Cross-posted from Hippies 4 Horses by Afroditi Katsikis

This is the text of the meeting of the Wild Horse and Burro Advisory Board meeting held on September 9, 2013 in Arlington, Virginia. This text was copied directly from the captioned text of the live-streaming meeting– no edits or corrections – only some spacing to make it easier to read. I can not assure you that it is complete as per the meeting – it is as complete as the captioning was on live streaming.]

[Joan Gullifoyle starts the meeting]
>> Okay, hi, everybody. We’re going to get started in two minutes. So, if you would take your seats and finish up your conferences. I can’t couldn’t see because I can’t see that side of the room. But Carol Lanne of the National Research Council National Academy of Sciences is here and she was very involved in the coordination of this report. I wonder if she would take a moment to explain in digs to the 11 questions that the BLM asked of the NAS, what the — what the report did and didn’t cover intensely just to help it be more clear why some things may be in and some things may not be in and what you’re leaving it to us to kind of do with the report if you wouldn’t mind.

>> Sure. Thank you, again for making the report your focus of this meeting. My name is Cara. I’m the study director for this particular project. And I’m a program officer at the National Research Council.
As you’ll find had your report on page 2 and 3 in the summary and then again on page 16 and 17, that’s a statement of tasks that the BLM and NAS agreed that the committee would look at.

Following that there’s a section in the report called bounds of the study which goes through what we were asked to look at and what we weren’t asked to look at. As you’re all aware there are many issues that are not related to science but national research committees are commissioned to look at science-based questions so that lends itself to the nature of the questions that are part of the statement of task which have to do with population estimates, range land estimates. Population growth. Genetic diversity population control. Things outside the bounds of science supp as how many animals should be on the range — because that’s a policy question, not a science question. Forage can tell you how much forage, how animals may use the forage, how different animals may use the forage but how many animals you’re going to have out there and what kind of animals they’re going to — for instance livestock and wildlife or some of all. That’s a question for policy makessers to make. So the purpose of the report in answering the science question is they can inform the policy conditions. That’s the role of BLM. Not the committee.

>> And Kara, if you don’t mind just for clarification, what BLM asked the NAS to do was look at these 11 questions about the science of the animals on the range. But not to look at perhaps the multi use nature of the rest of the agency for example. Is that right?

>> KARA LANEY: Yes, things like whether the law could be changed, whether different laws perhaps conflict with the Wild Horse and Burro should be changed, whether the allocations can be changed. Those are policy conditions that the policy was not asked to look for?

>> JOAN GUILFOYLE: And the logistical political educations that were asked for that were science based you’re leaving up it up to the receiving agency to determine the feasibility of some of it.

>> KARA: Correct. So the budget associated with recommended actions was not part of the committee’s purview. As you’ll note in the report, we do find that — we do go so far as to say an option will be expensive such as continuing to move horses to long-term holding will be expensive and the committee thought that its recommendations would be less expensive in the long run. You about we don’t go through and put a financial — attach a number to those actions or to the actions of gathering animals, implementing fertility control, anything of that nature.

>> JOAN GUILFOYLE: Boyd, just one more thing if I could add. The NAS, if I can say for Kara just for a moment, was concerned just as we were that there were a lot of interpretations about the report immediately after it was filed and to look at how we could correct that or they would correct that. They did do a posting on the Web site and I don’t know if you’re going to be here, Kara but I’d like that to be part of what we share with everybody.

>> Sure, I plan to be here the whole time.

>> JOAN GUILFOYLE: So we can go over that briefly now or wait until public comment period or Wednesday morning. It doesn’t matter. But I want to make sure that you all hear what they felt the media had misunderstood about the report and to be clear about what they said.

>> DR. BOYD SPRATLING: Kara, just so you know, we’re going to take latitude as a board to ask questions while we have experts in the room. We fully understand that a lot of the questions we ask were not part of the boundaries of the study and report. While we have people here that understand the dynamics out here, we’d like to ask questions to help us from our recommendations to the BLM.

>> Please make use of our expertise, they’re helpful to us and I hope they’ll be helpful to you, too.

>> DR. BOYD SPRATLING: Thank you. Next I’d like to introduce Dr. Robert Garrott. He’s a faculty member of the Department of Ecology at Montana State University.

>> ROBERT GARROTT: Thatch. I have two back to back talks one dealing wees mated population size and growth rates and then I’ll give another one on chapter 6 as well.
So as everybody can imagine, how many horse there are out there and how fast they grow is basic fundamental actions for the problem might take. So our job in this chapter was to look at those things and the objectives within that chapter was to review the the method used to inventory horses by the BLM and provide potential recommendations for improving the methods. And thirdly to review the data available to estimate growth rates. Typical growth rates of course on western range lands.

BLM spends about 1% of the wild horse and burro budget. Inventories horses, Marx sure the animals are accounted, HMA herd management area periodically and those counts are then used as your foundational knowledge for population management. Key attributes of any scientifically rigorous are good survey method. Scientific standpoint is that the methodologies be rigorous and standardized, that there be a statistical basis for how that’s done. That they’re consistently applied, they’re well documented and that the data that comes from these are complete, organized, and accessible. So that’s sort of the gold standard for what we’re looking for in a scientific inventory program for wildlife.

Our expectations before we actually looked at any of the BLM databased on our understanding of how the numbers will be generated and this partly comes from my experience in the 1980s when I worked on this issue by population dynamics of horses and potential of contraceptive control on horses in the 1980s and I worked with all the BLM offices was that there be a periodic count. So someone from the Bureau of Land Management most most likely wind up in a helicopter or airplane and flew through the herd management area and counted all the horses they could see. That count could be reported as a population estimate which you see this column here in this fictitious table that’s empty right now or that count could be modified to come up with a population estimate. If you think you counted them all, in other words, you had a perfect census of the heard management area, you just translate the number you counted into that population estimate.

But in general what we do know about counting large mammals and many, many places is that we don’t count them all. There’s a bias, what proportion you counted can be use if you have an estimate of that or even ball park figure, you can take and modify your count for the proportion you think you actually counted or the proportion you missed and come up with a population estimate. For example if we couldn’ted 422 animals and we thought we counted 80% of them, if you divided 422 by .8, you getten an estimate of 527 animals on that heard management area you’ve accounted for approximating what you think you might have missed.

But we also know that we don’t have the budget or the agency doesn’t have the budgets to count these populations every year. So, if you need a population every year. And you’re not going to count those years and you have some way of projecting that population from one year when you did do the count to the next years when you don’t do a count. And in general the expectation is you’re going to multiply your population estimate in one year by what you think is a growth rate of that population over the next year.

New England this case, this example, if think that the population is growing at 20% annually, we multiply 527 by 1.2 and we get an estimate the next year. And next year, 2003, there’s no count, you do the same thing again. Get a pop ration estimate. 2004, you get a count. You do whatever you’re going to do to that count to turn that into a population estimate and that’s — with the expectation of what was being done with the Bureau of Land Management. Then of course for national statistics if you do that for herd management areas every year and you aggregate population estimates for haul the herd management area, you get statistic how many horses you think are on the western range land throughout the west. So that was our expectation.

And to see if that’s what was actually being done, the NRC committee requested the BLM provide us records from 2000 to 2011 for all herd management area. The BLM was –’s response was that there was no — excuse me. No centralized database. The data was first among field office thought that request wasn’t manageable for what they asked it to do, they thought what would be more manageable would be if we requested records from agencies and suggested a maximum of 40.

And so what we did is the committee selected the 40. HMAs that requested data for and these are the distributions of the sample of HMAs that we received data from BLM versus the number of HMAs based on criteria that were available to be sampled. So we got records from 40HMAs represented there across those states based on the sample we requested.

And this is an example of what we got back. We provided BLM the national office a standardized table so we could get the data back in the same way. And I’m going to spend a little bit of time with this to point out some attributes of this. So this — I’d say, is a typical record for one of the herds we saw. And you request see that there’s attributes across the columns and there’s data filled in and there’s places where there aren’t any data. So those are the actual population counts. And the first thing you can see is that they’re irregular as far as when they were conducted. So sometimes we had two counts. Back to back and other times we had a year in between and sometimes in this case, a couple years in between. So there’s some inconsistency on how frequently this herd counted.

There’s a real inconsistency on when they were counted. And this is important because there’s a birth pulse where the population increases because of foaling and then throughout the rest of the year, there’s attrition, animals are being lost. If you count the animals at different times, you’re counting for a different amount of the seasonal mortality that occurs. And it makes the data less comparable from year-to-year.

So there’s inconsistency in these records in timing of surveys. And here you can see that there’s inconsistencies when the survey platform. Usually a helicopter was used. And then one year a fixed wing airplane was used probably because of logistics or perhaps budget. Fixed wings are much cheaper than helicopters. But the proportion of animals you can detect different survey platforms to be very different. So by changing your survey platforms from year-to-year, you’re probably adding additional variability to the count data that has nothing to do with what the population is doing. It’s due to your changes in the methodology you’re using to count them.

You can also see here that there’s also incomplete counts. Here’s that year that the fixed wing was used, only 70% of the area was surveyed. So that that adds a little bit of problem to interpreting those data as well. Now, if you look at the relationship between the population count and the reported population estimate. You see there’s a difference in this. There’s a first record. 190 were counted but population estimate was 164. You look over there at the last column on the right, the adjustment to the count was filled in as none.

But it — but the numbers are reported as being different. We’re not quite sure why those numbers were different there could be many reasons but we don’t know why. If we go down this column and look at how that number changed between counts and the population estimate each year, you see this year, the population estimate was a little lower than the count. Next time that incomplete count was 60, for some reason it was a really low count. But the population estimate — next year the population estimate was exactly the same. The next year it was counted, a little bit higher than the count. Population estimate was a bit lower. Next year is much lower. Next year little bit lower. Next year could be accounted for if there’s removal between population count and population estimate and that was accounted for.

We also had data on all the removals so we had the ability to cross reference these data to see if some of the discrepancies especially if they’re lower, sometimes that was the case. And then you had the years that there were no counts and the population was projected for those years as well. And we could find real consistencies if we adjusted the math just to see if there was a multiplier effect for population growth rate. If there was, it was very inconsistent from one projection to the next on the years that there weren’t counts. Here’s a different record that recommends another HMA that represents the best record we had at the 40. You can see that everything is consistent. Consistent use of both vehicles and horses. So these are ground counts, consistent time of the year, area was completely covered. Population counts were believed to be a census and they were always reported as population estimate.

Herd happens to be one that the USGS was conducting research project on supported by the BLM and so this might be a reflection of a research activity this herd is very much like the prior herd where the horses are all known individually by color patterns, there’s a lot of people to keep track of every horse. So this is an example of a small isolated herd. This would be represented of the horse record we received where there’s very little data filled in. There’s wide gaps, inconsistency where they’re counted and no population estimates filled in at all. We can just assume that the count was projected as population estimates since nothing else was provided in this table.

This gives you a representation of the types of data we received from the Bureau of Land Management from their inventory program. The other type of data you get is aggregated data which is reported on the BLM Web site that gives you an idea of the — excuse me, trajectory of the population range wide population throughout the west. Each of the data that have been reported on the Web site. This is important because it’s interpreted by public administrators. Gauge success of the program. It’s used in formal government review programs from the government accounting office reviews. And foundational data for planning and budge teary decisions that go to the Congress as well.

Given some of the data we saw in the field offices we weren’t quite sure of the national number. We started a conversation with Bureau of Land Management national office asking for an explanation how those national statistics came about. We were provided no documentation linking the national statistic to the field office. We compared the field office data from the sample of 40 HMAs and looked at what was reported on the Web site for those HMAs that were aggregate for national statistic. So Web site reports the sum but it also reports the number of population estimate for each HMA. I’m sorry, I had surgery just a little while ago on my throat.

And so when we did that, we found quite a few discrepancies between the field office data provided to us in given years and what was reported on the Web site. And we received no explanation of those discrepancies. So we just link — we can’t adequately link the field data to the national statistics.
So the committee conclusion regarding quite a few methodological flaws, inconsistent methods. It was also noted very often and with the public gave us testimony about movement of horses among HMAs. Which can confound at the HMA level that horsees are freely moving back and forth and adds variability to interpreting data. These are just straight counts.

Go up in a plane or helicopter and count all the ap malls you can see. So there’s no it at that time tickal method so there’s no proportion of animals detected which can be very substantial. The proportion that are missed. That’s giving a statistical range in what’s a plausible value given the data you collected and certainly inadequate recordkeeping and database management.

So we concluded that BLMs current herd inventory procedures don’t meet the modern standards management applied for most other systems where we’re required to inventory populations that we manage. Given how the data were collected and reported, we we concluded that the population estimates that are provided are likely substantial underestimates simply because we know we don’t count them all when we go up there and there wasn’t in most prop Asian estimates counts weren’t reported directly and if you didn’t count them all, it would suggest that the population estimates that are reported are underestimates of the number of animals actually on the range.

We also noted that this is the exact same conclusions that were made 30 years ago by the NRC committee that was in place when I was doing my Ph.D. in the 1980s and so this has been and this seems to not have changed since the program began. We also noticed there are attempts to improve the inventory program.

2010 wild horse burr and burro management handbook is published a rigorous set of guidelines for survey techniques and these are an excellent set of guidelines that mimics what I dictated as what would be ideal attributes there at the beginning of my presentation. They’ve also been working hard to aggregate HMAs where there’s a believe that horses are freely moving among HMAs and to essentially come up with more reasonable biological units to conduct surveys over which are called HMA complexes. So, if there’s no fences or fences are permeable between heard management areas, aggregate them together. Census and manage that as one population so it’s more interpretable data. That’s an improvement or could be an improvement if it it’s implemented.

And finally, BLM has had a partnership with USGS for about a decade to develop and test statistic rigorous survey methods. There’s a lot to be had from that 10 years as far as developing methodologies and horses can be counted well and scientifically rigorous. You can estimate detection probabilities and do a better job in that decade of collaboration certainly gives us some good science.

So our recommendations for improving population monitoring is that those two things that have been initiated, those guidelines from the 2010 handbook and the HMA complex initiative that those things actually be implemented well and consistently across west and evaluated on a routine basis. We also suggest that BLM should consider more intensive monitoring for what we call sentinel herds. So I’ll relate a little bit more to sentinel herds and why this is important, why we think this might be important in the presentation on the next chapter. But this is the idea that on some of your heards, a sample of your heards throughout the west that represent a diversity of ecological settings where BLM manages horses, that survey and inventory work should be done almost an annual basis in order to better understand population of horses.

We recognize there’s a budge teary constraint that’s firing that most HMAs only be inventories every two, three, four years, but in order to get the foundational knowledge like population dynamics, at least a sample should be monitored relatively routinely. Probably annually, to provide good data that can then feed into population management decisions and models that I’ll talk about.

We also recommend improved recordkeeping and development of a standardized and comprehensive database. And that all of this, inventory procedures. And data be made readily available to the public. We heard from a lot of public constituents that they don’t trust the numbers, don’t know where they come from and we think that that causes a lot of mistrust between many public groups that are concerned horse and burro management and the agency that’s responsible for the management.

Dealing with the second subject, population growth rates, our work primarily was limited to looking into literature review because there hasn’t been a lot of data available to estimate growth rates. So we looked at the papers that were published in the literature to estimate growth rates. We did conduct one novel analysis that provides additional insight and we looked at the age structure of about 168,000 horses that had been removed from public lands in the west. To try to get additional insight.

Essentially what we did is for each year, those horses are all age removed, that’s a lot of horses each year. So what we could develop is a young of the year to adult ratio is. And this is what those data look like. So this is sort of a moving window average of the young of the year versus an adult ratio for the horses removed from the range each year.

And you can see that generally between 20 and 25 young of the year say per hundred hundred adult horses so this is an index of population growth rate using those age data. We know that would be biased to growth rates because horses are moved to the range in order for t for it to be indicative of the actual growth rate, all the horses would have to be removed before the birth pulse. You had all the an halls that were going to die for the rest of the year up to the next birth pulse and I’ded y’allly that’s when you’d use these ratios since they’re throughout many more months, it’s providing a bit of an overestimate of population growth rate. But you can see where those numbers lie. 0 so our conclusion is while growth rates certainly virginiay from one herd to the next and within a heard — vary from one herd to the next both the population and age structure data from the horses removed is consistent with the idea that typical growth rates are probably in order of 15-20% annually. What’s that mean?

This is a graphic, a table, BLM estimates that there are 3,000 horses on public range lands in the west this year. If we want to project the number, say horses aren’t managed based on the growth rate you multiply that 33,000 by 1.2. And these are the numbers you get from that. So 20% growth rate would lead to a population doubling every four years if they weren’t actively managed and contribute willing in six years.

If you look at the same data for 15% population growth rate, if left unmanaged, horses would do you believe every five years. And triple — do you believe every five years and triple every 8 years until they became food and water limited like we heard about from Mike in a previous presentation. But that growth would probably go for quite a while when we start seeing that food and water limitation throughout the west.

So what’s that mean, these sort of numbers mean as far as BLM’s real dilemma and that is trying to manage the annual increment of horses so the population gets stabilized. So, if you multiply, this would be the annual increments for that 33,000 horses on the range that would accrue over one year, this current population if the populations were growing at 15 and 20% annually. You can see 6600 horses would be added to the population I’ll point out that for 10 years BLM has been removing an average of 8700 horses a year from the western range lands. Considerably higher than that annual increment at 20%.

And the national statistics would suggest that the population over that 10 years is approximately about the same. If the population is growing at 30% the BLM could remove an average of 8600 horses, I think that’s additional evidence that there’s more horses out there than the reported number. And it also provides evidence that these growth rates are realistic given that the off take of horses and the removal program.

So what would that mean if we looked at that annual increment at the different population levels. This would be the number of horses that those different population leveled. You’d have to remove from the range land just to keep the population in any one of those years stable. We know we can only adopt anywhere between 2 to 4,000 maximum. That’s the problem. And so you can see that there’s two things that are going to affect the annual increment that has to be removed. The annual population growth rate and the number of horses you have on the range.

So, if the ultimate goal would be horse management in the west is to only have to remove the horses that you can readily adopt 0 so you can get rid of long-term and even short-term folding facilities, you have to get the annual increment down to between 2 and 4,000 horses and there’s only two mechanisms to do that. And that would be to reduce the population growth rate which we have the NRC committee recommended at least three or four different fertility intervention technologies that could be used and/or you’d have to limit the number of horses on the range so that base population that that growth rate is acting on can meet those management objectives of 204,000 horses to be adopted. — 2-to 4,000 horses.

So in summary, we think the horse inventory procedures are not scientifically rigorous. That improvements to those inventory procedures have been initiated but we don’t know the extent to which actually been manipulated. Our implemented, excuse me. And whether or not that’s range wide throughout the west or not.

We definitely think that recordkeeping and database management has to be substantially improved. There’s no clear linkage between the national statistics and field offices or at least it wasn’t demonstrated to the committee. I’m sure there is a linkage but we don’t know what that is or we we couldn’t cover that and horse populations are growing at 15-20% annually. And with that, entertain questions if we have time.

>> Thank you very much for that presentation. Do we have any questions from the board. (off mic).

>> Have you done — in looking at nonlethal methods of slowing the growth rate, have — with what — what you’ve learned so far, do you feel that there is an opportunity to zero the growth rate based on the current population sizes?

>> ROBERT GARROTT: So the question — I don’t think the mic was working the question is do we think there’s a possibility to stabilize the populations using only –

>> Correct with the current based population?

>> ROBERT GARROTT: That wasn’t part of the charge of the NRC committee. I can tell you that there are a fair number of papers that were population models using horse data have been built and then what might be considered realistic fertility interventions with the tools available today and what might be viewed adds realistic treatment levels have been applied and in general, those modeling experiments would suggest that fertility control can help rethe growth rate but it will probably be very — reduce the growth rate but it will probably be difficult to stabilize the population useility control alone.

The fertility control is dependent on the number of the horses, the population of the population that can be treated. You can’t tread them all or if they do, it can be very expensive or difficult. And so it can help the problem but fertility control at least in the current forms we have probably are not going to — is not going to be able to essentially stabilize the population at whatever level.

So it can help. It can help substantially.

>> One of the tools that can be used.

>> ROBERT GARROTT: One of the tools, yes.

>> Thank you, again, appreciate that information and well-presented. I’ve got a couple of questions. So back to the estimating of the population size. And where you weren’t, I have my note here, did you question the local offices. So my understanding is you had to go to the local offices to get the data that you did for each one of those 40, is that correct?

>> ROBERT GARROTT: Not quite. The NRC communicated only with the folks at the national office. So the national — when we requested the data from the national office, the larger data request, the national office asked us if we would make a more modest request. We made a request to the national office. And I think they made a request to the field offices and then the data came through back to us through the national office. We didn’t have communications with the field offices directly.

>> All right. Thank you for that clarification. So off you’ve not had the opportunity to have the conversations with the local staff at all in any of the BLM offices?

>> ROBERT GARROTT: Not as part of the NRC committee.

>> Okay. .

>> On the new ways of counting that have been put in the policy book, there’s two, I can’t remember what they’re called. The handbook, yes. My question is are you familiar or have you seen them be what I guess I would refer to as ground troops or are they just models? I guess what I mean by that is for instance in the little book cliff because we do know how many horse there by name and picture. Have we flown that and done this process to verify that we are accurate?

>> That was one of the areas that USGS worked in. That they did do one of those survey methods which is called more creek capture where essentially that — every animal’s ID’d by his particular colors and patterns. They’re all individually marked. They’re intensively surveyed and known you actually very seldom do actually know the truth. So they use this herd because they did know the truth and then they applied a March creek capture technique where they didn’t use the known identities of those horses but you fly once and you photograph bands of horses and you essentially mark those bands you saw by the coat colors of that aggregation of horses.

Then you go out and do a second flight and now you’re considering those groups being marked groups, you know them. You saw them in the first light. And you go out a second time and you fly and you see unique groups that you saw the second time that you didn’t see the first time. You see the proportion of the marked groups that you saw in the first slide. And that can — stat thickly adjusted — statistically adjusted to allow you to estimate for the proportion missed.

So essentially you have your first flight identifies animals you know are out there. Groups you know are out there. Second time you don’t see them all and you see new ones so that provides you a way of estimating the proportion missed where you don’t have to know all the horses from the ground so you can do this on a — on a herd where you don’t have those individual IDs for all the animals. It does require, though, it’s hard to believe that this technique would be used for population these horses. It’s a methodology used for relatively small eyes isolated population where it’s realistic to do that and it take quite a bit of manpower to go through those photographs and identify them all. So what USGS did is identified a suite of techniques and they evaluate several more that didn’t work out very well.

And it would be that you’re not going to use — you probably wouldn’t be using this same methodology for every herd management unit. You have a suite of scientifically rigorous methods that were matched to the ecological conditions and the survey conditions on the various ranges.

>> Thank you.

>> DR. BOYD SPRATLING: Dr. Bray.

>> Dr. Bray: Dr. Garrott? Robert Bray. Clearly you and your colleagues on the committee have provided a comprehensive review of the literature so thank you very much for that. You reference that the field data was — could not be linked to the national statistics that were imported. Was there any pattern in those differences, less, more, highly variable. Any numbers to say they were consistently X percentage? Differences.

>> The only thing we’d have there is the population estimates reported for the 40 — the sample 400 HMAs that we were given and compare that against what was on the Web site for those herd management units. And for those where we had the field population — field office population estimates and we could compare with what was on the national Web site there was no consistency. Less or more?

>> I would also offer that of those 40 HMAs, I think something — I have to look back at the report. I can’t quite remember the figure but I think we only had about 50-60% of those population estimates filled in. From the field offices. But there were population estimates filled in in all the national statistic.

>> Was there a subsequent requests when you did not receive one the first time? Did you have a delivery request?

>> ROBERT GARROTT: There was about six months of communications back and forth to try to understand.

>> Dr. Bray: Is there any reason they were ignored or not responded to.

>> ROBERT GARROTT: They were always responded to. We always gout responses and I think that the records of the committee would have all the email responses went back and forth about that except for a couple conference telephone calls that weren’t recorded.

>> And finally with the 40 HMAs that you requested numbers from were they randomly selected or were they identified by the national law office as to what was going to be provided.

>> ROBERT GARROTT: No, the national office didn’t select that. The committee selected that.

>> Was that random or how did you go about making those 40?

>> It was a systematic examine and let me explain that a minute. We took the most recent population estimate for each HHMA, we ordered based on population size, since we could only get 40, we decided we didn’t want to burn up our sample of 40 by getting records for an HMA where there’s only 12 horses or 30 horses, we wanted to have something that represented both a range of horse sizes but but to get the best information we could from those 40. The present population had to be at least 40 or 50 animals. It couldn’t be a mix of burros and horses because we wouldn’t know how to split up that number between those two. So it was only horse only HMAs and then when we had that listing, then we took every third. So — when that list — we got a systematic sample across the range of horse sizes. That took us up to about 336 and then we added 4 in the 80s we he had population data that weren’t on that list that would give us more data that could reflect population growth rates to bring it up to 40 and that’s how it?

>> ROBERT BRAY: And one finally — my voice normally carries so I don’t worry about a microphone.
One final question, when you look at that pattern of differences between field data versus national numbers, can you give me a sense of high and low and how they varied? They were off by 4% or –

>> ROBERT GARROTT: All I can say is sometimes they were right on the money. It was exact same we got in the field office. Sometimes they might minor differences on the order of physician horses but it wasn’t unfrequent, infrequent to have differences of hundreds of horses. And the proportion would depend on the herd. We looked for patterns. We looked for patterns and could not find consistent patterns. So I think the NRC committee report we said we think that the national estimates are based on probably many hundreds of somewhat subjective independent judgments because there were certainly judgments being made at the field offices when they reported population estimates. And there must have been judgments being made at the national office as well after they got the field data.

>> ROBERT BRAY: I probably know the answer to this. But 15-20% foaling rate, do you think that’s a real number?

>> That wouldn’t be the foaling rate, it would be population growth rate.

>> ROBERT BRAY: I’m sorry, do you think that’s a real number.

>> ROBERT GARROTT: I’m not sure what you mean by that.

>> ROBERT BRAY: Do you think it could be substantially higher?

>> ROBERT GARROTT: Probably not. It’s plausible based on the biology of the animals, but probably not much higher. Some of the literature has reported population growth rates up to 28%. And the way that’s.

>> DON: I think in all of those records it’s where you have a good sequence of population counts. Say, 8, 10 counts in consecutive years and you can look at the trend of those counts and estimated a growth rate from that and that particular methodology has generated some estimates of population growth that exceeds 20%, up to about 28% there’s confidence limits on those and that’s when those higher estimates would cover 20%.

>> That model wouldn’t be based on accurate counts of animals.

>> It’s .
>> You how do you define an accurate count if you don’t know truth.

>> You could have some variability in what proportion you count every year but, if you had consistent methodology, and you counted those populations over over a number of years, you could statistically estimate the growth rate even if you don’t know the proportion you counted. And you get a legitimate and scientifically rigorous estimates of growth rate without knowing truth. How many horses are out there or any wildlife in our populations other than those few that are tracked by individual animals and they’re all named and you’ve got 20 people growing up out there that loved him and watch them and keep an eye op them for you?

>> ROBERT BRAY: Thanks again.

>> ROBERT GARROTT: Yep.

>> I’ve got a couple questions on the growth rate, do you have a percentage of it’s interesting, I spent my entire career working, one of the places I work is Yellowstone National Park and there’s more science on the demography of one out population in Yellowstone National Park than this entire species on the western range lands and indeed there’s been very little science done and most of it was done in the late 70s and 80s on population demography, growth rates, foaling rates, serious signs to understand horse vital rates, survival reproduction age first reproduction pregnancy rates.

And so there isn’t much science there and that’s the idea of sent until populations could help us get a little bit more of that. But certainly pregnancy rates at least 50-60% based on — 50-60% pregnancy rates based again on research done way back in the 70s and 80s, where a lot of horses were bred and certainly levels were assayed for horses being removed from the range land and part of a big research project that I was part of there in the 80s. So the foaling rates, the pregnancy rates as much as we can tell in the foaling rates would certainly support the idea that populations could routinely be growing at 20%. (off mic).

>> That would be more valuable than the actual growth rate of the herd I think as far as knowing population growth expression. Am I incorrect in thinking that way?

>> ROBERT GARROTT: So you don’t know which animals are pregnant or which animals are fertile most likely when you’d be treated them. So you’re going to — in a practical standpoint, you’d be treating at least Mary oriented contraception. You could be — wouldn’t be able to carefully target which animals were probably treated. In order to do any effective management with contraception and I’ll touch on that in the next presentation, you certainly have to have — or I would think you’d want to have some model of population dynamics for the horse herd that includes pregnancy rates. As well as foal survival.

>> TIM HARVEY: That’s where I’m going, I’m trying to figure out how many Maries you would optimally treat to get that rate you’re trying to get earlier to achieve the number of growth rate you were talking about that would be sustainable.

>> It would depend on the demography of the herd. So even though our conclusion is that typically herds in the west are growing on average of 15-20%. There are probably herds that are growing much less than that.

>> TIM HARVEY: I’m grasping for is there a number we need to treat 15% of the mares, 40%. The number that gets treated is quite low. So what I’m trying to get my mind around is mares need to be treated 0 on average. Can you comment or can you throw a number out there how many would have to be gathered and treated in order to affect some of the changes that we’re looking to do?

>> ROBERT GARROTT: So effect some — so — there have been a number of population modeling studies looking at just that question. If your target is the reduced growth rates, and the mare that you stop from having a foal for a year or two is helping to contribute to that. If you’re looking for a proportional decrease, it depends on what your state agoal is. If it’s to reduce the growth rate by athe least half, I think all the studies that have been published thus far with limited demographic data on horses it would suggest that at a minimum for mare oriented couldn’t Secretaries only you’d have to be treat — couldn’t sent only you’d have to be treating 20% of the mare.

Chemical vasectomy could be used, no one has incorporated any modeling that experiments where both mare and stallion contraceptive tools were used in combination. The bottom line is that to have a noticeable and a measurable effect that you could measure with good population inventory techniques, you’d have to be treating at least 30-50% of the annuals, of the animals and you’d have to repeatedly do that at least for the mare oriented an animals that were available.

>> Can I follow up on that, you’re saying you’d need to treat them. Is that accounting for the fact that only a certain percentage are going to be effective in contraception. Are you saying that 30-50% have to be couldn’t sented or treated?

>> ROBERT GARROTT: We’re getting way beyond the NRC report. The NRC didn’t do any modeling of that. So I’m speaking partly from the modeling that’s been done in the past of which I was part of when I was involved in the wild horse research in the past. I think that the NRC committee did strongly recommend that fertility control, there’s been 30 years of research on fertility control specifically targeted horses. That the problem of excess horses as defined by appropriate management levels that BLM has now is a real issue I think you can see by the end of that presentation that reducing growth rates can contribute to helping BLM solve the problem.

How it can contribute, how much it can provide, this would be a management experiment because nobody knows. But we do know there are multiple pools that can inhibits reproduction effective with side effects in the context of behavior in ways that were viewed to be acceptable. That could be applied. You about it’s not something that we’d recommend that BLM be able to say that for the next 10 years, we’re going to increment the number of horses treated with TZP or something by a thousand. It needs to be done if it was going to be credible, it needs to be done as science and as an experiment because it’s very uncertain how much of an effect it can have. But we’re very comfortable with the fact that it could have an effect. It could have a measurable effect.

>> Did you have a follow-up question?

>> TIM HARVEY: I’m just thinking numbers. I’m a numbers guy and I’m just listening to within 5-20% growth rate wondering what we have to do to get that growth rate down to something that can be handled by the adoption program. So looking at the growth rate and how does that shake out from foals in the ground and natural mortality westbound the herd? And you know, because that’s going to knock that number down a little bit. I think that again we’re outside the committee. Commit he it’s work.

But there was — some work done outside of the committee. And it’s a pretty easy math exercise if you look at that last table. If you applied if you aenough horses on the range for horses, I think the upper level for appropriate management is something like 23 or 24,000 horses on the range land. If you had that many horses on the range land and you reduce the growth rate by 50%, so dropping it down to 710%, you’d have annual — 7-10%, you’d have annual inyes 3,000 horses that could be adopted equivalent to adoption demand.

So there is the potential that in the long run, effective application of contraception with a base population of what is your appropriate management level now, beyond appropriate management level could eliminate the need for removals beyond what the adoption demand can fake. I’m saying that’s not going to happen in a year (take.) we’re looking at — you’d be looking at a fair amount of time to ratchet up to a nationwide program like that. And a fair amount of science to understand how to do that. But it’s certainly within reason that that could happen.

>> TIM HARVEY: I think there needs to be activity going on that’s what we’re looking at as advisory board. We have to address this issue from a couple perspectives. One is crisis management situation that is looming over us, the elephant in the room. And then you’ve got the long-term. So in trying to come up with a process and a plan that allows you to basically address the problem from more than one perspective or angle, it seems to be the way to do it. The questions are geared at long-term management. One is not exclusive of the other.

>> ROBERT GARROTT: The crises you have now with the budget, BLM isn’t going to remove as many horses this year from the range and they typically removed for the last 10 years which means that next year the annual increment will be bigger than it was this year because you didn’t remove as many horses as you did last year and the year before. And so the longer the budget crisis that BLM faces now that’s going to curtail the amount of money that can be applied to active management of horses on the range applied to horses that need to be maintained in long-term holding facilities that is a real problem that you have right now.

Because the more money that has to go into that, the long-term holding facilities, less money you have to manage, if the budget stays the same. You have to maintain the horses in captivity. The only place it seems you have left to go is not to actively manage as many horses on the range, which means that there will be more out there and larger increments.

>> Page 22.

>> Yes,?

>> TIM HARVEY: One of the things I find — I don’t know the right word for it. I’m looking at perceptions BLM is saying there’s more horses out there. There’s certain citizens groups that main there are nowhere near as many, many out there and have you guys come along with your study and this really is fond that there are a lot more horses out there and appear to be more horses out there than a lot of citizens think.

And I think from an outside agency coming in with that information, I’m hoping will give some credibility to where the BLM stance has been on how to manage these horses in this crisis situation. I firmly believe the facts that you guys have presented. And I’m hoping that maybe the reason I’m kind of asking you to reiterate some of this reaffirm the validity of the percentages in the growth rates and stuff is so that maybe some of the citizens groups that have been fighting the BLM on some of these actions can maybe participate in the process to help instead of fight with them so much over it. And that there’s — the fact that an outside agency has come in and verified that these populations are growing at this exponential rate is really going to create a really, really poor situation in a short period of time.

>> ROBERT GARROTT: The NRC committee certainly does not support the idea that removal program that BLM has had ongoing for the last 30 years is managing the population to extinction. And indeed, 195,000 horses have been removed from the range lands thus far, at least according to the records that the committee receives. And one can do the math on those removal rates and there has to be a substantial base population on the range in order to sustain 195,000 horses removed over that period of time.

And the last 10 years on average 8700 horses have been removed from the range consistently over the last 10 years. And you saw when I added 33,000 multiplied by 1.2 at 20% annual growth rate, that only comes up with the 6600 animals. The disjunct there is we’re not sure how the national statistics go with that.

>> When our facilitator starts to dance around over here, I know our time is limited.

>> That’s extremely important.

>> It is important.

>> I really wanted to understand –

>> Unfortunately Dr. Garrott is going to give us the next presentation so we can carry over questions in the next segment. Joan?

>> JOAN GUILFOYLE: Thanks for saying you agree we’re not managing these animals to extinction. I hope that is a myth that will now die. I appreciate that. I had a couple questions about what you said Bob we’ve been frustrated. Now that I see the best example and the worst example, I think I understand now why it’s been difficult to get a grasp of this. So that was very illustrative. 179 HMAs, different methods, different timing, I can see it.

I appreciate that you recognize that we have made improvements. A handbook, you’re looking at complex as USGS methods that we’ve done training on twice with field people. When we read this chapter, we said absolutely yes and we have been able to put some money aside. I want the board to know, I’ll update you on some of these things as we go along. We’re able to get money into the USGS agreement to help us and be in charge of the design for our field folks on using these as a side or simultaneous double couldn’t methods if those are the most appropriate and depending on the HMA, we absolutely agree. We have to know the number that we’re talking about so we can do the rest of the management so thank you for that. And I’m — I always say this when we get input from external folks that it really does help us — helps me manage and improve the program and so this was — this was — this is an on one to us. — onius one to us.

We met with Dave Powell on the committee that Kara set up last Thursday and he did a presentation for us on all the chapters and to the chapter you’re referring to, the population survey estimation one, one your recommendations is that we have’ centralized database and we agree with that and would love to be able to do that.

The interesting point he also made that in a lot of the public testimony that the committee got was really that the public, the people don’t understand that data that we’re putting out there. It isn’t so much that we’re trying to confuse anybody or hide anything. But they do not understand it. Which is our job to do a better job with that.

I think not understanding and not believing, I appreciate that this report and U.S. GS’s methods will enable us to convey what the actual facts are and that they won’t be disputed and they will be out there and clear for everyone to understand. So I wanted to say thank you. And Boyd, I don’t know if — the BLM as you know, we’ve been looking at this report quite heavily too and we do have a person in charge of looking at this chapter and I just wanted to inite that person who I think is Dean, if you had any questions and comments and he’s saying no. So so that’s all that I wanted to just do, Boyd, thank you.

>> DR. BOYD SPRATLING: Dr., I’d like you to go ahead and proceed on the next presentation of population models and evaluation of models.

>> As he gets a glass of water and rests his throat. I want to let you know, in order for everybody to hear you well, these microphones require that you do have them reasonably close. But also directionally, they almost have to be facing your mouth. (off mi>> Like that when you moved your microphone away.

>> ROBERT GARROTT: Chapter 6. NRC dealt with population models, statement and tasks to evaluate. Population models are models in general. And the first thing I’d like to say are some people think models are esoteric but they’re actually extremely useful tools for any manager because you can understand they can understand and help explain and predict a dynamics of populations and we’re all in the business of managing populations putting treatments on populations and some expectations of where they’re going to go afterwards.

Population models are very useful to any wildlife manager. In particular on this case where we’re doing active management and maybe thinking about even some new and more aggressive management with new tools is allowing population models allow you to manage various treatment options that you might have and predict the consequences from alternative management action. It’s useful in any agency.

So objectives for the chapter was to provide a brief description for models that had been developed for wild horses over the years, in particular we were asked to evaluate the win equis model which si amodel the BLM contracted to be used for management of wild horses in the west and also comment about alternative models that may also be useful. So that’s what we’ll cover here in this. WinEquus.

Very briefly, the first population models for wild horses in the west were probably developed in the late ’80s and early 1990s. I was part of the group of scientists doing that throughout the west when the same issue was really hot in the 80s when there was really aggressive removals going and excess of animals with no place to put them. Soon after basic models were developed and that was trying to put together enough survival and fecundity data to understand how these — how these populations operated.

Then once we had that, then we could start manipulating either survival or fecundity in models to look at various — how those interventions would affect population size and growth rate so that really got started by the mid 1990s. And that’s when the WinEquus model was developed as well. And then going down a little further. 2,000. Mike alouded to ecosystem models where you’re modeling everything from the climate, its effects on the forage and plant base and all the herbivores, not just horses that might be off taking that forage and trying to understand essentially how horses fit into all these ecosystem processes and Mike was involved in that and 2000.

So that’s a brief history of types of models and when they were developed for horses. We believe we spent a lot of time looking at the WinEquus model because this is the model that was built to inform routine BLM management of wild horses. It’s individual based model which means the model is keeping track of every individual animal in the population whether it’s male or female. Whether it’s 2-year-old or 5-year-old or 25-year-old. So it’s age and sex structured individually-based model. It provides output for up to 20-year prediction on the population given whatever you put in to the model. And it’s used by all that — am all the HMA planning that involves management interventions.

So you see this model in all the intervention documents. The EAs and the gather plans.
The real strength of the model is it’s easier to use with minimal train. It’s flexible and you can change lots of input parameters and basic mechanisms influencing population like density dependence and how variable the climate is affecting vegetation. It effectively simulates management scenarios which at that time was female fertility control and removal or a combination of those two things. It provides informative output and it’s very well documented.

All models require data. What has to go into the model are some sort of initial age and sex distributions to the population being monitored. Age specific foaling rates and if desired, the user of that model can turn on parameter values that will implement density depence which Mike talked about earlier as well as environmental casteicity, that’s variation of the amount of forage that might be available based on climate. And of course interventions like removal and female contraception.

The council reviewed HMA gather plans and what we saw was that in general the WinEquus model was used to assimilate alternative management actions for no removals to removals only. Maybe it’s several different levels. And perhaps contraceptive treatments as well.

The plans would provide basic model output in those plans, those gather plans, management plans. Often they were account and pasted from the computer output as an appendix and very often with no interpretation at all. So the output of the models is an appendix at the back of the report but no interpretation of the model. Based on both the reportses and interactions with BLM, representatives, the committee really couldn’t determine if the use of the WinEquus model actually informed the management decisions.

Whether it was used to justify management decisions independent of the model results. Or whether it was simply a boilerplate requirement of management plans. In other words, planning management plan was written, it was required that everyone put some WinEquus model output in the plan and so it was run and put in the appendix. We just tell.
— we just couldn’t tell. It’s probably or it could be a little bit of that each depending on which herd management plan you’re looking at. Just couldn’t tell. Not a weakness but something to consider with the WinEquus model is there are many decisions and assumptions in setting up that model. So somebody has to sit in front of a computer and actually make all those decisions and put those into the — before they can run the model. Those decisions actually dictate the performance and output of the model.

But WinEquus also has the ability for the user not to set any of those things. It just uses default data sets, default parameters. And so, if you choose not to do anything, you can open up the model and run it and not set anything because our default parameters that allow it to be run. The management plans typically didn’t provide any of the information about how the model was set up for the run. So you get a simulation that was put in the management plan but all the information on how the model was set up, whether or not density dependent was added to it, environmental stoichasticity, what age specific survival and fecundity rates were being used as demographic of herd being modeled. So almost universally, not quite, like we looked at one management plan that did a very good job of telling us exactly how the model was set up before the run.

But there was no information about how the model was set up for those particular runs. So without that information, a critical user can’t really evaluate how well that model was mimicking that population and critically evaluate the output. Alternative models, we were asked to look at alternative models. The WinEquus model is a model built to emulate a population of horses and you’re managing primarily on the population or HMA level. So looking into the future, so planning what kind of a population management BLM might use. In the future a little bit beyond WinEquus, in other words, things that could be done to improve a population model for the future would be to have survival and fecundity and age structure data that better matches the target population.

WinEquus has three default data sets for these things that all come from three different herds that were researched in the ’80s and whether or not any of those default parameter sets are even legitimate for those herds that they were generated from 20 or 30 years later. Might be a little bit questionable. There’s a future to better match the populations you’re going to be modeling and one thing that can be done is use herd specific age and data from gathers and removals.

Often times those are substantial removals and they aren’t selective. There’s information on the age and sex structure from previous removals of the population that you’re actually model for. They could be used rather than the default data set. This brings back the idea again of using demographic data from closely matched sentinel populations. So, if you improve inventory techniques and then you also identify a suite of populations that you apply those inventory techniques to routinely in extremely arid environments, mountainous snow environment, across a sample of populations that represent the difference with ecological settings that horses are found in throughout the west, then you could at least say that the herd I’m going to manage comes from a very arid desert environment and we have sentinel population that’s provide fecundity and survival data that are similar to that, parameterize and model with that, it doesn’t come from your herd but it’s ecologically similar.

So you have a suite of default data sets so you match the default data set that you’re going to use with the ecological conditions of the population that you’re modeling. It also might be important have the capability to model both mile and female contraceptive techniques which the WinEquus couldn’t do. If you can’t get the population growth rate to meet your management objectives with just male or female contraceptive techniques, it may be a combination of the two can do a better job.

Right now we have no models that can apply both removals, mare oriented contraception and male oriented contraception and it may be that that could be a useful addition in the future. If those sorts of interventions are going to be considered. The other thing we learned since WinEquus has been built there’s been good studies where horses have been manage the pretty heavily with fertility control. These are primarily the shackle Ford banks and as teeing island situations and from those studies, as Mike alluded to, when you shut down reproduction in horses and don’t have that additional energetic costs, they’re healthier, they live longer.

So there’s a feedback there that we know enough about or start to know enough about to be incorporated in population models future. Also in those studies, there’s an indication that repeatedly for the PZP vaccines that if you reedily treat mares with PZP, when you withdraw that treatment, more times animal has been treated with PZP, the longer it is for her to return fertility to the point with enough treatments they may indeed be sterile. These sorts of demographic feedback provide a means of trying to develop models that are more realistic for what we’ve learned in the science thus far.

Another type of model that might be useful is a comprehensive model of the wild horse and burro program. So right now we’re talking abouted month eels models just for a specific herd unit. We know there’s 172 of those out there. So that would all relate to the free ranging herd populations. BLM manages more horses in captivity than they do on the range. So entire program includes the demography of horses in short-term fast its and long-term facilities and movement of horses between those.

That’s your model. Having a more comprehensive program might provide ensites for future especially long range planning budgetary planning and things like that that could be helpful as well. Finally, a different type of model adaptive resource management models, short-term for it. ARMS. I think could be very useful moving forward. This is the idea that BLM managers need to make important decisions about what tools or combinations of tools they want to use to manage horse populations that’s made with incomplete and imperfect information about how the horses will respond to management actions. Adaptive resource management models is a structured way to make decisions scientifically incredible manner where you continue to learn so you reduce the level of uncertainty as you continue on in the management program.

The premise here is there’s a lot of uncertainty and the more we could reduce the uncertainty, the better we could find management. Decisions have to be made in the face of that uncertainty we’ve got to manage even though we don’t know perfectly how many animals are out there, how good the contraceptive treatments are going to be, what sort of feedback might be there and we have to keep making decisions over and over again. So, if we have monitoring in place or it could be implemented, then we can actually learn every time we make one of these management decisions.

Here’s the process. If we do X, Y, and Z, here’s the objective. We want to reduce population growth rate from 20% to 10%. So we have an explicitly stated objective. We say okay, how can we get that done? With are our management alternatives. We could have PZP vaccine at a certain level. We could have chemical vasectomy at a certain level. Combination of those. Combination of removals and both fertility interventions, all the options available. And you make predictions about those options, which those options will best meet your management objectives or predicting will meet them. That’s your population model. So you manipulate the population model based on interventions and make a prediction. That model was your best knowledge. That’s your best guess, you implement it and follow up and monitor and see if you got what your model told you, whether or not did meet your objective?

Were the predictions met? If they aren’t, then you either chose the wrong management alternative or your model is not quite right yet. So you get chance then to go back and say well, I should have chosen another objective you need to change your model because there’s other feedbacks we don’t know. Next time you go back and make decision you’ve reduced uncertainty and you’ve improved your ability to manage over time.

This arc RMS model is being used to learn as you go because managers are experimenters. They’re researchers that learn how to do things better if you do it in a structured way and we think this could benefit BLM in the and the wild horse and burro program and models themselves if a model like this was implemented. So in summary, we think models are essential tools for management, that the WinEquus model is scientifically sound but its application for informing management has been poor or at least as much as we can understand from how it’s being used with the documents.

We think it’s implementation to help inform management has not been what it could be. We think substantial improvements could be made for planning future population management or improving on models existing. And we think models of free ranging and BLM could be useful especially in the context of the budgetary constraints you have. It costs a lot of money to manage the captive program, understanding the dynamics of that — those captive horses and your options there and how that influences how much money you have for free-ranging horse manage the could be useful.

and we think implementing adaptive resource management models could strengthen the scientific credibility of the program going forward. With that if I have questions, I’ll take them.

>> DR. BOYD SPRATLING: Questions. John.

>> I’d like to follow up on the adaptive management plan you’ve presented there because as you know, we’re hogtied on a lot of avenues to control these horses. My question is of all the practices that you presented which of these do you think would be most likely to be able to be presented and get on the ground so they could actually control the number of foals being produced.

>> That’s completely based on what’s likely depends on the political actions taken by all the people that care about horses. I think that any combination of those three fundamental management actions you have, removals, female oriented fertility control or male oriented fertility control, all of them can help. They all have costs: They all have proponents and opponents.

So the policy on what could be used effectively in the political arena I can’t — I can’t tell you. But I can tell you that no matter what you do as — no matter what the agency does, if there’s no assessment of how well that is done then it’s difficult to sell it to the public over time. So going out and saying well, next year we’re going to treat 500 horses with some contraception, if there’s no explicit objective, and no follow-up to tell how well you did, you simply will have a eroded public support for continuing along any track that you don’t follow up with reasonable monitoring and something like this adaptive resource management model.

Which tools you use, that say political decision and is right now an economic decision as well. None of them are cheap. If you believe that horses have to be active managed, that the public will not accept self-limitation as we heard about the consequences, that they have to be actively managed, all the active management tools you have are expensive. They’re all invasive. They’re all going to require capturing and haled handling a lot of horses which a lot of people don’t like. And so there’s going to be political obstacles to any of those tools or combination of those tools and I don’t know how you get there.

>> From your perspective, do you think you got any choice?

>> ROBERT GARROTT: Well, you could just simply stop managing horses and they would self-limit. And I think you can look and see what’s going on in Australia right now with 400,000 horses and catastrophic mortality happening because of the drought and there’s no reason to think that that wouldn’t happen here. I think — well, it would violate every mandate for responsible management of public uses.

So I think you have to manage horses, yes. Most of us agree that self-limit is not an option. Then we have to go back under today’s parameters we’re operating under. To get a handle on foal crops. So, if that’s what we’re looking at, then we have to go back to what your suggesting, I think you’re on the right track, getting it implemented is my biggest concern.

>> ROBERT GARROTT: There’s only two ways to reduce the annual increment. You reduce the population growth rate and/or base population that’s growing at that rate. So you have reproductive intervention of a suite of tools that can be used to reduce population growth rate and reducing the number of animals, base population, you only have one tool there right now and that’s removal and captivity. That’s expensive and prohibitive because you have no place to put them.

>> John: Alternative is expensive also. Thank you.

>> I have a quick question. I’m looking at the enormity of trying to implement these changes. I’m realizing this is outside injure study area. But it seems to me it would make sense to implement them and decide on a couple different passes. There is no one answer. We’re looking at a quiver of ariose, not one silver bullet. Does it make sense for you — for the BLM to perhaps select several HMAs that we would focus on and implement using several different tools and permutations of some of the things that you folks have come up with? And go head and implement some of these changes and see what the results are over a year or two. Room than trying to broadcast every single horse on a ranch in 179HMAs, as a scientist I would think it would make more sense to approach it individually so you can also see what’s going on. You can then judge the results of what you’re doing. A little bit easier?

>> ROBERT GARROTT: This is again where the agency is in a really hard spot. So, if you there’s been 30 years of couldn’t ceptive research on wild horses and you could pick a couple HMAs in the west and do a several different types of treatments. You could have replicates. And you’ll get good science out of that. You’ll learn from that especially if will follow up well from treatment and modeling. So that’s a dilemma. That’s good science and you can do that at the small scale and be really cautious but you still have the issue you’ve got to do something with your annual increment right now. And so they aren’t going to help you with the national level. They’ll help you get good science and understand what the effects are of these controls.

You know, that will be a 5, 10 year process to try to fine tune the technique. Reproductive inhibition techniques before you go a little more aggressively with the national population and mean while you’re still going to have to take care of the annual increment. So whether or not BLM moved forward with any sort of management level applications of contraceptive treatments are keep it very small scale couple HMAs. That’s a policy decision that needs to be made.

Using sentinel populations to get more frequent and additional data. If you were — wouldn’t you also agree that those would be the places fairly quickly if you decided to start working on some of these sentinel pop Asian las that’s that where we would do adoptive management experiments. More about demographically would be better places for to start and herds that are intractable or like a herd that only gets surveyed every 6 or 8 years. Simply because you don’t know the form answer of that herd to begin with. So the more you know about a herd, the better able you are to tailor your first experiment, your first treatments.

It tends to be that the herds you know most about are small herds and the problem isn’t as much the small herds as it is the really big wide ranging herds in remote areas that are most subjected to drought. But in the ideal world, you know, a lot about the herd before you started so you really would be able to parameterize your models well and have a lot of confidence in them.

Go small herds because it would be easier to know what you’re doing. And ideal world when all your worried about is science and not managing what — not managing the national population. Thanks.

>> The LLC out of Florida that owns the rights to the ARM model. Who are the individuals behind that?

>> ROBERT GARROTT: The adaptive resource –
>> .

>> Yeah who are the individuals who have that LLC?

>> ROBERT GARROTT: Well, I would say that the — the primary people behind adaptive harvest management are the scientists out of paw tux 8, the USGS scientists. We heard a presentation from one of them with Jim Nichols. The committee.

They’ve gone all over the nation and have workshops on adaptive resource management. They’ve built technical models, there’s several books published on that. They’re sort of the experts although it goes back before the paw tuxet people got involved. I think — oh, my, I should know this. I think it was actually generated for fisheries. Ocean fisheries in particular. Can’t remember the name of the guy who wrote.

>> CALLIE HENDRICKSON: Just go back real quick. Discrepancy of numbers were you able to get the dates for the reason you got the numbers, for the Web site for the total number is as of usually, like, February 28 of whatever year. I just was wondering if there was any reason that sort of date might have been part of the reason why you had such different numbers, I don’t know.

>> ROBERT GARROTT: We just had no information. I think that’s plausible that the range manager that did a count in December knows that the population estimates supposed to be for the end of February. And estimates that maybe it was a bad winter and he had a 5% mortality rate between that 3-month period and in good faith adjusted that count by what he thought might have been the population mortality between the count and when the population numbers to be reported. We just simply don’t have it. And to — and to be fair, there wasn’t — when we asked for the numbers, to the national office, and the national office went to the field offices, I assume they didn’t ask for a big long diatribe on the numbers they fill in the table and send it back and we have limited ability for individual field area people to provide all the rationale. It would have been nice to have that.

>> If I could, Boyd, folks in the audience who work for BLM can explain how this data transferred. I would invite them to do that. It seems to be a littlen clear. Is that Zach or Dean or somebody.

>> Dean is who reinteracted with a lot..

>> Can you gentlemen explain how it worked.

>> I’m Zach Ryan hold, the senior wild horse and burro program specialist. The way rereceived the information was exactly that. We received a request to the committee either through Kara. She then relayed that it request to myself. I then relayed that request to the state lead who then spoke directly to the field. The field provided that information. It was in a table. We did speak to the committee and with Bob and explained to him that there has been a number of adjustments at the date and time of reporting for various different reasons over the last 30 years.

And those were primarily due to program decisions. Either it was based on when the budget was falling — the budget cycle ended or began. Whether or not we were trying to adjust it to capture the whole — the whole crop after the full crop trying to get it to be in sync with the public lands records and when those are actually reported. So it has varied over the years when that is reported to the national office. And you know, none of it — it was all in order to try to get it in sync with some sort of program that was occurring at that time. Or for budget reasons. There weren’t any hidden underlying acts to try to deceive the public or anything like that. It was merely we were trying to bet it in sync with one reporting system or another.

>> Dean, did you want to add anything to that or no.

>> Dean: Just a couple clarifications. The national office did not manipulate the data any year since I’ve been involved since 2013. But we get HMA is what they reported to us (off mic) and there’s a lack of information about how they’ve been manipulated the data, increased the estimates since the last survey. But two things it contribute to the inconsistency. There’s not information about removal which creates oh, my gosh, how did this number go from this to that and had that information been available, it might have been more easy to interpret but the main message is here. There’s good findings.

We need to do a better job of recordkeeping and reporting to the public explain what methods we’ve used the findings are sound and good and that’s the direction we desire to go and that we’re headed. So –

>> As I said, we want to — we want to analyze each HMA to see which method is best for that HMA including the two that the USGS recommended. We have to do a better job of that and we are going to. Thank you, Zach and Dean for clarifying some of that.

>> Doctor, I have a couple questions then because modeling tends to have my eyes glaze over and I have a tough time getting around it as a clinician. But I’m going to go back and ask a question about actual counts and methods. And I think we’ve — you mentioned that Mark recited or recaptured dash would not be practical in the west where we have a few thousand horses as opposed to a couple hundred. Account — what model or what attempt at counting would account for human error because I think we have to sympathize with people sitting in that aircraft on any hot bumpy day or whenever they happen to have their flight time be up in the air.

How do we standize the differences in people sit the sitting there doing the count because that I can see would be tremendous differences to human oriented ear and differences being able to spot and see the horses in different locations. PGA or whatever is covering the ground. How do you feel about that? What would your recommendation be in that sense? Since wide life became a science in the late 1930s, one of the primary activities in research is figuring out better ways to inventory animals and statisticians have developed a lot of innovative tools to do that.

So we do know that from other studies that there are many places that horses are counted very effectively especially the wide open sagebrush plains. You see dust trails of those when you start flying, you can find them well and couldn’t them well. There’s no trees. There’s been a good science that tug that you might only miss 10 to 15% of the animals under those conditions. Basin rank country or book list aerial survey you may only count 50%.

So fundamentally without having estimates of detection probability for each of your senses, you could provide your herd management areas based on cover topography, those things that make it more difficult to count animals. In the ballpark when we’re counting the desert, we’re probably not missing more than 20%. When we’re counting some mountainous area that has juniper all over it, we’re more likely to couldn’ting maybe sex 0 to 70% dash counting maybe 60 to 70% at the most and just make an adjustment and it’s just an approximatization based on what we know about how cover and topography affects it in a more rigorous way one of the methods at USGS suggests is double observer survey.

If you’re going to use a plane, you have people independently recording which animal groups they see and when and then you have the observer one independent observer in the plane that’s seen so many groups of animals on the survey. A second observer that independently saw so many groups of animals on that survey. They identify which animals. The first oftener saw and the observer saw. Which animals the first observer saw and the second didn’t. Which the second observer saw and the first didn’t and from that you also know statistically the probability that both observers missed some animals.

And so USGS used one of those methodologies of double triple or even quadruple observers on the plane that recorded data independently and statistically you can get some of those things. There are other mechanisms too.

>> But logistically, you’re sitting in an aircraft, looking out the right window or left depending where you’re sitting in the craft. If you’re going to have that true double or simultaneous double count, you’d have to have people sitting on the same side of the aircraft, I would think. I mean, I can see so many variables sitting there. I can fully understand the difficulty in coming up with legitimate numbers.

>> I would say 1930s, the tools are interest to make adjustments and appropriate Constance limits on those. Right now we have population estimates with no ability to say how precise those are. When you use those tools even though they’re imprecise, you can get scientifically rigorous population estimates that adjust for animals not detected but that also give you confidence intervals that provide — so instead of saying we have 33,000 animals or let’s say 1200 animals in a herd, based on the rigor of our scientific data, that’s our best plan estimate. But the — but given the data, it could be anywhere between 800 to 1600.

>> I understand there are actually tables or data out there that okay, given a certain type of cover, certain type of habitat that you would use a certain percentage of accuracy. Are those tables –

>> Yes, you could just do the ballpark approximation, it would be better than nothing based on some of those attributes. But there are no tables that — to tell you that the ideal thing is to be able to collect data about detection probability at the time we do the survey. You can use the same aircraft, the same observers, different day or in the morning versus the afternoon. Long shadows versus bright sunlight and the proportion missed will be different.

So ideally you’d like to get that information for that particular survey U USGS evaluated some of those techniques, they flip flopped that several of them were practical. Could even reduce the cost of inventorying it because it might be better done with fixed wing and helicopter and if they were scientifically valid and rigorous.

So there’s lots of different ways that one can go and that would be up to BLM to decide if they want to change their inventory techniques orfy them how they’d want to go about that.

>> Hello. Boyd, I think you — I just want to be clear the USGS. Excuse me at Fort Collins simultaneous double count and mark recite and there are several people in the audience who have been trained on that and have done them, I’m sure, someone back there is going to give us a little clarification who has actually done them on horses. Is that someone out of am I eyesight.

>> It was said not in these words. Simultaneous double count part of what gives you confident limits excludes excitability of animals and also things like the experience of the observers and the position in the air crave craft. So a lot of those variables are factored into what gives you confidence limit and that’s why you have a limit of low confidence to high confidence that includes variables like snow cover, sightbility, bias, based on environmental positions, but position in the aircraft and experience that of the observer can be factored in.

>> So there parameters that are set.

>> That’s how you arrived at your statistical estimate of your 90% –

>> Thank you.

>> I’m sorry, this is Dr. Al cane. AFIS veterinarian assigned to work with the wild horse and burro program.

>> Hi, in the NAS report there was some space dedicated to infrared technology. I’m curious in the areas where environmental issues, tree cover and stuff like that, I know in the am ill Terry applications they use it in the cooler part of night to get body counts and who’s out in areas. It seems to me that might be a very effective cost effective way. I know that there’s restrictions on the drone flying because of aver air time allowances.

But it seems to me with the unmanned aircraft using infrared technology flying at night so that the images quality is way better at night than it is during the day especially at hot regions, is that something that would give you more accurate counting or is there problems with the horse thing that I would be unaware of using it for horses?

>> Well, the USGS team looked at that. And they didn’t think at least at this time that it was practical. Partly because that had to be done on contract and cost of forward-looking infrared system mounted on the plane that they had to contact would be considerably higher than what’s being paid right now. The other thing about infrared that was an issue is whether or not you can get an image of a higher resolution that you can tell different species of large body mammals. Obviously the lower you fly and the finer the spatial resolution you can get outline to define species but that means more intensive flying more time in the air.

This time you’d have to look at the report I think they said they didn’t think it was practical primarily because it cost so much to contract all those remote techniques had the potential down the road to be used that they still had the same issues of detection. So you still have the issue of having to do the research to figure out what the detection probabilities are and how they vary over the different types of situations that you might encounter and that very different terrains that you have. So those are all possibilities but USGS scientists, I think did not think that they were effective at least now. That surprises me.

>> DR. BOYD SPRATLING: We’ve conditioned over our agendized time.

>> No way.
>> No way!

>> Is there time for the person would led this chapter review to just see if he has any questions before we –

>> DR. BOYD SPRATLING: Okay. Yes.

>> Roger, okay I’m getting a no. Anybody else from BLM.

>> Okay, never mind.

>> DR. BOYD SPRATLING: So tomorrow we invite you to be back. Does the public have access to the agenda? You’ll see what we have coming up tomorrow at 8:00 a.m. he vine it you to return.

>> Did I hear correctly that we might ask Kara to say a few words to start off tomorrow morning.

>> Okay. Okay. And just a final reminder we will close tomorrow 3:00 to 5:00 for two hours of public comment. Sign up and we’ll allocate the time based on the number of people, mike, Bob, thank you so much. Mike is back with us tomorrow. We very much appreciate it.

>> Thank you, Dr. Garrott we appreciate your time and Dr. Cuff Coughenour.
(Applause.)
With that we’ll adjourn until tomorrow morning.

 

Watch the live-stream  here