Incorporating Human Dimension Objectives into Habitat Planning and Delivery
The incorporation of human dimension objectives within the North American Waterfowl Management Plan (NAWMP) has compelled waterfowl and wetland managers to consider whether and to what extent landscape characteristics such as public land access, the type, amount and location of wetlands, and site infrastructure will increase support for wetland conservation among user groups. We demonstrate how model outputs can be used as an objective metric to evaluate the benefits of alternative habitat acquisition and restoration projects. These data and methods show promise for incorporating human dimension objectives into habitat delivery and understanding potential tradeoffs relative to biological objectives.
icon search Searchable Transcript
Toggle between list and paragraph view.
[00:00:01.217]Well, than you everybody for having me out.
[00:00:03.994]Before I get started, I'd like to really thank
[00:00:07.505]Mark and Chris for inviting me to participate
[00:00:09.704]in this seminar series.
[00:00:11.133]It's been a great day, a lot of good discussion,
[00:00:13.728]met a lot of great people.
[00:00:16.336]And before I get too far into it,
[00:00:18.749]I need to recognize my coauthors,
[00:00:21.088]Tony Roberts of the Atlantic Flyway office,
[00:00:24.180]Scott Knoche of Morgan State University of Maryland,
[00:00:27.361]Paul Padding, another Atlantic Flyway representative,
[00:00:30.109]and Bob Raftovich, one of our assessment biologists
[00:00:33.119]in our branch of Harvest Surveys.
[00:00:37.928]Just to give you an idea of what I'll cover today,
[00:00:40.929]I'll start with some acknowledgements
[00:00:43.156]that kind of laid the groundwork for this project,
[00:00:47.039]go into some background about waterfowl conservation
[00:00:50.339]in North America and get into really the meat of it,
[00:00:54.049]the purpose of this current analysis and our methods,
[00:00:58.075]the results and the potential applications
[00:01:00.309]to conservation planning and delivery,
[00:01:04.097]and then a few recommendations about how we
[00:01:07.069]can move these kind of efforts forward.
[00:01:13.098]So before I get into it, they hypotheses
[00:01:17.867]and really the intellectual meat that went into this
[00:01:21.787]has come from a variety of waterfowl biologists involved
[00:01:25.465]in the North American Waterfowl Management Plan community.
[00:01:29.029]I was lucky enough to serve as a facilitator
[00:01:32.010]for two structured decision making efforts.
[00:01:35.136]One was by the National Wildlife Refuge System
[00:01:38.544]and they were looking at how to allocate refuge funds,
[00:01:42.996]land acquisition funds each year across the country,
[00:01:46.474]in part to achieve the goals
[00:01:48.243]of the North American Waterfowl Management Plan.
[00:01:51.254]So we spent a week together really hashing those ideas out.
[00:01:55.925]A few months later on a very similar question,
[00:01:59.231]really the same question but a different scale,
[00:02:01.720]there was a group brought together to think
[00:02:04.341]about the same question on the Platte River.
[00:02:06.797]How do we use our resources strategically
[00:02:10.424]on the Platte River System to achieve
[00:02:12.500]the North American Waterfowl Management Plan goals?
[00:02:15.217]And in particular with both of those, we were starting
[00:02:17.562]to wrestle with the idea that now we were trying
[00:02:20.310]to achieve biological goals that really benefit
[00:02:23.510]waterfowl populations, but also this new objective
[00:02:27.320]of trying to benefit or more explicitly
[00:02:30.191]provide full recreationists on the landscape.
[00:02:34.959]Through this whole process, there's been a guiding committee
[00:02:38.210]overseeing the North American Waterfowl Management Plan.
[00:02:41.201]They've put a lot of thought into this material
[00:02:44.235]that we're gonna see today and through the Flyaway System
[00:02:47.780]just spending a lot of time discussing these issues
[00:02:50.302]with a lot of great dedicated and smart biologists
[00:02:53.536]who have really added a lot to this.
[00:02:56.823]And then finally the Cornell Lab of Ornithology
[00:02:59.233]for sharing data.
[00:03:02.933]So just some background.
[00:03:04.553]I think many of you are probably familiar with a lot
[00:03:06.757]of this, but really our waterfowl hunters and birdwatchers
[00:03:11.461]are really gonna provide our reason for being
[00:03:14.041]in a lot of ways, why we do the work we do.
[00:03:17.541]In 2011 in the U.S. Fish & Wildlife Service survey
[00:03:21.445]about dual recreation, there's an estimated 1.5 million
[00:03:24.831]waterfowl hunters spending 17 million days hunting
[00:03:28.548]across the country and spending $1.3 billion
[00:03:32.105]on trip expenses and equipment.
[00:03:35.261]Similarly birdwatchers, 47 million of them in the country
[00:03:39.187]spending $40 billion into the national economy
[00:03:42.426]and supporting an estimated 666,000 jobs.
[00:03:46.867]So really generated a lot of money,
[00:03:48.117]31 billion in employment income.
[00:03:50.286]So they're really contributing a great deal
[00:03:52.698]into our national economy.
[00:03:56.768]Beyond that and perhaps more importantly for us
[00:03:59.147]is they provide the financial and political support
[00:04:02.027]for wildlife conservation in North America.
[00:04:06.567]For example, National Wildlife Refuge System.
[00:04:11.008]They get funds to acquire lands through two sources.
[00:04:14.429]One is the Migratory Bird Conservation Fund,
[00:04:16.707]that's the Duck Stamp which I suspect
[00:04:18.628]most people are familiar.
[00:04:21.994]Yeah, the Duck Stamp.
[00:04:24.065]And since it's inception it's generated
[00:04:25.753]over $800 million and helped purchase
[00:04:28.555]6.5 million acres of habitat in the Refuge System
[00:04:32.114]and most of that, well all of that,
[00:04:34.084]the Migratory Bird Conservation Fund goes towards wetlands
[00:04:37.028]and associated uplands for waterfowl.
[00:04:41.715]They also provide the political support
[00:04:45.271]for the North American Wetland Conservation Act,
[00:04:48.031]which was really put in place shortly after the original
[00:04:50.571]1986 North American Waterfowl Management Plan
[00:04:53.671]and it was designed to provide funds
[00:04:56.431]to really go out and get additional habitat,
[00:04:59.021]wetlands and uplands, to achieve the North American
[00:05:01.519]and has been supported since than
[00:05:03.531]by hunters and birdwatchers.
[00:05:06.062]It's generated over $1 billion of federal funds,
[00:05:09.333]which has been matched by more than $3 billion
[00:05:11.841]in non-federal funds.
[00:05:13.553]To get a knock-up grant, you usually have
[00:05:16.081]to have three to four dollars to match
[00:05:17.971]and your proposal to be successful.
[00:05:22.383]Supported over 2,600 projects to conserve
[00:05:26.245]33 million acres of wetlands and uplands for waterfowl.
[00:05:34.219]So despite that political support and the amount
[00:05:36.975]of money that we have in the System,
[00:05:39.024]it still isn't enough to do what we believe
[00:05:41.126]has to be done on the ground.
[00:05:43.354]Getting back to the Refuge System, they have
[00:05:45.590]requests upwards of $350 to $400 million
[00:05:49.691]to acquire new lands into the refuge system.
[00:05:53.539]In 2004, 70 refuges specifically had $350 million asked.
[00:06:00.984]We only had $42 million in the Land Water Conservation Fund
[00:06:04.287]and just as a side, that is a tax on offshore oil drilling.
[00:06:08.658]And then we had $15 million in the Duck Stamp Fund.
[00:06:11.967]So we certainly didn't have enough money to do the work
[00:06:14.797]we wanted to do and this happens every year
[00:06:17.537]so we have to ask the question, "How do we decide
[00:06:20.498]"to allocate these funds across refuges
[00:06:23.345]"to acquire new lands?"
[00:06:27.288]So with the political and financial support that hunters
[00:06:30.715]and birdwatchers and other conservationists supports,
[00:06:33.728]we need to keep them active.
[00:06:36.688]And with the waterfowl community, we're becoming
[00:06:38.878]very concerned over this because we're seeing a decline
[00:06:41.843]in waterfowl hunters, in fact all hunters nationwide.
[00:06:45.597]Historically we have seen a tight relationship
[00:06:48.952]between the number of duck hunters that are always
[00:06:51.064]buying the Duck Stamp each year with the number
[00:06:54.945]of breeding ducks and our breeding counts
[00:06:59.604]up in Canada every spring.
[00:07:01.705]So going back to 1955, as the numbers drop
[00:07:04.126]hunters drop out for a year, but when the duck numbers
[00:07:06.665]come back up they returned and kept buying Duck Stamps.
[00:07:10.470]Starting in the late '80s, early '90s
[00:07:13.413]this relationship started to fall apart
[00:07:17.769]and we are now looking at duck numbers
[00:07:19.889]that are historical highs and we're just not seeing
[00:07:22.566]that response in Duck Stamp sales.
[00:07:26.844]I'd like to point out before I get too far
[00:07:28.712]that I stole this from a publication by Vrtiska et al.
[00:07:33.466]So this is an issue we've been wrestling with a great deal
[00:07:36.282]and worry that if this continues and we continue
[00:07:40.273]to see this increasing proportion of American society
[00:07:44.072]that's not involved in outdoor recreation at all,
[00:07:47.135]that we're gonna lose the financial support
[00:07:49.561]and political support for wildlife
[00:07:52.231]and specifically waterfowl conservation.
[00:07:59.258]So that's where we've been wrestling with it as a community
[00:08:03.025]and to kind of put this into perspective,
[00:08:05.391]the 1986 North American Waterfowl Management Plan
[00:08:08.720]that was the first one released that was brought together
[00:08:11.501]in response to declining abundance of waterfowl
[00:08:14.511]in the drought conditions across
[00:08:16.370]the Prairie Pothole Region in particular.
[00:08:18.754]It had one real explicit goal to restore the abundance
[00:08:21.082]of North American Waterfowl populations.
[00:08:23.569]And we're gonna do that, it was very specific
[00:08:26.237]by focusing on habitat conservation.
[00:08:28.767]And we're gonna do that through priority regions
[00:08:32.417]and we're gonna bring together partners around tables,
[00:08:34.868]we're gonna call them joint ventures,
[00:08:36.328]they're all gonna contribute whatever resources they can,
[00:08:39.088]identify areas on the ground in those regions
[00:08:41.759]that are most important to waterfowl
[00:08:43.772]and protect it, restore it, manage it
[00:08:46.041]so we can grow the duck populations.
[00:08:49.931]Within that plan, recreation opportunity,
[00:08:54.015]concerns about Harvest Management
[00:08:55.519]were really only implicitly recognized.
[00:08:57.570]It was a biological plan focused on habitat delivery.
[00:09:03.091]So we've been doing that for more than 30 years now
[00:09:05.979]and understanding the relationships
[00:09:07.877]between waterfowl and habitat has really
[00:09:10.228]become our Bread and Butter.
[00:09:11.868]We're very good at studying these relationships
[00:09:14.344]and taking that information and developing
[00:09:17.405]a wide range of decision support tools and models
[00:09:21.095]and frameworks to really help identify again,
[00:09:23.735]how much habitat do we need, where do we need it,
[00:09:27.275]and how do we manage it?
[00:09:28.891]This is just an example of a decision support tool
[00:09:31.866]coming out of the Prairie Pothole joint venture.
[00:09:34.448]One thing you'll notice that this one
[00:09:36.237]is a multi-species priority because since
[00:09:39.632]this original 1986 plan and the original joint ventures,
[00:09:43.544]which were waterfowl habitat joint ventures, we've expanded,
[00:09:48.029]we have joint ventures across the entire country,
[00:09:50.960]and they've become all bird joint ventures.
[00:09:52.967]So they're planning now for more than just waterfowl.
[00:09:57.425]We are really pretty good at it.
[00:09:59.079]Here in the Rainwater Basin Joint Venture,
[00:10:01.597]they've developed a wetland prioritization model
[00:10:04.670]that identifies which wetlands they should be working on
[00:10:08.104]again based on mostly biological issues.
[00:10:11.801]Current density of ducks using it, proximity
[00:10:14.144]to protected habitat, distance from human disturbance.
[00:10:17.724]So again this is focused on identifying what waterfowl need
[00:10:21.966]to achieve our biological goals.
[00:10:25.681]In my neck of the woods, folks see American black duck.
[00:10:28.677]90% of the population winters along the U.S. Atlantic coast
[00:10:32.813]with a good portion in the Chesapeake Bay watershed,
[00:10:36.142]what you see outlined here.
[00:10:37.907]We've been using a bio-energetics approach
[00:10:39.803]to estimate how much food do we need on the ground
[00:10:43.095]to support the number of ducks for the duration
[00:10:45.333]of the non-breeding season that we believe we need.
[00:10:48.894]And then how much food is actually out there?
[00:10:50.755]How many ducks can we support?
[00:10:52.407]So we can identify under watershed levels
[00:10:54.754]areas where we don't have enough habitat and enough food,
[00:10:57.723]areas where we might have excess food.
[00:11:01.639]We've taken that where we're now modeling in urban growth
[00:11:05.595]and sea level rise out to 2030 and 2080
[00:11:08.724]to again inform this kind of decision.
[00:11:11.515]So we're actually pretty good at doing this.
[00:11:14.471]We have a lot of experience, a lot of people
[00:11:16.146]thought really hard about it.
[00:11:19.237]But in 2012 we had a revision
[00:11:21.439]to the North American Waterfowl Management Plan
[00:11:23.418]which has really caused us to go through a lot
[00:11:25.386]of introspection and scratch our heads
[00:11:27.368]and kind of scramble around.
[00:11:31.907]We went back through and had really detailed discussions,
[00:11:34.937]thought about what are our goals?
[00:11:36.696]What are we trying to achieve here?
[00:11:38.534]And two of them came out the same as it's always been.
[00:11:41.551]We want to achieve desired abundance of waterfowl
[00:11:45.742]across the continent of different populations
[00:11:48.952]and we want to provide enough habitat on the ground
[00:11:51.347]to support those waterfowl populations.
[00:11:54.474]But for the first time, the implicit discussion
[00:11:57.823]of recreation, hunting, maintaining the hunting tradition
[00:12:02.806]really became explicit and it was identified
[00:12:05.386]as a third goal.
[00:12:06.718]Often we call it the third leg of the stool
[00:12:09.054]and it's growing the numbers of waterfowl hunters,
[00:12:11.536]other conservationists, and citizens who enjoy
[00:12:13.883]and actively support waterfowl and wetlands conservation.
[00:12:17.605]So now this is the goal, a sociological goal
[00:12:20.316]that's on par with our biological goals.
[00:12:23.626]And this has caused us to think real hard
[00:12:25.816]about our programs, our planning tools, what we do.
[00:12:30.353]How do we now use these same tools, these same program
[00:12:34.492]of the North American Wetland Conservation Program
[00:12:38.174]to acquire lands, restore lands, to actually recruit
[00:12:42.784]and retain hunters on the ground, to facilitate
[00:12:46.044]participation in birding with the idea that the more
[00:12:48.894]that they participate in these outdoor recreation activities
[00:12:52.995]the more they will become this connected to place
[00:12:56.948]and become active supporters, both by financial support
[00:13:00.303]by buying Duck Stamps or by providing political support
[00:13:04.039]that would allow us to do this good work on the landscape?
[00:13:09.588]How, you're talking about a bunch of duckologists?
[00:13:11.758]We really don't know people that well
[00:13:13.928]and we didn't really have a background on it,
[00:13:15.966]so there was a little bit of guidance
[00:13:17.628]that came out in the 2012 plan.
[00:13:19.660]And that was really to do this we need to focus resources
[00:13:22.918]again on important landscapes for birds and people.
[00:13:26.109]So again can we do for people kind of the same planning
[00:13:29.735]that we do for ducks?
[00:13:32.855]And so we really needed to start investigating
[00:13:35.261]how habitat management and land management
[00:13:39.250]influences hunter recruitment and retention,
[00:13:41.907]influences what they do on the ground.
[00:13:44.238]And that's not just for hunters but for birdwatchers too.
[00:13:47.415]They're a big constituency out there.
[00:13:50.839]A lot of people go birdwatching, we saw those numbers
[00:13:53.629]and waterfowl viewing is actually one of the earliest
[00:13:57.238]activities birdwatchers get into.
[00:13:59.024]They're easy to see, they're colorful,
[00:14:00.821]they're easy to identify so they're an important part
[00:14:04.098]of birdwatching actually.
[00:14:07.056]And that's really what we need to focus on
[00:14:09.569]is trying to understand how land management
[00:14:11.863]influences hunter recruitment
[00:14:13.476]and birdwatcher recruitment and retention.
[00:14:17.735]So these groups I talked about earlier,
[00:14:19.623]the National Wildlife Refuge System as the end group,
[00:14:23.085]the Platte River group, they spend a lot of time
[00:14:25.026]thinking conceptually about how these systems work.
[00:14:28.618]This is a typical process and the decision analysis.
[00:14:32.536]You've got to map out the system and be able
[00:14:35.558]to make predictions so you have to understand
[00:14:37.717]all of the relationships.
[00:14:39.047]And this is just one example of what
[00:14:42.201]we went through there on the Platte River.
[00:14:43.785]You can see all the interconnected relationships.
[00:14:48.252]We have some fundamental objectives
[00:14:50.325]of maintaining waterbird populations,
[00:14:52.777]threatened and endangered populations,
[00:14:55.195]and then this conservation participation.
[00:14:57.827]If you follow this all the way through,
[00:14:59.323]what it really comes down to is fowl regimes
[00:15:01.905]on the Platte River and river access there
[00:15:05.033]and that's really what we focused on.
[00:15:07.052]River access influences participation
[00:15:09.234]in both hunting and birdwatching.
[00:15:11.773]The more they get out there, they create this connection
[00:15:14.153]to place and then they become active conservationists.
[00:15:20.299]All the other groups in all the other discussions
[00:15:22.379]I was involved in revolved around these same discussions,
[00:15:27.284]came up with a lot of the same hypotheses,
[00:15:29.719]and then we started to think, "Can we get beyond hypotheses
[00:15:32.326]"and start drawing functional forms
[00:15:34.352]"and think about how can we actually
[00:15:36.145]"fit data to these forms?"
[00:15:37.724]So we spent time on chalkboards and flip charts
[00:15:41.004]trying to draw out these relationships
[00:15:42.873]and these are just some examples of what we came up with
[00:15:45.645]is that if we start thinking about this
[00:15:48.038]in a population dynamics kind of framework,
[00:15:50.689]we're thinking about survival and recruitment
[00:15:53.415]and those are positively related to accessibility
[00:15:56.748]of recreational land so now we're talking
[00:15:58.669]about survival and recruitment of hunters and birders.
[00:16:02.300]The one that comes out always is this travel distance.
[00:16:05.672]People don't want to go too far,
[00:16:07.269]they don't want to have to drive too long
[00:16:08.670]to go birdwatching or hunting
[00:16:11.179]and so it's negatively related to that.
[00:16:14.011]Of course there has to be birds in the area.
[00:16:15.830]You have to have some chance to see a cool bird.
[00:16:18.050]You have to have some chance to call in a duck or two
[00:16:20.228]and maybe take a shot at him, you don't necessarily
[00:16:22.161]have to fill your bag.
[00:16:24.030]But there needs to be some birds in the area
[00:16:26.910]and probably the larger the recreational area the better.
[00:16:29.891]So we have a whole bunch of these functional relationships
[00:16:32.705]and hypotheses kind of floating out there.
[00:16:35.648]But this is usually about as far as we could get.
[00:16:38.341]We're duckologists, we haven't been
[00:16:41.013]in the sociological literature.
[00:16:42.431]We don't know what data is out there,
[00:16:43.814]we don't know what analytical techniques are out there
[00:16:47.262]to kind of start addressing these hypotheses
[00:16:51.268]and this is usually about where we would stop.
[00:16:55.015]The only way I can summarize it is that retention
[00:16:57.476]and recruitment of birders and hunters is influenced
[00:17:00.065]by access to recreational areas.
[00:17:02.776]Access is a function of the amount of public land
[00:17:05.876]or at least land, it can be privately owned,
[00:17:08.044]but that is available to the public to recreate on,
[00:17:11.156]the amount and quality of those lands, travel distance,
[00:17:15.020]presence of birds, density of hunters.
[00:17:19.119]This is different for hunters and birders.
[00:17:21.850]For hunters we want low density and for birders
[00:17:23.991]we think it's more of a social activity
[00:17:27.140]so the more birders you have out there,
[00:17:28.801]that might be a good thing actually.
[00:17:30.690]And then infrastructure.
[00:17:31.822]Does the site have a boat ramp?
[00:17:33.381]Doe it signage and trails for birders?
[00:17:35.801]And those all might influence where someone
[00:17:38.315]chooses to recreate and how often they choose to recreate.
[00:17:44.486]But as I said, that's usually about as far as we got.
[00:17:46.958]We didn't know how to actually test these hypotheses,
[00:17:50.479]what tools are out there or what data are out there.
[00:17:52.691]We usually talked about conducting surveys.
[00:17:56.739]And then one day it occurred to me after seeing
[00:17:59.038]some other modeling efforts on other things
[00:18:02.370]that we might actually have data,
[00:18:04.306]particularly just on that travel distance.
[00:18:06.140]That's the thing that we really keyed in on,
[00:18:08.160]that people don't want to travel far.
[00:18:10.047]Half an hour, an hour, two hours, we didn't know.
[00:18:13.721]And it occurred to me that we might have existing data
[00:18:16.242]within the Waterfowl program actually
[00:18:19.153]to at least start looking at this,
[00:18:21.661]trying to describe some general patterns.
[00:18:24.148]And that is data that comes out
[00:18:25.576]of our Harvest Survey program, our Band Encounter data,
[00:18:29.251]and then the eBird data that would provide
[00:18:32.366]trip information really.
[00:18:34.310]And then could we then apply these results
[00:18:36.139]to our habitat planning and delivery?
[00:18:38.030]So that's really the objectives of this project
[00:18:41.182]that I'm talking about now.
[00:18:44.671]So with these data you can do a lot of simple summaries
[00:18:47.616]just on how far people are traveling,
[00:18:50.135]how many people are traveling, out of state,
[00:18:52.405]in state, same county, things like that.
[00:18:54.997]A lot of ways you could look at that and that's interesting
[00:18:57.865]in itself, but then what I'd really like to focus on
[00:19:00.874]is the application of Discrete Choice models,
[00:19:03.777]also known as random utility models, and then using
[00:19:06.451]the output from those models
[00:19:08.005]in decision support applications
[00:19:10.446]or actually decision analysis.
[00:19:15.910]So I'm not an expert in the Discrete Choice model.
[00:19:18.678]At one point we hit a wall and were able to reach out
[00:19:22.668]to Scott Knoche at Morgan State University,
[00:19:26.175]who is a Resource Economist, and he helped us
[00:19:29.199]run through these.
[00:19:30.478]But the Discrete Choice model have been used commonly
[00:19:33.617]actually in natural resources management.
[00:19:36.589]It's also called a random utility travel cost model
[00:19:39.368]and what it does is it allows you to evaluate
[00:19:42.063]relationships between site selection and the attributes
[00:19:45.052]of those sites, like the presence of boat ramps or signage
[00:19:47.707]or things like that, and substitute sites
[00:19:50.816]and figure out what people are actually
[00:19:52.665]selecting for in the landscape.
[00:19:55.690]The model we used was a maximum likelihood estimation
[00:19:58.651]and once you fit a model, you can then start changing
[00:20:01.635]the parameters in that model to predict changes
[00:20:04.824]in site selection and participation
[00:20:07.965]as a result of those changes in those attributes.
[00:20:10.400]So if you add a boat ramp, will more people
[00:20:12.849]come to that site over others?
[00:20:15.021]Which is really useful for what we want to do.
[00:20:19.289]I'll just mention now that there are a variety
[00:20:21.932]of formulations of these methods that can be used
[00:20:24.820]and some of the more recent ones are more flexible,
[00:20:27.781]particularly the nested logit and mixed logit models.
[00:20:31.896]And then I've been reading recently
[00:20:33.750]about hierarchical formulations of the models
[00:20:36.245]where you can actually integrate multiple
[00:20:37.969]different data sets into the same model
[00:20:40.548]to explore these relationships.
[00:20:45.280]So the data sets that we're using really come out
[00:20:49.107]of our Harvest Management programs.
[00:20:51.336]We have a branch of Harvest Surveys
[00:20:53.104]that has two monitoring programs,
[00:20:54.897]one's called a diary survey
[00:20:56.598]and the other's a parts collection survey.
[00:20:58.728]And both of these surveys or monitoring programs
[00:21:01.225]were set up to allow us to estimate
[00:21:03.329]the total continental harvest or U.S. harvest,
[00:21:07.258]Canada does the same thing, the U.S. harvest
[00:21:09.548]of waterfowl by species, the days of field,
[00:21:12.349]things like that.
[00:21:15.129]And we use it in two different ways.
[00:21:17.058]One, we take a sub-sample of hunters and we ask them
[00:21:19.989]to complete a diary survey.
[00:21:22.439]That's this form here that you can't see too well,
[00:21:25.600]but what we do is we ask each hunter each year
[00:21:28.456]that's in the sample to record every day
[00:21:31.468]that they go out hunting.
[00:21:33.090]Tell us what county they go hunting in,
[00:21:35.459]the number of ducks, the number of geese,
[00:21:37.480]number of brant, and the number sea ducks they shoot.
[00:21:40.040]More importantly, if they don't shoot anything,
[00:21:42.350]they still are supposed to report that day.
[00:21:45.229]And so with that we now know where the hunter lives
[00:21:47.578]and we know at least on a broad scale
[00:21:50.419]where he or she went to hunt.
[00:21:54.056]Parts collection survey is a different sample of hunters
[00:21:57.153]from the diary survey, but those that are participating
[00:22:01.008]in that are asked to send us a wing from every duck
[00:22:04.459]they shoot or a tail from every goose they shoot
[00:22:07.496]and that way we can get the species composition
[00:22:09.750]of the harvest sex and age ratios.
[00:22:12.339]But again, we know where that hunter lives
[00:22:14.478]and they record on the form where they hunted.
[00:22:17.169]So we can't get information on days that they don't
[00:22:20.048]shoot anything, but we can get a lot more
[00:22:22.358]on the species composition of the harvest.
[00:22:26.428]So those are two data sets we have available to us.
[00:22:28.758]The third one is our banding program
[00:22:31.492]and this is really the workhorse
[00:22:35.504]of our Harvest Management program.
[00:22:37.628]We go out and band ducks every summer
[00:22:40.116]between June and August.
[00:22:42.356]Catch a duck and put a band on them and we'll release them.
[00:22:45.334]A hunter shoots a duck, they can return it to BBL
[00:22:48.147]and they'll get a certificate that says,
[00:22:49.932]"You shot a mallard that was banded in Saskatchewan,
[00:22:53.026]"it was a female" and all kinds of other information.
[00:22:55.838]And we use that to estimate harvest rates.
[00:22:59.811]Until 2011, if that hunter wanted a certificate,
[00:23:03.153]they had to provide their address so now we know
[00:23:05.565]where they live and with the banding information
[00:23:08.043]we have pretty fine-scaled to the ten minute block,
[00:23:10.783]one minute block of where they actually harvested that bird
[00:23:14.246]so we now know where they traveled to and from.
[00:23:17.305]So those are three existing data sets we have
[00:23:19.615]to look at travel movements of waterfowl hunters.
[00:23:23.574]For birdwatchers it's a little more difficult
[00:23:26.173]and the one we came up with was eBird.
[00:23:28.703]I think most people are probably familiar with this,
[00:23:30.999]but this is a voluntary program for birders
[00:23:35.210]to share their observations and what they're seeing
[00:23:37.323]out there when they go birding.
[00:23:39.240]So to do that they have to create an account,
[00:23:41.591]they have to give you their name,
[00:23:42.871]and then you could put in your information
[00:23:44.230]about where you live, some information
[00:23:46.241]on your education level, your work experience,
[00:23:49.849]things like that, how much money you make.
[00:23:53.175]And then you upload your observation.
[00:23:54.890]I saw five mallards today and they usually pinpoint
[00:23:57.041]exactly where they were on the map
[00:23:58.716]so now you know where they live
[00:23:59.912]and where they go out to go birdwatching.
[00:24:03.965]Now we have these trip logs on both
[00:24:06.279]waterfowl hunters and birders.
[00:24:16.624]So that really allows you to start looking
[00:24:19.265]at the Discrete Choice model sets.
[00:24:21.943]For the explanatory variables, again we used a variety
[00:24:24.571]of publicly available data sets.
[00:24:27.000]The National Landcover Database provided
[00:24:29.090]habitat availability information.
[00:24:31.339]The National Wetland Inventory also provided
[00:24:35.269]Protected Lands Database allowed us to identify areas
[00:24:38.241]that are available for recreation and then within
[00:24:41.161]that category which ones are open to hunting.
[00:24:44.972]And then we used U.S. Census data to get information
[00:24:47.085]about abundance and demographics
[00:24:49.081]of the general populace at the county level.
[00:24:55.520]I'll give you some background on the amount
[00:24:58.977]of data that we have for this.
[00:25:00.770]Between 2005 and 2013 was the period we were looking at
[00:25:04.420]and this was only for the Atlantic Flyway states,
[00:25:06.738]so we're talking Maine through Florida
[00:25:09.794]and it does include West Virginia.
[00:25:13.106]So for the Harvest Surveys, we're looking over
[00:25:15.617]161,000 records with almost 27,000 unique individuals
[00:25:20.800]and on average we're sampling over 3,000 individuals.
[00:25:24.666]In the Atlantic Flyway, a hunter takes on average
[00:25:26.942]four hunting trips per year.
[00:25:30.544]And the eBird data, a much larger data set.
[00:25:33.422]We had 1.5 million total trips during that time period
[00:25:36.482]with 19,000 unique individuals providing observations.
[00:25:40.606]About 4,000 per year and they take on average
[00:25:44.094]20 birding trips per year.
[00:25:45.899]I should note here that when we're taking
[00:25:47.534]four hunting trips per year, that is just
[00:25:49.722]during the hunting season of course,
[00:25:51.455]whereas 20 birding trips per year is the 12 month calendar.
[00:25:55.809]I would also note that the eBird data,
[00:25:58.230]the number of people contributing to eBird
[00:26:00.412]was growing exponentially during this time.
[00:26:02.791]It's really a massive data set.
[00:26:06.400]I'm sure a lot of people have different opinions about
[00:26:09.137]for different uses, but for this
[00:26:11.785]it was pretty interesting and useful.
[00:26:15.750]So I told you we could do a lot with this.
[00:26:17.711]You can just start exploring doing some
[00:26:19.325]basic simple summaries with these data
[00:26:21.100]to start getting some idea of how hunters and birdwatchers
[00:26:25.100]move across the landscape.
[00:26:28.804]The biggest thing that we're focused on
[00:26:30.231]is how far are they willing to go.
[00:26:31.781]Again, that was the thing everyone focused on.
[00:26:34.060]Kind of guessed maybe about an hour
[00:26:35.772]is how far they're gonna go so maybe 60 miles or so.
[00:26:39.739]And we just kind of implicitly assumed that
[00:26:42.320]with the negative linear relationship.
[00:26:44.920]So first thing we did was just start plotting distance
[00:26:47.342]that hunters were taking and birdwatchers.
[00:26:50.747]So in the top left, you can see a histogram
[00:26:53.802]of the trip distance of Atlantic Flyway hunters,
[00:26:57.001]so this is pulled across states.
[00:26:59.232]You can see that the vast majority of the trips
[00:27:03.075]happened within 50 kilometers of a person's residence
[00:27:06.433]and certainly within 100 kilometers.
[00:27:08.468]They're not going very far.
[00:27:10.058]When you looked at this at the state level,
[00:27:11.561]it was the exact same pattern across each state.
[00:27:15.983]Yeah, exactly the same.
[00:27:17.842]So it's not only a negative linear negative relationship,
[00:27:20.306]but it was negative exponentially.
[00:27:21.997]They're really not going far at all.
[00:27:24.155]Looked at birders and we saw the exact same relationship.
[00:27:26.499]In the bottom right there, that's a histogram
[00:27:28.978]of birding trips in Florida, so just at the state level.
[00:27:34.028]And again, you see basically the same pattern
[00:27:36.104]and when you pulled this across the Atlantic Flyway
[00:27:38.202]it was the same thing.
[00:27:39.223]So they're not going far from home to participate
[00:27:41.611]in either birdwatching or hunting.
[00:27:45.977]To tell you the truth, that was it.
[00:27:47.363]That was really all I ever wanted to do with this,
[00:27:49.577]was to look at this data and call it good,
[00:27:51.622]but then we started to explore the Discrete Choice models.
[00:27:56.474]Here's some other ways you can start to visualize
[00:27:59.882]the data and just explore it.
[00:28:02.434]We love to make maps often and so we started doing that.
[00:28:07.038]On the left here, you can look at the proportion of people
[00:28:09.280]who hunted in their county of residence.
[00:28:11.591]So the darker blue there, and this is just for New York,
[00:28:14.918]the darker blue there means that they don't have people
[00:28:17.901]coming from out of the county to hunt there.
[00:28:19.671]Any trips that are happening there are mostly
[00:28:21.257]just by county residents.
[00:28:22.890]The lighter areas means that a lower proportion
[00:28:26.455]of those trips was from county residents
[00:28:29.694]so those are areas people are moving to to hunt.
[00:28:33.661]On the right is the average distance traveled
[00:28:36.880]to a county to hunt.
[00:28:38.560]So if you look closely at it, it's kind of a mirror image.
[00:28:42.087]Hunters are traveling further to hunt
[00:28:45.539]in these northern counties in New York
[00:28:48.670]and not so much down here, just north of New York City.
[00:28:52.939]So just some real simple ways to start exploring these data
[00:28:56.242]and start to think about the relationships
[00:28:58.222]between land management, public access,
[00:29:01.853]and recreational opportunity and participation.
[00:29:08.081]So what we found there actually was that somewhere between
[00:29:11.922]55 and 60% of all hunting and birding trips
[00:29:17.393]occur within the county of residence
[00:29:19.776]and then 92 to 93% of all hunting and birding trips
[00:29:23.540]occur within state of residence.
[00:29:25.412]So for the most part we're not going out of state
[00:29:28.201]for the majority of our hunting or birdwatching trips,
[00:29:31.415]in either case.
[00:29:33.706]So given that, what we decided to do was to apply
[00:29:36.132]these Discrete Choice models at the state level.
[00:29:39.518]It was a little more trackable for us
[00:29:41.933]and we just subjectively chose New York and Georgia
[00:29:44.873]as two examples to do this.
[00:29:47.593]We fit those relationships, those hypotheses
[00:29:50.496]I talked about early and ran the Discrete Choice model.
[00:29:54.935]What we found for New York was that there
[00:29:56.465]was a negative relationship to travel distance.
[00:29:58.977]That's not surprising, but a positive relationship
[00:30:02.129]to the total wetlands in the county, total area
[00:30:05.116]of the county, and whether or not the county
[00:30:06.938]was coastal to a Great Lake or to the Atlantic coast.
[00:30:10.728]And again, that's not overly surprising.
[00:30:12.854]They're gonna go where the ducks are
[00:30:14.418]and ducks are right along the coast there.
[00:30:18.107]For New York, we did have one interesting result
[00:30:20.409]that made us scratch our head and we found
[00:30:21.880]a negative relationship to the amount
[00:30:23.578]of public land in the county.
[00:30:26.296]And that really made us pause, thought really hard about it.
[00:30:29.999]If you look into the published literature,
[00:30:32.180]actually a vast majority of waterfowl hunting trips
[00:30:34.871]occur on private land.
[00:30:36.788]Most of our wetlands are on private land,
[00:30:38.966]so maybe that's the explanation.
[00:30:41.047]But also in New York, the vast majority of the public land
[00:30:44.167]is actually forested land up and at around state parks,
[00:30:47.731]areas that the ducks aren't in
[00:30:49.965]and so it could be as simple as the public land
[00:30:52.679]doesn't hold the wetlands.
[00:30:54.049]So there's some really easy ways to improve that model
[00:30:57.381]that we didn't get to at the time.
[00:31:00.651]That's what we found.
[00:31:02.025]In Georgia, again negative relationship to travel distance,
[00:31:05.734]positive relationship to coastal counties again,
[00:31:08.465]and in Georgia we found a positive relationship
[00:31:10.272]to public lands so they're going to those counties
[00:31:12.146]that had particularly national wildlife refuges
[00:31:15.799]down around Okefenokee.
[00:31:19.821]So some things that you can do here,
[00:31:21.840]thinking about this in terms of how we usually
[00:31:25.285]start planning for waterfowl and habitat,
[00:31:27.973]we often try to find targeted areas that we need habitat
[00:31:31.890]or the best place to put habitat on the ground
[00:31:34.521]that would elicit some response in population dynamics,
[00:31:38.262]increased survival, increased abundance,
[00:31:42.642]So the first thing we did was just real simple.
[00:31:44.993]What would be the percent change in hunting trips
[00:31:47.991]to every county if we increased the amount
[00:31:50.141]of wetlands by 10% simultaneously across these two states?
[00:31:53.872]So you just get this really kind of simple index
[00:31:56.458]of predicted percent change in trips.
[00:31:59.601]So New York, these are hunters.
[00:32:01.642]They're in St. Lawrence County, we would expect
[00:32:04.193]to get about 1.5% increase in number of trips to that area,
[00:32:08.069]whereas in some of these other counties of lighter shade,
[00:32:10.811]you can add more wetlands there
[00:32:14.078]and you still wouldn't see more trips there.
[00:32:16.000]They're just not going to go to that area
[00:32:17.582]because it's too far or for a variety of other reasons.
[00:32:20.435]So now you can start thinking real simplistically,
[00:32:22.519]"Where is the best place to work if we really
[00:32:24.783]"want to focus on recruiting or providing opportunity
[00:32:27.642]"for hunters and assuming that's gonna feed
[00:32:30.971]"into retention and recruitment?"
[00:32:32.985]Same thing with birders and we're seeing similar patterns.
[00:32:36.417]In Georgia, we have one county in particular
[00:32:39.171]that stuck out for hunters and then another good one
[00:32:42.881]that's down around Okefenokee Swamp actually
[00:32:46.050]and that is where we would see the biggest
[00:32:48.481]bang for our buck for birders too.
[00:32:51.182]Again this is kinda typical of how we approach
[00:32:53.820]thinking about targeted areas for waterfowl.
[00:32:57.736]Every one of our JVs across the country
[00:33:00.043]has developed biological focal areas,
[00:33:02.832]priority areas, and different names
[00:33:04.762]for waterfowl, landbirds, waterbirds.
[00:33:08.139]Here you can see the focal areas identified
[00:33:10.393]by the Atlantic Coast Joint Venture for New York
[00:33:13.412]up along the St. Lawrence valley, the Hudson River.
[00:33:17.328]And I'll draw your attention to the Finger Lakes region
[00:33:21.136]if you're not familiar with New York,
[00:33:24.068]a really important area for waterfowl,
[00:33:26.596]a great place a lot of birds migrate
[00:33:28.276]from right over the Great Lakes
[00:33:30.274]shoot right down through the Finger Lakes area
[00:33:32.078]and then down through the coast.
[00:33:33.317]It's really important for waterfowl.
[00:33:36.055]Now with being able to kind of develop similar
[00:33:43.506]using that previous map where we just predicted
[00:33:46.138]change in trips based on increasing wetlands
[00:33:49.319]in the county, you can start looking at the overlap
[00:33:51.638]and a simple way to think about this
[00:33:53.589]is just to prioritize your habitat delivery
[00:33:56.039]in those areas that are both good for ducks
[00:33:58.468]and good for people.
[00:34:00.299]So it becomes real clear in New York along the Hudson River
[00:34:03.869]is a great place to work, let me find my mouse,
[00:34:07.537]for ducks and people.
[00:34:09.849]Along the St. Lawrence valley,
[00:34:11.279]particularly St. Lawrence County, great place
[00:34:13.141]to work for both ducks and people.
[00:34:15.120]But I told you about the Finger Lakes region,
[00:34:17.883]already has some docks, but we can do a whole bunch
[00:34:20.376]of work there potentially and we're not gonna increase
[00:34:23.296]a whole lot of the number of hunting trips
[00:34:25.827]or birding trips into that area.
[00:34:27.957]So now what do we do?
[00:34:30.649]Do we keep putting money in there for waterfowl or not
[00:34:34.404]because it's not giving us a benefit for hunters?
[00:34:37.927]So we have to now start thinking about the potential
[00:34:40.303]trade-off between providing habitat for ducks
[00:34:42.853]versus providing habitat for people.
[00:34:51.309]To me that's useful, that kind of fits into how we do
[00:34:55.762]a lot of planning and informed decisions
[00:34:58.702]for waterfowl in general, but I think it's even
[00:35:02.005]more interesting to drill down into this
[00:35:04.315]and really apply it to a decision
[00:35:07.107]and the way that we actually deliver habitat
[00:35:09.310]on the ground is through a variety of these
[00:35:11.267]competitive grant programs.
[00:35:13.267]So the National Wildlife Refuge System,
[00:35:15.037]not so much a grant program, but all the refuges
[00:35:18.117]are competing for that limited money.
[00:35:20.366]They're all putting in the proposals
[00:35:21.865]and trying to say why they're the most important
[00:35:23.826]and they're handing out money that way.
[00:35:26.691]North American Wetlands Conservation Act,
[00:35:29.893]we have knock-up grants to acquire uplands
[00:35:34.107]and wetlands for waterfowl.
[00:35:36.782]This has been the primary tool used by the joint ventures
[00:35:40.077]to secure habitat for waterfowl and currently right now
[00:35:44.016]there's actually 2% of the waiting goes to public access,
[00:35:49.286]whether or not the acres on there provide public access.
[00:35:53.760]And we want to see if we can use this
[00:35:56.344]rather than general text kind of vague way
[00:35:59.634]that they're scored right now.
[00:36:01.205]Could we use this Discrete model to come up with a more
[00:36:04.472]objective and intuitive metric to evaluate
[00:36:10.100]And so from those Discrete Choice models
[00:36:12.042]we can estimate the change in trips for waterfowl hunters
[00:36:15.644]and birders based on each one of these knock-up proposals.
[00:36:18.959]So what we did was just take four historic
[00:36:20.793]knock-up proposals, submit it to the ACJV in New York,
[00:36:24.841]and ran them through our Discrete Choice models.
[00:36:28.548]And what this eventually leads to
[00:36:30.235]is this multi-criteria portfolio analysis.
[00:36:34.063]Basically which project should we fund,
[00:36:36.370]which one maximizes your utility?
[00:36:41.302]So here just a simple summary of those four projects
[00:36:44.154]that we looked at.
[00:36:45.475]Three of them are acquisitions, the fourth one
[00:36:48.400]includes acquiring new lands and also restoring
[00:36:50.963]some existing protected land.
[00:36:53.819]Within the proposal, you know the total acreage
[00:36:56.380]that they're gonna acquire to restore,
[00:36:59.003]how much that is wetlands, and then they
[00:37:01.375]also have to tell you how much of that land
[00:37:02.924]is gonna be open to recreation.
[00:37:05.186]In some of them it's all open to recreation,
[00:37:07.594]some of them it's open to birding everywhere
[00:37:09.684]and hunting in some of it and you can
[00:37:11.861]pull that information out.
[00:37:13.802]You can also identify whether or not
[00:37:15.334]it's along the Great Lake coast or along the Atlantic Coast,
[00:37:19.184]how much it costs, and then of course what county it's in.
[00:37:22.294]So what you can do is pull that information out
[00:37:24.223]from each knock-up proposal, put that back
[00:37:26.432]through your Discrete Choice model, and you predict
[00:37:29.444]the number of trips that will be taken statewide
[00:37:32.405]across every county, so it'll be the distribution
[00:37:35.354]of trips and the total number of trips.
[00:37:38.683]And this is kind of what it looks like.
[00:37:40.153]You have every county in the state in the left-hand column
[00:37:43.059]there and then the number of trips taken
[00:37:45.704]so in the status quo, this is what we got
[00:37:48.145]from our original Discrete Choice Model.
[00:37:51.384]How many trips are taken in each county?
[00:37:53.624]We sum that up at the bottom for total number of trips
[00:37:56.414]and then you can look at the change.
[00:37:58.591]You plug in proposal one, run the model again.
[00:38:01.446]Now you change the distribution of trips
[00:38:03.536]across the counties and also the total number taken.
[00:38:06.315]This one we estimate would generate six new trips statewide.
[00:38:10.913]Proposal two, same thing, estimated fifty
[00:38:13.872]additional trips for waterfowl hunters.
[00:38:15.878]Proposal three is nine and proposal four gives us 51.
[00:38:20.014]So this is really a nice intuitive objective metric
[00:38:23.066]that we could plug into a Decision Analysis framework
[00:38:25.738]to allow us to really decide how to allocate funds
[00:38:28.408]across the landscape.
[00:38:33.002]And then you can think about how to extend this to beyond
[00:38:36.482]just a single objective Decision framework,
[00:38:38.654]just providing habitat for hunters and birders,
[00:38:41.478]and combining that with our biological objectives,
[00:38:44.075]so taking all those tools that we've developed
[00:38:46.078]over 30 years to tell us where we should
[00:38:48.307]put habitat on the ground for ducks.
[00:38:51.091]We're really good at that and we can evaluate
[00:38:53.546]these proposals to determine which ones
[00:38:55.129]are best for ducks and which ones aren't.
[00:38:58.130]For most of our non-breeding habitat joint ventures,
[00:39:00.946]we are using some form of bio-energetics modeling
[00:39:04.010]in most cases, estimating duck use-days on the ground.
[00:39:07.968]So from each model you can think about how many
[00:39:11.848]duck use-days it provides, does it increase duck use-days,
[00:39:15.415]which of course would only happen if you
[00:39:17.203]are restoring degraded habitat.
[00:39:19.593]If you're just acquiring it, you're not really
[00:39:21.054]increasing the duck use-days, how many acres.
[00:39:24.663]And then with the Discrete Choice modeling
[00:39:26.050]we can now plug in the number of trips
[00:39:28.389]and this gives us a really nice way to do
[00:39:30.584]the trade-off analysis and decide how to allocate
[00:39:33.820]those limited resources to those projects
[00:39:36.130]that maximize utility for both ducks and people.
[00:39:41.800]So what we found or what we believe
[00:39:44.588]is that Discrete Choice Modeling is really
[00:39:46.481]a useful framework for us to explore and start to understand
[00:39:50.618]what factors drive participation and site selection
[00:39:54.237]of hunters and birdwatchers on the landscape.
[00:39:57.049]And then we can incorporate those results into our planning.
[00:39:59.788]The assumption that if we're putting more habitat
[00:40:02.999]on the ground where people can access it,
[00:40:05.379]the right types of land in the right places,
[00:40:08.769]they're gonna participate more, they're gonna buy
[00:40:11.248]more Duck Stamps, they're become active conservationists
[00:40:14.149]and provide us financial and political support.
[00:40:17.972]However, I would say this is a really pretty rough
[00:40:20.991]first draft at applying these data sets
[00:40:24.697]and these tools by a bunch of Duck Heads
[00:40:27.629]and we really should be reaching out
[00:40:29.168]to the recreation management and resource economics fields.
[00:40:32.573]These guys use these tools all the time
[00:40:34.521]to study these questions and I think they could bring
[00:40:36.909]a lot to the table for us.
[00:40:40.589]I'd also say future work should be working
[00:40:42.874]at the state and county scale.
[00:40:44.760]As I said the vast majority of trips
[00:40:46.156]are happening within the state and then even
[00:40:48.524]actually within the county of residence.
[00:40:52.044]The benefit of doing that also might be
[00:40:54.215]that I think a lot of state agencies have additional data,
[00:40:57.065]perhaps better data than we do at the federal level
[00:40:59.883]to support this kind of modeling effort.
[00:41:03.205]Many states, at least in the Atlantic Flyway,
[00:41:05.853]conduct a limited quota hunts for waterfowl
[00:41:08.581]so you have to apply to hunt in certain areas.
[00:41:10.883]They know where you live, they know where you want to hunt,
[00:41:13.054]even if you don't get drawn.
[00:41:15.324]That can be used in this Discrete Choice Modeling.
[00:41:18.479]They may also have more information, spacial information
[00:41:21.557]on where their WMAs are, how much is open to hunting,
[00:41:24.763]what kind of hunting, are there boat ramps,
[00:41:27.170]just some of those other hypotheses and factors
[00:41:29.600]that we didn't get to look at may be available
[00:41:32.541]at the state agencies.
[00:41:34.738]Other partners may have data to contribute,
[00:41:37.043]National Wildlife Refuge System perhaps
[00:41:39.023]or maybe some of the non-government organizations.
[00:41:42.269]The U.S. Fish & Wildlife Service I think
[00:41:44.947]we can maybe add to this process going forward.
[00:41:48.871]We've actually had some discussion on those Harvest Surveys,
[00:41:52.528]so that diary survey and the parts collection survey,
[00:41:55.199]if we should add questions on those envelopes
[00:41:58.252]that ask them, "Did you hunt on public or private land
[00:42:01.431]"and what was the nearest town?"
[00:42:02.606]And that would give us a little more final spacial scale
[00:42:04.926]to conduct these models.
[00:42:09.613]And so with that, I'll take any questions
[00:42:12.079]that you may have and thank you for coming out.
[00:42:15.623]Just a simple one and I really enjoyed
[00:42:18.323]kind of the approach you took.
[00:42:19.969]I always like to compare
[00:42:21.688]what the computer tells us to what people
[00:42:24.569]are already kind of thinking or doing
[00:42:26.237]and whether they've got some of these algorithms
[00:42:28.814]running around in their heads sometimes.
[00:42:30.842]So when you had the county in New York
[00:42:33.519]and the county down in Georgia that would have gotten
[00:42:36.981]the biggest bang for your buck, I'm just interested,
[00:42:39.267]do you know those states well enough
[00:42:41.138]that the people are already kind of thinking,
[00:42:44.099]"That'd probably be a good place" or was that
[00:42:45.753]a complete surprise with those being the best?
[00:42:48.493]I've asked the reps in those two states
[00:42:51.837]and they thought it made sense.
[00:42:54.139]One of the reason those came out is because people
[00:42:56.088]are already going for those areas.
[00:42:57.777]St. Lawrence County there in New York right on the river,
[00:43:00.716]they have a lot of public land and a lot
[00:43:03.347]of waterfowl hunting opportunities so they felt
[00:43:05.556]that made a lot of sense.
[00:43:07.246]And then also on Georgia, that's an area where they've done
[00:43:11.025]a great deal of wetland conservation through NAWCA
[00:43:14.833]and have a lot of public land for waterfowl
[00:43:17.028]and so I thought that made sense.
[00:43:18.893]And of course Okefenokee Swamp for birders,
[00:43:21.216]it's a big birding place so it's providing
[00:43:23.429]them more opportunity in there to get in and see birds.
[00:43:26.978]I thought it made some sense.
[00:43:30.421]It is interesting, in a previous version of this
[00:43:34.384]before we really got into this I went to the reps
[00:43:38.371]at Atlantic Flyway and told them what we were thinking
[00:43:40.565]of doing and just kind of ran it by them
[00:43:43.881]and so a rep from Georgia went ahead
[00:43:47.490]and got to me bird data, got his banding data
[00:43:49.955]and his Harvest Surveys data and did some other ways
[00:43:53.655]of looking at these, some other simple summaries
[00:43:56.278]and he had very different areas where he worked.
[00:44:01.231]But then ours, he thought ours made a lot of sense too
[00:44:04.869]and ours lined up a lot with where their focal areas,
[00:44:06.942]right where they already thought was good for waterfowl.
[00:44:17.961]In the 2012 acquisition and inclusion
[00:44:21.369]of the sociological components focused on viewers
[00:44:23.880]and hunters and those resource users, I guess I'm curious
[00:44:28.626]about your thoughts if we were to sort of take a step back
[00:44:33.319]and include non-users in that conversation
[00:44:37.330]from a sociological perspective, from those that we might
[00:44:40.721]be able to pull into the field
[00:44:43.310]for that political support or that conservation support
[00:44:47.451]from a non-user prospective, those who may never be
[00:44:49.910]non-users, but recognize them as important conservationists
[00:44:53.210]in the field.
[00:44:55.290]Yeah, there's been a great deal
[00:44:56.455]of conversation about that.
[00:44:58.976]We are really serious about this, we've been thinking
[00:45:03.013]about a variety of stakeholders and those that may never
[00:45:06.383]go out to a wetland and how to engage them.
[00:45:09.591]It become difficult to identify who they are
[00:45:12.274]and how to work with them.
[00:45:14.232]We are funding a very extensive, a big effort
[00:45:18.453]to conduct a national survey, we're doing a survey
[00:45:21.833]of hunters, doing a survey of birdwatchers,
[00:45:23.675]and then a general public one to try
[00:45:25.495]to get at some of those pieces.
[00:45:28.933]You're right, we've got to think about all of those.
[00:45:30.885]This one was specific to how do we consider
[00:45:34.656]at least these two constituents just in our current
[00:45:37.324]operations of how we give a habitat,
[00:45:39.403]but those broader conversations, they're being had
[00:45:41.965]in other vendors.
[00:45:47.166]Do you think you reach a point
[00:45:48.737]of diminishing returns when you're adding opportunity?
[00:45:51.848]So if you have a space that already has a lot of opportunity
[00:45:54.898]and a lot of use, if you have 50 visits that you could
[00:45:58.184]add through a management action, are they more valuable
[00:46:00.306]there or are they more valuable in a place
[00:46:02.274]that has less opportunity?
[00:46:07.469]I'm thinking as if you ran that through a Discrete model
[00:46:11.459]and you had relationships that you felt really reflected
[00:46:15.070]the system, then it would tell you where you get
[00:46:19.689]the biggest bang for the buck so it should tell you that,
[00:46:22.447]"Yeah, you have enough resources in this area.
[00:46:25.188]"You're gonna get with this one project
[00:46:27.718]"over in the other county, it's gonna give you more trips."
[00:46:31.214]I guess I'm assuming it's gonna come out in the analysis,
[00:46:36.624]but maybe it wouldn't, maybe you do get to some point
[00:46:38.953]where you have enough land in the area.
[00:46:41.159]I don't know, that's a good question.
[00:46:42.355]I guess I'm assuming it would come out in the analysis.
[00:46:50.241]So as a birder I always wonder
[00:46:52.537]why you don't have more observation towers
[00:46:56.121]or birdwalks, something like that in
[00:46:59.006]the National Wildlife refuges and certain places.
[00:47:01.635]This talk seems to be about delivering habitat.
[00:47:05.065]If you want to bring people in, birders, who are mostly
[00:47:07.978]the people using these refuges, wouldn't you want
[00:47:10.311]to include access?
[00:47:14.986]Yeah, so we've talked about those hypotheses and factors
[00:47:19.044]that have come out.
[00:47:21.089]I recognize that how you actually manage the opportunity
[00:47:24.270]on those public lands by putting in trails
[00:47:27.129]or putting in signage could be really, really important.
[00:47:31.264]I didn't have data on those aspects to test that
[00:47:34.711]with that analyses.
[00:47:37.329]I do think if we pursue these kind of techniques,
[00:47:41.232]if we work with the partners and particularly
[00:47:43.696]land management agencies, the state agencies,
[00:47:46.786]in some cases they're gonna have more information
[00:47:49.052]about what's available, how that land's management,
[00:47:51.889]and you should be able to work those into your analysis.
[00:47:55.173]That would be ideal because I think those
[00:47:56.989]are important factors.
[00:47:58.879]The other way to do it is you can do
[00:48:01.500]this kind of Discrete Choice modeling and this is gonna
[00:48:05.283]refer to Reveal Discrete Choice modeling
[00:48:07.558]because it's what people actually do
[00:48:09.784]whereas you can also just conduct surveys
[00:48:13.714]and ask them what they think, what they feel.
[00:48:17.185]It's data preference stuff so you can get it
[00:48:18.834]at that way too as we can incorporate it
[00:48:21.465]with these data sets.
[00:48:27.911]So it looks like the majority
[00:48:29.535]of the trips most of these hunters and birders
[00:48:31.445]are taking are within 50 kilometers or so.
[00:48:35.153]By providing additional habitat over these very localized
[00:48:38.373]areas within states, do you think you're missing out
[00:48:40.515]on recruiting new hunters and birders?
[00:48:43.255]Missing out on which hunters?
[00:48:44.934]On recruiting new ones.
[00:48:48.088]Because they have to travel throughout the entire state
[00:48:50.174]to get to these hot spots more or less.
[00:48:54.951]I would think probably not and the reason I would say that
[00:48:57.534]is because most of our recruitment of new hunters
[00:49:00.654]is through existing hunters.
[00:49:04.785]Wherever the existing hunters are, they're gonna bring out
[00:49:06.965]the new hunters and they're in these population centers
[00:49:10.654]and they're going pretty close at home
[00:49:12.346]so I suspect we're not missing out on too much of that.
[00:49:21.495]So Pat, in some of the work we've done
[00:49:23.846]and some of the work that other folks have shown,
[00:49:26.566]tradition is a really important factor in picking sites
[00:49:31.229]people go to so I hunters tend to traditionally go
[00:49:34.332]to the same place over and over and over again.
[00:49:37.261]And in your models, I'm wondering if that creates
[00:49:40.010]a potential lag in your response.
[00:49:42.462]So you've got habitat improvement or access improvement
[00:49:45.333]that you're gonna put into place, I would agree
[00:49:47.092]that if you're in a place that people aren't already going,
[00:49:49.721]the initial response is gonna be low,
[00:49:52.023]but I think that there's gonna be a timeline in there.
[00:49:54.931]And once that area is identified, then you quickly
[00:49:58.693]get very rapid response to it and then out of that
[00:50:02.797]becomes a tradition.
[00:50:04.191]So your Finger Lakes example, if the birds are there,
[00:50:06.823]you can increase access and the initial response
[00:50:09.172]I would also predict would not be great
[00:50:12.102]because no one ever goes there to hunt now,
[00:50:14.328]but when their knowledge of the access comes into play,
[00:50:18.113]can you get that timeline?
[00:50:19.773]Is there ways to incorporate that
[00:50:21.616]into the data you're looking at?
[00:50:24.272]So I think there are ways and that's kind of why
[00:50:27.027]I through that little disclaimer in there
[00:50:28.763]about the nested logit model and the hierarchy models.
[00:50:32.521]I have seen examples where people are trying
[00:50:35.691]to model more complex dynamics like that
[00:50:40.804]and also take into account the actual repeated measures
[00:50:45.295]nature of this because you're making the same decision
[00:50:48.021]repeatedly over time and how do you incorporate that
[00:50:50.610]and I suspect if you can work in that repeated
[00:50:55.063]at every decision process you could incorporate
[00:50:59.545]that knowledge into it and that lag.
[00:51:16.199]So we did talk a little bit about that
[00:51:19.038]and the big challenge there is you're gonna have to build
[00:51:24.292]a time series of land cover,
[00:51:27.833]existing changes and open lands, loss of lands,
[00:51:31.609]and that would be a great thing to do.
[00:51:35.377]I can't wait to read that paper. (laughs)
[00:51:39.957]I hope someone does it, it's just well beyond me. (laughs)
[00:51:44.013]This is really just to hopefully encourage
[00:51:45.826]a lot of smart people to say we have it.
[00:51:47.727]That's really it, it really is a kind of joke.
[00:51:52.217]The motivation behind this was to say that we are sitting
[00:51:55.475]on data sets, that we may not have to spend a whole lot
[00:51:59.503]of money to start really addressing these questions
[00:52:02.551]and making better decisions and just exploring
[00:52:05.482]these relationships and that would be one of them, too.
[00:52:09.410]So yeah, there's a lot more
[00:52:10.822]I think that could be done here.
[00:52:14.004]Any more questions?
[00:52:19.110]Alright, well join me in thanking Pat for coming out here
[00:52:21.419]and giving his talk. (audience applauds)
Log in to post comments