnel, and risk to the Olympic mountain goats.

Of particular interest is my conclusion that no goats need be killed (and few, if any, should be removed from sensitive areas at this time for any reason). Another important conclusion is that although the government claims that the "No Action" program (Alternative 2) is the most costly of the three presented choices, it appears to be the least expensive.

If you hate mathematics, click here to skip the following:

Risks to Rare/Endemic Plant Taxa   (SEE UPDATE!!!)

There exists some risk that mountain goats may increase in numbers and harm or destroy large enough numbers of rare/endemic plants such that certain populations would no longer be viable. This risk has not been quantified by the NPS but appears to be extremely low since it is dependent on a chain of "ifs": If mountain goats are a significant threat to these plants, if the goat population increases naturally, if more effective means of contraception are not developed and employed (if necessary) in future years, if the increase exceeds a point in which the rate of plant damage could exceed plant recovery rates, if barrier means such as fencing are not effective in sensitive areas, . . . Note that the Draft Environmental Impact Statement (DEIS) tends to view the "ifs" in discrete ("black and white") terms. In real life (and risk analysis), things are more complex. In risk analysis, the probability of an event happening decreases (usually, dramatically) when a chain of possible events must happen to produce the event. As shown earlier, the probability of a particular outcome is the product of the probabilities of the independent events (scenarios) leading up to the end result. (Note: Not every event in life is entirely independent. Dependent events should be appropriately combined in a rigorous, mathematical analysis. However, I will not attempt an exhaustive formal analysis. I present above, the chain of events, in a less mathematical way to introduce the stated concept.)

Many, if not most, of the detrimental events that are expressed in terms of confidence (in the DEIS) appear to be rated as likely by the NPS. However, there are factors that appear to reduce the likelihood of these harmful events. Obvious factors are possible human error in analysis and oversimplification in the DEIS (a tendency toward presenting risk in discrete terms). This does not necessarily mean that shoddy science was employed in the document--it simply means that the DEIS document took a "black and white" approach, as opposed to a risk analysis approach, to the perceived problem. For example, statements like "If the National Park Service takes no action, the mountain goat population size will increase," are "black and white" statements. There appears to be no acknowledgment to the possibility that the population has already peaked. Of course, attaching probabilities to predictions is difficult. However, some estimate of confidence is better than none--even if it is simply an opinion of the author. (It may be that such estimates were employed in referenced documents. If they were employed, the DEIS did not usually present calculations--formal or otherwise--of the perceived risks.)

I would like to mention some other factors that reduce (or, in some cases, appear to have potential to reduce) the risk to plant taxa. These factors are organized below by scenarios envisaged in the DEIS:

DEIS Scenario:
Peninsula Goat Population Increases from an Estimated 288 in 1994 to a Maximum Sustainable Population of More Than 1600 Animals
(Note: The "more than 1600" figure is obtained by DEIS projections of 1402 in the ONP and 200 to 250 goats "outside the park [but on the peninsula]".)

According to the DEIS (page 23):

According to DEIS reported research (page 180):

The NPS contention, that the goat population will increase, conflicts with the census data suggesting a possible ongoing population crash. (The falling numbers cannot be explained simply by limited goat removals.) In fact, the evidence suggests that the peninsula-wide figure of 1175 goats is not sustainable. (Evidenced by the simple fact that the population was not sustained at this level.) It has been asserted by the Fund for Animals that the peak seen in 1983 may have been a product of ONP salt-lick baiting programs. Such baiting may have produced a manmade congregation of goats, increasing mating opportunities while the area and its surroundings still offered habitat from which goats could temporarily feed. An unnatural population expansion may have resulted. Thus, it may be that only man can produce such numbers. Regardless of the cause, nature apparently did not (eventually) tolerate numbers that high. It seems unlikely that it will in the future either.

In any event, the DEIS contains research summaries in Appendix B indicating that, normally, goat populations increase slowly in the ONP. Moorehead (1969) estimated the 1969 population within the park to be "at 250 to 300 and increasing slowly." Moorehead (1977) also later documented a long-term historical increase of only 2% per year. Since sustained population increases are necessarily asymptotic (limited, in this case, by natural factors), any sustained increases should be smaller in years to come. In fact, the 1994 estimate of 288 goats in the entire peninsula compares well with a 1954 estimate of 200 goats within the ONP alone. (See page 196 of the DEIS.) Therefore, in terms of risk analysis, the possibility of any sustained substantial increase in the near future cannot be rated high. The probability of a population explosion must be rated extremely low.

DEIS Scenario:
More Effective Means of Contraception are Not Found and Employed in Future Years

Research, as summarized in the DEIS, presents mostly skeptical views of contraception. However, Plotka (1994), in Appendix B of the DEIS, does admit that "under certain conditions, one or more of the techniques described in the 1991 report [which included contraception and other control techniques] could help control the size of the populations." McCullough (1994) also admits that "It is virtually certain that the technology of contraception will advance rapidly in the next decade." "Advances in contraception technology will surely make population control by contraception more efficient in the future." However, McCullough feels that field conditions present a "bottleneck" on efficiency. (Most of the skepticism presented in the DEIS hinges on the difficult field conditions. Problems--and possible solutions--related to field conditions are addressed later in this document.) But, McCullough does admit agreement with a 1992 panel in that "some [limited] control can be achieved by contraception."

Kirkpatrick (1994) states that there are "existing techniques and others on the near horizon that might have application in the control of goats in ONP."

Gill and Baker (1994) state that "Research is currently underway which, if successful, could significantly alter the conclusion that the lethal methods are the preferred control/extermination option." "In our own research with hormonal toxins, we are on the threshold of producing an effective contraceptive which will sterilize treated individuals for a lifetime with a single injection. It looks like the amount of contraceptive will be minute enough to encapsulate the drug into 'biobullets' which can be remotely delivered. These developments suggest that the conclusions of your review ought to be considered very tentative."

Jessup (1993, and again in 1994) states that the NPS should consider using removal "and/or" (emphasis added) shooting along with contraception "with the goal of reducing and controlling the mountain goat populations." "Given the low recruitment rate the ONP mountain goats are showing, I really believe that this could be accomplished." So, it would appear that, even in the present, it may be possible to accomplish reduction and control. Since Jessup uses the word "or", in his analysis, it appears that he accepts the present possibility that contraception without lethal shooting might actually work.

From Conservation Biology (Volume 7, No. 4, December 1993, pages 954-957), we see that Kirkpatrick has rated (in a 1993 interview with the Seattle Times) the success probability of population control technology for wild mountain goats as "very, very high." So, it is clear that contraception presently has, at a very minimum, an excellent chance of succeeding. (Please also see suggested improvements in capture and drug delivery elsewhere in this document. These suggestions could significantly improve the probability of success since they deal with the most difficult issue: field conditions)

If a combination of methods is employed, I would rate contraception a very likely success. (That is, I would rate the risk of failure, if population control was attempted today, to be very low provided that a combination of methods was employed.) More importantly, the future is more promising. (I would rate the risk of failure of future population control efforts to be extremely low--especially if a combination of methods was employed.) Given the slow or nonexistent rate of population increase, it appears that the situation can wait for the technology. It is not even unreasonable to think that a simple and effective contraceptive bait will eventually developed. However the technology companies (and academia) will not rapidly proceed in their wildlife contraception efforts unless wildlife managers show a track record of accepting imperfect technology over lethal techniques. The wait for perfect technology will be infinite. Should the technology fail to improve (which appears very unlikely), other options will still be available.

DEIS Scenario:
High Numbers of Goats Threaten Plant Taxa

There is considerable evidence that if the goat population eventually returns to its high of approximately 1175, plant taxa will survive. Some of the evidence for this is simple: When the population peaked at approximately 1175 in 1983, plant taxa did, in fact, survive. Note: This fact lessens, to some extent, the need (and costs) for vegetation monitoring, detailed for Alternative 2 in Appendix E of the DEIS.

Of course, the 1402 maximum sustainable population projection represents a higher density (than the 1983 census of 1175 goats)--in part because the 1402 projection is confined to the park where the 1175 figure was for the entire peninsula. However, actual field data of a peak (and a subsequent population "crash" to 288 goats) should take precedence over an extrapolation "based on census data and availability of potential goat habitat" (however scholarly). Also, it would also appear from my reading of the DEIS that, at the time of the maximum sustainable population estimate, the "crash" data were not available. In this case, it appears that the projection should be completely reevaluated. In addition, any future projection should include an estimated time table of population growth.

The possibility (discussed later) that mountain goats are native to the area significantly reduces the possibility that goats threaten plant taxa. In fact, if these animals are native to the park, then the risk that naturally occurring goat populations threaten plant taxa is obviously about as close to zero as one can get.

In my reading of the DEIS, I did not notice any mention of possible positive effects that mountain goats have on the local ecology. I believe that any positive effects should not be largely ignored in an objective analysis: they should be explored fully. I would ask the following question: Assuming that the ONP mountain goats are exotic, do native mountain goats in other areas have any positive effects?--are they at all important to local ecologies? If so, are these positive effects existent in other ecologies but somehow nonexistent in the ONP? It seems reasonable to assume that, in some respects, herbivores may benefit the very plant populations that they feed on (and sometimes wallow on) by destroying marginal individuals, allowing stronger ones to survive (much as predators actually benefit prey). Clearly, herbivores have some role in the environment--whatever it is. This role is surely a complex one and cannot be measured by simply quantifying one or two variables likely to show only a negative view of herbivores. There may be some possibility that, overall, mountain goats are, in fact, harmful to the ONP ecology. But the criteria by which they have been judged in the DEIS would find mountain goats incompatible with any ecosystem. In fact, the criteria by which the goats have been judged would find all herbivores incompatible with the entire planet!

If, in the end, a more balanced analysis still finds the mountain goats to be an ecological liability in the ONP, then at least a better understanding of the extent to which they are a risk can be determined and weighed against other considerations. I would rate the present risk to plant taxa as extremely low. However, if, in the near future, the projection of 1402 goats was somehow realized (and other protective measures were not employed, such as removal, contraception, fencing, etc. . . .) then I would rate the risk to plant taxa as very low.

This analysis is supported by a 1987 park Environmental Assessment that stated "there is no apparent danger that these species will be extirpated." (See National Park Service 1987a:22.) This is especially significant in light of the high goat population at the time. In fact, the possibility exists that high goat numbers (should they ever actually occur) may have complex indirect benefits on plant taxa (and wildlife). So, there is some risk associated with removal (or even population control).

DEIS Scenario:
Barrier Means, Such As Fencing, Not Effective in Sensitive Areas

Although the DEIS does discuss the possibility of using fencing in connection with site restoration programs, it takes the position that "it is not feasible to construct/maintain fencing around plants of concern because of size of areas, steepness, rocky substrate, and snowpack. Fences would interfere with natural ecosystem processes, such as snow accumulation and melting, and could harm the plants they were intended to protect." However, it is difficult to imagine that fencing could not be used in a number of situations. It seems that fencing could at least be used to protect the more sensitive, smaller populations of plants in situations where they are accessible--provided that fencing is not placed within a given radius (which should vary depending on slope and direction) so that plants do not experience microclimatic alterations (due to shade, wind, etc. . . .). In fact, from the research summaries presented in the DEIS, it is clear that exclosures can be very effective in the ONP. (See pages 162, 167 and 193.) The aforementioned radius limits should address possible objections. Layout and design of fencing should be such that normal movement patterns of wildlife are not seriously disrupted.

Options such as long-life, battery-powered electric fencing should also be considered. Although year-round protection may not be possible because of snow-caused electrical shorting, protection during the most vulnerable periods should be possible. Note however, that new, low-impedance models are resistant to ground-out problems (due to wet conditions.) Electric fencing does not pose microclimatic alteration risks to the plants they protect even when very near them. Electric fencing is safe to plants if placed only a few yards away. (The fence wire must be placed so that it does not contact rare plants. For this reason, when protecting tall-growing plants, it may be wise to keep fencing some distance from areas likely to be populated in the future. A few extra feet may be required to prevent the possibility of ground potentials and/or contact from fallen wires.) For very hard to reach areas, 25-mile range solar-powered electric fencing (with U.L. listing and batteries that are constantly recharged by the sun for up to 3 days of operation in the dark) is available commercially for an economical price. Electric fencing is easy to install in soft soil but can also be installed in rocky areas with some difficulty. Maintenance of a properly installed fence is minimal. Other types of fencing may be appropriate in special situations. Fencing, in general, has another beneficial effect: Keeping humans from trampling vegetation. (The threat of man is acknowledged on pages 105 and 202 of the DEIS.) Reasonable care must be taken to avoid trapping animals within the exclosure. The idea of using electric fencing should be evaluated for wildlife safety by, of course, a qualified biologist. Note that a possible limitation of electric fencing is that it should not be used near cliffs since startled animals may flee in haste and fall. In some cases, a combination of fencing types may be used depending on slope, direction, soil, proximity to cliffs, etc. . . .

Other barrier methods should also be possible. Solar powered noise devices can be located near sensitive populations of plants. Testing would be required to see which sounds--if any--frighten mountain goats. (Periodic gun shots and rattle snake sounds would be good first bets--rattle snake sounds may get a lot of effect for very little loudness.) A motion-controlled unit (power information not available) is commercially available between $1,200 and $1,500 and can be equipped with a video camera for test verification. However, with modern digital sound storage chips (complete with sound retention during power failures) running for a dollar or two, and low-voltage solar power supplies running for only a few dollars (the most expensive component would most likely be the speaker), it seems likely that inexpensive units could be found with a little effort. Digital recordings could also contain warnings to humans. In any event, if field testing was successful, such devices could be designed and produced economically even in small quantities. A possible limitation of noise devices is that they must not be placed in areas where they will seriously interfere with threatened or endangered wildlife. (According to the DEIS, helicopter noise, for example, is to be avoided in owl and murrelet habitat "as much as possible and would not be likely to have an adverse impact on either species.") It may also be wise to keep intermittent noise devices away from cliffs.

A very oversized mountain goat model may frighten other goats. (Animal behaviorist may be able to summarily dismiss or confirm this idea. Otherwise, testing would be necessary to make sure that such a model indeed frightened goats instead of attracting them.) A cougar, wolf, human, or bear model might also succeed. Models would have to be placed away from (or on the southern end of) sensitive populations so that they do not cast shadows creating microclimatic changes. Incorporating sound into models would not be difficult, and periodic low-energy mechanical movement is not out of the question. As with noise devices, a possible limitation of animal models is that they must not be placed in areas where they are likely to seriously interfere with threatened or endangered wildlife.

It seems likely that most any movie studio (such as Universal Studios, Paramount or Disney) would be more than happy to lend excellent animal models for testing since it would be for a good cause and because it would probably earn them a story on the nightly entertainment news shows. Should testing prove successful for any of the barrier methods, design contests (engineering schools would likely participate) could be held for practical, low-cost solar-powered implementations that are light, stackable, and easily transported on foot (when disassembled). If such methods seem to be too involved for widespread use, it should be remembered that the DEIS makes the point that the smallest of the subpopulations of astragalus "contains only 37 plants in an area approximately 65 ft. by 80 ft." Such sensitive, small populations are prime candidates for barrier methods.

Overall, I would rate the ability of barrier methods to protect plants to be very high. I would rate the risk of failure to be very low (even if all other DEIS projections somehow become true).

The use of barrier methods should increase confidence concerning the health of the most sensitive plant populations and, therefore, reduce the need for both goat and vegetation monitoring. Note that in the three DEIS presented alternatives, vegetation monitoring was the largest single projected cost. Thus, barrier methods have great cost-reduction potential.

Having considered the various senerios presented in the DEIS, it appears that risks to rare/endemic plant taxa are extremely low. The "extremely low" conclusion is supported by the mathematical reality that when a chain of unlikely independent events are each necessary for a result to occur, the result is extremely unlikely. (Note: Even if one or two events were likely but two or more were unlikely, the result would still be very unlikely.)

It should be noted that, it can be argued that other "chains" of events may be hypothesized which might cause harm. Such an analysis is beyond the scope of this paper and, unfortunately, the data presented in the DEIS. (However, possible combinations of events have been hypothesized in preparation for this informal analysis and determined to be insignificant.) In parallel with this concern, it should also be noted that real-life events are seldom completely independent and the subject analyzed here is not an exception. However, a worst-case analysis that assumes that dependencies conspire against plant taxa indicates the same conclusion: that plant taxa are at little risk.

In fact, plant taxa may be at greater risk without the goats. This risk can be weighed against the extremely small risk of not removing the goats. However, since the risk due to the presence of goats is so small, the conclusions of this paper are not dependent on the claim of risks associated with removal.

Finally, many of the factors that can or do reduce risk to vegetation also reduce risks to archeological resources, which are clearly less difficult to protect. For this reason, I have not devoted additional space to archeological issues.

Risks to Personnel

It appears that risk to personnel has been well assessed. Personnel risk was a major concern of the DEIS. However, I feel that I may be able to offer a few rather non-conventional ideas that could be evaluated for significant potential risk reduction (as well as cost reduction). Some of the suggestions presented here are engineering-oriented suggestions. They may seem a little farfetched to the non-engineer. However, I believe that they would be viewed as promising by other engineers.

The DEIS makes it clear that the most difficult hurdle in population control is not contraceptive technology, but field conditions. Helicopter safety appears to be the overriding concern in removal and contraceptive-delivery operations. I have argued elsewhere in this document that there is little need, at this time, for field actions such as capture operations (except, possibly, in the most sensitive areas.) Taking no action eliminates virtually all personnel risks.

However, should field actions truly become necessary, or should the NPS insist on them, I propose that the ONP consider the use of radio-controlled model aircraft for use in a variety of operations. Use of R/C aircraft may have potential to drastically reduce risks to personnel and radically reduce costs. R/C aircraft may also increase operational efficiency (as measured in personnel hours) in a number of situations.

One use would be in connection with drive netting. The DEIS reports that the use of drive nets was abandoned because of expense and efficiency. Low-speed, low-altitude flying was also reported as dangerous. R/C aircraft are virtually free of major personnel hazards. A helicopter may still be necessary for personnel transport and transport of animals to staging areas. However, total flight hours could be significantly reduced and limited to safer operations. During times of the year when some goats are located in more accessible areas (or near such areas), transport may not even be a need.

Radio-controlled helicopters are commercially available in a variety of sizes for far lower costs than real helicopters (although model planes are even less expensive) and could be assumed to be frightening enough to motivate goats to flee. (Different scale aircraft may have different effects on mountain goats. It is possible that the smallest scale aircraft would be viewed as some type of pest to avoid, where larger models would cause great fear; noise devices could also be installed to assure a reaction.) R/C aircraft have reasonable flight-time capabilities. However, if longer times are required, multiple aircraft can be used on a time-contengency basis (with refueling rotation). In addition, multiple R/C aircraft could fly simultaneously to exert a high degree of control on the direction of fleeing. (With real helicopters, multiple aircraft in close proximity presents an enormous safety concern.) The park service could elicit the help of hobby groups for advice (and even direct participation) in tests. I think it likely that many would be delighted to see their hobby applied for practical means. The idea seems worth a test. (It is my hope that any test would be limited to moving animals to a specific location. Again, I do not feel that other actions are generally warranted at this time.) Good quality radio control units would be a necessity because of the terrain. The operator may have to be positioned in higher ground in some situations to avoid signal loss.

According to the DEIS, "During prolonged aerial pursuit of individual animals during darting or netting capture attempts, animals can sometimes be maneuvered toward moderate terrain. As the animals become exhausted, they are captured 'manually' by using heli-jumping." It stands to reason that R/C aircraft could also be used to force goats into moderate (or even ideal) terrain where they can be captured "manually" or, at least, darted if necessary. In some situations, driving goats might require minutes. In other situations, personnel may have to intermittently travel by foot as individual goats are slowly driven, in steps, to an appropriate area. If a real helicopter is not used to then transport the animal to a staging area, contraceptive (or possibly sterilization) procedures must be performed on site. Silastic implants of MGA (as well as other contraceptives) could be reevaluated in this context since a major objection to this method was based on the difficulty of capture. Also, other contraceptive methods should be forthcoming in the near future. (See Gill, page 183.)

Other possibilities for R/C aircraft are more complex and less likely. However, a cursory investigation into the possibilities is warranted. One possibility is the use of net guns with large R/C helicopters. Radio control of net guns is easy to accomplish. However, the weight of three or four-barreled net guns would likely be too great for even the largest commercially available R/C helicopters. A solution to the weight problem might be smaller, lightweight nets, opened by three or four small model rockets (controlled, of course, by commercially available radio controls and fuses). Again, I believe that R/C enthusiasts (as well as college engineering classes) would jump at the chance to participate in a design contest--because the challenge is so interesting. (Note: A limitation to the use of model rocket motors and possibly even the use of some R/C models may be the prohibition of use during any dry season.) The possible loan of larger military drones (assuming helicopter models exist--which seems likely) is also worth a simple inquiry (even if the request is likely to be rejected).

Another less likely possibility (but worth a cursory investigation) is the use of military drones (or even commercially-available, R/C model aircraft) for biobullet delivery. (Note: Biobullets have been successfully manufactured, but have not been tested in goats. Serious humane concerns have been expressed about their use--especially in large calibers. In terms of concern for the goats, the use of biobullets is, of course, better than lethal methods.) Although I feel that biobullets (and large-scale goat management in general) will not be necessary, I do think it wise to develop a delivery system as an alternative to lethal shooting.

The use of biobullets would have the advantage of being applied remotely since no animal recovery would be necessary. I would be somewhat surprised if military drones have not already been fitted with light weaponry. In any event, as an engineer, I do not find it difficult to envisage the installation of a servo-controlled, biobullet gun (modified for weight) on a commercially available R/C model helicopter. Dual, dart and paint-pellet delivery is another possibility. Commercially-available, very-small lightweight cameras that broadcast reasonable distances on conventional TV channels (available from security specialty companies) could be easily mounted on model aircraft for targeting use. Convenient, lightweight battery-powered receivers are available for about $150.00 that can be carried in a pocket. Design of the targeting system is no more complicated than orienting the gun and camera in the forward direction of the model helicopter and drawing a cross-hair on the TV receiver. Another possible targeting system involves a simple sighting laser (commercially available) or laser pointer. The beam can be observed from the ground to confirm proper aim. (Note: A limitation of pointers costing approximately $100 is that they are not effective in bright sunlight. Brighter units are available at higher costs.) The only real engineering questions are weight, gun safety, and the survival of aircraft during the shock of firing. These problems appear to be manageable. I do not believe that the design questions are outside the skills of any good model-design hobbyist--let alone, an engineer.

A modification of the above idea might include a quiet, fan-powered R/C balloon. Goats may not be frightened by such devices since they make little noise. However, gun safety and wind problems might make greater design demands of anyone putting together such a system. It is also possible that a radio-controlled, square balloon (approximately 20 to 40 feet in size), or a simple square structure made from tubular balloons (which are easier to purchase), could carry an already opened, lightweight net on its under side, ready for R/C release. (The suspended net would also act as a balance to keep the balloon upright. Although I will not include details here, a heat-controlled ballast to control altitude should be feasible. As mentioned for other ideas, inexpensive lasers and/or television cameras could be used for drop-net targeting. A long line would remain connected to the middle of the net, or a ballast would deflate, to prevent loss of the balloon once it was free of the weight of the net.) Of course, balloons would not work if mountain goats are frightened by quiet floating objects directly above them. Again, from an engineering point of view, this idea may be more practical than it may sound to some. In fact, a low-tech solution might also be possible: A balloon could be positioned above individual goats through the use of hand held cords. Capture distances could still easily exceed those of net guns, which are generally limited to 32 feet. (Net guns have been used with some success.) In any event, I believe that unusual technical solutions deserve consideration (and possibly a try) and should not be rejected simply because they are unusual. A possible reduction in personnel risks and a reduction in taxpayer expenditures gives these odd ideas enough credibility to warrant an examination. A possible legitimate objection to the use of balloons would be if wind conditions are usually unfavorable in the areas in question. (Again, I hope that any testing would be conducted in such a way that no harm occurs to animals.)

Another strategy might involve using a model of a small (nonthreatening) goat in combination with a salt lick as a bait site for dropnetting. Although plywood decoys have been tried, more convincing decoys might be attempted. According to the DEIS, in 1982, 47 mountain goats "were captured using a dropnet with salt bait" on Klahhane Ridge alone. The total cost of the operation was $21,130. It would appear that if this method was improved and applied (which I do not think will be necessary) at various sites, large numbers of animals could be captured at a reasonable cost. (The 1994 estimated population is 288 32 and appears to be declining.)

Another idea to drive individual goats into drive nets would entail the use of portable battery powered electric fencing. A thin (and light) electric fence wire would be strung out (in a long as possible distance) between a number of personnel (or volunteers) who would operate as a moving fence, changing shape as necessary to direct and control goats (and even enclose them). The personnel would not actually come in contact with the fence. Instead, each person would use an insulated pole to hold the wire as needed. The person carrying the power unit would drag a ground wire or wear insulated, metal-fabric boots (for grounding) identical to those used by certain high-tension, electrical-wire workers during "hot" maintenance. An alternative would be a two-wire fence with no ground. In either case, a very small battery, a small (but standard) battery-powered electric-fence control, and a little tape is all that would be necessary to create the power unit. The fence could be marked with tied-on sections of survey tape to encourage the goat not to contact the fence. It should be noted that if a mobil electric fence is not, by itself, sufficient to capture goats in a majority of situations, it may be helpful once goats are "squeezed" into the drive net area by other methods. A variety of methods, each appropriate to the situation, should be employed if control becomes necessary. The moving electric fence idea, while feasible, would likely increase capture trauma and thus should be evaluated for possible capture mortality problems.

It should be noted that, in recent years, there has been much renewed interest in (and actual use of) non-lethal "weapons" by the U.S. military. This has been fueled by mission changes as military personnel are increasingly involved in difficult, police-like actions. A cursory investigation into progress in this area (and application to mountain goats) seems warranted.

It is my belief that all goats captured by any method should not be killed. As an alternative, they should be eventually or immediately returned to the ONP in a temporarily or permanently sterilized condition. (It is also my belief that most of the Olympic mountain goats should be left undisturbed for reasons that will become clear later in this document. In addition, any procedures should, of course, be conducted with great concern for animal welfare. Health risks and the induction of trauma in these animals are acceptable only if greatly minimized and only where truly necessary. Unfortunately, this ethic must bend somewhat in light of the lethal alternatives being considered by the ONP--mortality rates from capture and other humane issues would be a moot question under a lethal shooting program.) If sterilization methods (such as vasectomy and tubal ligation) are employed that do not affect goat behavior, then reintroduced goats can redirect the sexual capacities of never-captured goats so that natural, reproductive efforts of other goats will be unsuccessful. (This reduces the need for future operations and therefore reduces personnel risks. On a per capita basis, lethal methods do not have this advantage: Killing a male goat, for instance, actually increases the reproductive success of other males instead of inhibiting it.) This may be especially true of large healthy males who are likely to enjoy a mating monopoly as they suppress the mating attempts of other males. However, sterilized females would also have an effect since the propagative energies of males would sometimes be spent nonproductively. As an alternative to shooting, it is even worth considering introducing sterilized animals to the ONP that have been removed from other areas. Although a temporary increase in population certainly occurs, the long-term effects should be favorable. If sufficient numbers of sterilized animals are introduced, the population will surely crash (humanely, provided that numbers do not approach peak population figures) due to lack of reproductive success. Although this takes time, I believe that the time is affordable if barrier methods are employed in the interim to protect plants. However, it well may be that the introduction numbers required are prohibitive. This is a question for biologists with appropriate computer models. In any event, every reintroduced animal exerts some control. Alone, this may not be an answer; with other non-lethal methods, it may be significant. Radio collaring of these animals might also help locate other animals. Finally, if these animals were slightly habituated to a feed pan and/or salt lick, they might be humanely used in baiting situations for capture or contraception efforts.

As previously mentioned, I do not necessarily claim that any single idea presented here will, alone, facilitate control of mountain goat populations. More importantly, it appears that control efforts are not necessary. Control, if it becomes necessary, may depend on many non-lethal strategies that are targeted according to the time of year, area, sex, age, weather, and availability of new techniques (including experimental ones). Removal from sensitive areas could be by any of the non-lethal means already reported in the DEIS, as well as means mentioned above, and by additional means suggested by other members of the public (and the scientific community). Removed animals (and, possibly, additional animals) can be sterilized and reintroduced as an additional control. Contraceptive means can be also delivered in the field (without removal operations) as outlined in the DEIS, as well as by means mentioned above, and by means suggested by other members of the public and the scientific community. Again, it should be mentioned that, considering the present population level and a historically low rate of increase, the introduction of major control efforts can wait for emerging technology. Waiting for soon-to-emerge technology may also reduce personnel risks since future contraception delivery methods may be safer.

In general, it appears that risk to personnel is very low if helicopters are not flown at low speeds when near the ground. Any techniques that reduce the need for helicopters in such situations increases safety. A plan (such as Alternative 2) that includes little or no aggressive field action is the safest.

Risks to the Olympic Mountain Goats

Risk to Olympic mountain goats is a valid concern that the DEIS recognizes. The document details injury and fatality rates related to capture techniques and sterilization. It also includes a reference to "the park's humane philosophy." The DEIS reports a commitment of a panel of scientists (who commented on the contraceptive efforts) to "developing non-lethal control techniques for wildlife." The question of killing mountain goats to achieve management goals is also raised in the context of public sentiment and concerns of ethicists. It is safe to say that the public view concerning animal issues is in a rapid state of change, accelerating toward more empathy for individual animals. A substantial degree of weighting to the park's humane philosophy (and the public's concern about the plight of the ONP mountain goats) is not an unreasonable inclusion in an analysis of competing risks. (Ironically, in my analysis, there is a solution--Alternative 2--in which the risks are not "competing." Instead, they are all minimized.) With respect to shooting the goats, the lethal risk is close to a factor of 1.0 (a near certainty) for each goat that will be shot under two of the proposed alternatives. Another way of looking at it is, if x number of goats are planned to be killed, then the "risk" is close to a factor of 1.0 that approximately x goats will, of course, be killed. So, the "risk" part of the analysis is, of course, trivial (since the risk is intended and highly controlled); the difficult part is a weighting (formal or informal) to the significance of the lethal events. I do think that virtually everyone would agree that, all things equal, it would be better if the goats were not even disturbed--much less killed. Few would rate the importance of the lives of these sentient creatures as trivial, many would attach rights (which are almost absolute) to the Olympic mountain goats as a population, and others hold that individual goats have rights. Therefore, a substantial weighting--a weighting that recognizes a remarkable shift in public opinion and philosophical tenets--should be applied to these concerns, especially if NPS mandates are unaffected, or even helped, by such consideration.

In addition, there is an entirely different sort of risk associated with the mountain goats. The DEIS has concluded that the goats are a non-native species. (The DEIS traces today's goats to a human introduction in the 1920's.) Again, this conclusion is not presented (in the DEIS front-material) in terms of probability. There is considerable risk that the DEIS conclusion is wrong. If the DEIS is wrong and the mountain goats are killed (or even just removed from the ONP), a naturally occurring population will have been eradicated from its native range. It is by no means certain that mountain goats are exotic to the ONP. Evidence that mountain goats are a native species includes the following facts:

These reports--all written before the release of a few goats in the 1920's--are substantial evidence that mountain goats are native to the ONP and an important part of the local ecology. However, the DEIS presents some interesting research that claims goats were not in the area before the 1920's. The goat wool reported in 1844 and the mountain goat items reported in 1917 are hypothesized to be the result of trade. The published reports of mountain goat sightings are criticized for a lack of detail such as a description of the animal(s), date of specific goat sightings, and [specific] locations.

I agree that, in one respect, original notes of the explorers would constitute better evidence since it would indicate diligence. However, there is evidence that at least one author was clearly credible--and no evidence that the others were not--and all publications were credible. The detail, if we believe the authors, appears to be sufficient to establish that:

  1. the animals sighted were goats (since two of the sightings were by explorers--not causal observers--and since goats are difficult to mistake),
  2. the date was before the 1920's (since the articles were clearly written before then), and
  3. the location was the area in question (since there exists a map of one expedition, and since the titles and/or details in the articles clearly identify the general area.)

Another point brought up in some of the research presented is that there should have been more references to the mountain goats if they really existed in the area. However, a small population, dispersed in such a territory, would only be detected infrequently--not regularly. Infrequent contact can especially be explained by the nature of mountain goats: they live in higher elevations--especially in the summer when human travel is more prevalent. So contact could easily be infrequent. Out of the few who made an unlikely contact with goats, even fewer would actually write about it. Even fewer would likely submit the writings for publishing. Even fewer would actually be published. And even fewer of the published accounts would survive and be discovered.

Clearly, the DEIS position might be correct. Just as clearly, there exists, at a minimum, a credible possibility that a native population of a species could be exterminated. This risk should be analyzed not in the "black and white" terms of the DEIS but in a more realistic estimation of risk. Otherwise, the wrong decision may be made.

The credible possibility that mountain goats are native to the ONP means that they may play a beneficial role in the ONP ecosystem. Removal of goats would thus represent a possible artificial change in this system and warrant a very large expenditure in monitoring of all vegetation and wildlife for possible changes. It is important to note that while the present concern is for rare/endemic vegetation, goat removals could have unforeseen indirect effects on such vegetation as well as threatened and endangered wildlife. It seems clear, that extensive monitoring (under "goat elimination" Alternatives 1 and 3) would be necessary because of the possibility that the goats are native. This fact appears to make necessary, a restructuring of the reported cost estimates in favor of Alternative 2. It is important to note that Alternative 2 represents no rapid environmental changes, while the "goat elimination" programs would produce lightning-paced, radical ecological change with possible unforeseen effects.

In addition, there is a financial risk, that removing the goats will prove to be ecologically unsound and therefore in need of reversal. This may occur if detrimental ecological changes take place, if new historical information is discovered, or if the presently available historical data is viewed differently by future administrations. Thus, an expensive emergency capture and release program would be required to restore the goats to appropriate numbers.

For all practical purposes, Alternative 2 has been "field tested" for decades and with a peak environmental loading of 1175 goats in 1983. Thus, it would appear that monitoring efforts (and costs reported on page 224) should be far less for Alternative 2 than for the "elimination" programs (Alternatives 1 and 3). Therefore, if my analysis is correct, Alternative 2 should be the least expensive of the three choices presented in the DEIS.


There is an extremely small risk that rare/endemic plant populations (not listed as endangered or threatened by the U.S. Fish and Wildlife Service) may be destroyed by mountain goats. This risk is extremely small (and possibly insignificant) in light of the chain of unlikely events necessary for it to occur. Furthermore, the fact that mountain goats surely have beneficial effects on at least some environments is of great importance. Should the Olympic mountain goats be eliminated, the effects of the total loss of any beneficial effects on the ONP ecosystem is unpredictable since the possible positive effects of mountain goats is an issue that seems to have been largely unexplored. (In fact, in the DEIS, mountain goats are simply assumed to have detrimental impacts when no data is available. For an example, see page 84 of the DEIS.) Since risks to plant taxa (due to total elimination of goats) have not been adequately explored, some presumption of risks must be made. However, I would rate the risks from goat elimination as only moderate to low, but not as low as the extremely low risks from goats remaining in the park. In other words, there is risk with either choice, but I claim that it is ecologically riskier to eliminate the Olympic mountain goats. Goat elimination may cause the very problem that it is intended to solve and could create other unforseen problems for wildlife.

There are also risks to personnel to be considered, but these risks can be virtually removed if the goats are left undisturbed. If the park insists on actively managing the goats (which appears to be undesirable and unnecessary), some of the ideas I have presented may help reduce risks to humans and animals.

Should the ONP remove or kill all of the Olympic mountain goats, there is also a risk that the animal population destroyed will be a native one. Elimination of a native species from an ecosystem is clearly unacceptable to everyone.

Contraceptive means appear to exist (and future means are expected) that could at least reduce the goat population should it begin to increase at its historical rate of about 2%. (However, evidence indicates that the population may presently be experiencing a crash.) Furthermore, present and future means of capture and contraception may be even more successful if a number of new and old suggestions are employed on a situation by situation basis. A combination of removal (if absolutely necessary) of some goats from sensitive areas (with sterilization, marking, and reintroduction elsewhere in the park), various contraceptive methods (delayed unless an immediate need develops), plant barrier methods (keeping humans and goats out of the most sensitive areas), and monitoring appears to be the best strategy--particularly since all risks can be minimized. (This course of action also appears to be the most fiscally responsible one.) This should also satisfy public concerns about the welfare of the Olympic mountain goats while meeting all park objectives.

David Lee Winston Miller is a candidate for a Masters of Science degree in Computer and Information Science, specializing in General Systems Theory at the State University of New York--Utica/Rome. He also holds a Bachelor of Arts in Audiology from the University of Tennessee. Mr. Miller is presently employed as a Project Engineer specializing in Control Systems Technology and jointly holds mechanical patents in a couple of countries. He has been actively involved in animal and environmental issues for well over a decade. In addition, Mr. Miller has served, as a volunteer, in a variety of board and executive positions with a number of organizations addressing such issues.

This document was originally published in July 1995

Copyright 1995, 1996, David Lee Winston Miller. This document may be linked or copied in any form provided that it is reproduced in its entirety with copyright, references, and links noted or included. Mountain goat graphic modified from "Wild Animal Silhouettes" by Ellen Sandbeck, Copyright 1992, Dover Publications, Inc., 31 East 2nd Street, Mineola, NY 11501. This page has been accessed  times since 11-07-96.

Homepage         Comments