Using Decision Tools to Design the Everglades Headwaters NWR

Video Transcript
Download Video
Right-click and save to download

Detailed Description

The U.S. Fish and Wildlife Service’s (FWS) Everglades Headwaters National Wildlife Refuge (EHNWR) is strategically located in Florida to protect upland and wetland habitats. The location of the new refuge was targeted to: address pressures from urbanization and climate change, provide improved quality of water flowing southward to the Greater Everglades, and protect habitat for species of concern (e.g., Florida panther, Florida grasshopper sparrow, Everglades snail kite). Acquiring land for a large protected area such as the EHNWR comes with challenges because it typically takes many years to gather the funds to purchase all necessary parcels of land and could be complicated by future conditions such as climate and urbanization changes across the landscape. To meet the FWS-defined objectives for the EHNWR, including the protection of five target habitats, Southeast CASC supported researchers used Marxan with Zones as a decision tool to select configurations of parcels under different scenarios of urbanization and protection. The designs researchers generated met FWS habitat goals within fee and easement zone restrictions, and they found refuge configurations that fell well below the mandated size limit.


Date Taken:

Length: 00:35:59

Location Taken: Everglades, FL, US

Video Credits

Stephanie Romañach, USGS Ryan Boyles, USGS


Katie Poston:  [0:05] Welcome from the US Fish and  [0:06] Wildlife Services National  [0:08] Conservation Training Center in  [0:09] Shepherdstown, West Virginia.  [0:11] My name is Katie Poston, and I  [0:13] would like to welcome you to our  [0:14] webinar series held in  [0:16] partnership with the US  [0:17] Geological Survey's National  [0:19] Climate Adaptation Science  [0:21] Center. Today's webinar is  [0:23] titled "Using Decision Tools to  [0:25] Design the Everglades Headwaters  [0:27] National Wildlife Refuge."  [0:29] We're excited to have Stephanie  [0:31] RomaÒach with us today. To  [0:33] introduce our presenter today,  [0:35] we have Ryan Boyles, who is the  [0:37] Southeast Climate Adaptation  [0:39] Science Center's Deputy Director.  [0:42] Welcome, Ryan.

Ryan Boyles:  [0:43] Thank you, Katie. Very pleased  [0:45] to introduce to you Dr.  [0:47] Stephanie RomaÒach. She's a  [0:48] Research Ecologist with USGS.  [0:50] She's based in Fort Lauderdale  [0:51] with the Wetland and Aquatic  [0:52] Research Center. She's done a  [0:55] lot of work in central and south  [0:56] Florida working with Everglades,  [0:58] Everglades restoration.  [1:00] She  [1:00] leads a fairly large research  [1:01] program working with decision  [1:03] makers on a lot of the complex  [1:05] problems that they're dealing  [1:06] with in terms of how restoration  [1:09] funding should be allocated.  [1:11] She's accommodation of field  [1:12] observations, ecological models  [1:14] to look at impacts and connect  [1:15] that back into decision making.  [1:16] That's one of the things.  [1:17] While  [1:18] this specific project that she's  [1:20] talking about was funded several  [1:22] years ago, it fits very well  [1:24] with the theme of this webinar  [1:25] series around the science that  [1:27] you can use. She brings decision  [1:29] analytics in and decision  [1:31] science in in combination with  [1:33] the ecological models and field  [1:35] observations that she collects.  [1:38] She's done a lot of work in  [1:40] Everglades in central and south  [1:42] Florida, but also has prior  [1:43] experience working in Africa  [1:45] around conservation issues there.  [1:47] Thank you to Stephanie for  [1:49] coming back to talk about  [1:51] something that she probably  [1:52] hasn't thought a lot about in  [1:53] the past year. I think the topic  [1:55] certainly still resonates with  [1:56] us today, especially the process  [1:58] that you went through to look at  [2:00] the reserve design for  [2:02] Everglades headwaters.  [2:04] With  [2:04] that, we'll turn it over to  [2:05] Stephanie, and say thanks to  [2:07] everyone for joining. Hope you  [2:08] get a lot out of this.

Stephanie RomaÒach:  [2:09] Thank you, Ryan. Thank you,  [2:10] Katie. That is a nice setup to  [2:14] say that yes, this project was  [2:16] funded some time ago. It was  [2:19] completed in 2015, but we only  [2:22] published the work maybe a year  [2:26] or two ago. Year‑and‑a‑half ago,  [2:27] maybe. It doesn't make it any  [2:30] less relevant.  [2:31] This was also  [2:32] one of my favorite projects to  [2:33] work on. It will be fun to talk  [2:35] about it again. My apology would  [2:37] be if you ask me a detail that I  [2:39] just can't recall, because the  [2:41] technical details ‑‑ it  [2:42] definitely has some problems.  [2:45] Like I said, it's just as  [2:46] relevant today, and the work by  [2:48] the Fish and Wildlife Service is  [2:51] ongoing. As Ryan and Katie said,  [2:54] I'm going to be talking about  [2:56] the work that Brad Stith, Fred  [2:58] Johnson, and I did to help the  [3:00] Fish and Wildlife Service with  [3:02] designing the new Everglades  [3:04] Headwaters National Wildlife  [3:05] Refuge.  [3:07] Designing a wildlife  [3:08] refuge is not too dissimilar to  [3:13] buying a house, in some ways,  [3:15] although that might seem odd to  [3:15] say. Yes, it is a little more  [3:18] computationally complex, but the  [3:20] process, in a way, is similar to  [3:23] buying a house or any other big  [3:24] decision that you make.  [3:27] I would  [3:27] guess that if you're buying a  [3:29] house, you don't just get on  [3:30] Zillow and put an offer on the  [3:32] first house that pops up on your  [3:34] search. You would probably think  [3:36] about things that you care about ‑‑  [3:39] quality of life in general, but  [3:40] you'd probably be thinking about  [3:42] quality of the schools your kids  [3:43] would go to, or the layout of  [3:45] the house, or size of the house,  [3:47] or traffic patterns, or whatever  [3:48] it would be.  [3:51] You'd think about  [3:51] what you really want first, and  [3:53] then you'd find a house that  [3:55] matches those objectives that  [3:58] you have. If you're designing a  [4:01] wildlife refuge, chances are,  [4:03] just like if you went in Zillow  [4:05] and put an offer on the first  [4:06] house you found, if a refuge  [4:10] manager just puts an offer on  [4:12] whatever is the first parcel of  [4:13] land that comes up for sale in  [4:15] the area, chances are that  [4:16] that's not going to work out too  [4:17] well.  [4:18] The refuge manager will  [4:19] have to consider his or her  [4:22] objectives for the refuge when  [4:23] deciding which parcels of land  [4:25] to buy. I'll get back to all  [4:27] those details about details  [4:28] about objectives and that in a  [4:29] minute.  [4:31] I'm getting used to the  [4:32] slide‑changer. I'll tell you a  [4:34] little bit about reserve design  [4:36] in general first, and about the  [4:39] refuge, and then we'll get into  [4:40] more details. I realize that we  [4:43] are talking about a refuge, and  [4:45] I'm calling this "reserve design,"  [4:47] and that's because the field of  [4:49] study is generally called  [4:51] reserve design.  [4:53] It deals with  [4:53] identifying parcels of land that  [4:55] are in need of protection for  [4:57] whatever reason ‑‑ to protect  [4:59] particular wildlife species or  [5:01] populations or habitats of  [5:03] interest, ecosystem functions or  [5:04] whatever it might be. That's  [5:07] done in another sense just  [5:09] [laughs] like buying a house,  [5:10] where it's secured on an open  [5:11] market. You do your purchase.  [5:14] Some land can be put into  [5:15] easements.  [5:16] There are challenges.  [5:19] Usually, funding is not  [5:22] available upfront to buy all of  [5:25] the land necessary for an entire  [5:27] refuge or any other kind of  [5:29] protected area. What that means  [5:31] is that results in incremental  [5:33] implementation of that refuge or  [5:38] other kind of protected area.  [5:42] What that leads to is because we  [5:44] have to do this incrementally,  [5:46] it means these purchases are  [5:47] made over many years and in many  [5:49] cases, made over many decades.  [5:52] That subjects the area to  [5:54] changing condition.  [5:55] You might  [5:56] have properties that are lost to  [5:58] residential or commercial  [5:59] development. You don't know  [6:01] where you're going to have  [6:02] funding in 10 years from now in  [6:05] this case of Fish and Wildlife  [6:06] Service and what is congress is  [6:09] going to authorize. What are the  [6:10] funds can be in 10 years from  [6:12] now or even next year? We just  [6:13] don't know.  [6:15] Just like with any  [6:17] real estate transaction, what's  [6:18] the probability of a successful  [6:20] purchase or successful land  [6:22] transaction? As we know, there  [6:26] are a lot of environmental  [6:27] change. We have no idea what a  [6:31] particular price might look like,  [6:33] depending on management or  [6:37] nature or other factors. Things  [6:39] could change. Whatever we think  [6:41] might make a good refuge, now  [6:43] that might be quite different in  [6:44] 10 years from now.  [6:46] Everglades.  [6:47] Just getting into the study area  [6:49] a little bit. If you have worked  [6:51] in the Everglades or you're  [6:52] familiar with the Everglades at  [6:53] all, chances are you have seen  [6:55] this figure a million times or  [6:58] something like it. If you look  [7:00] over at the left panel, this is  [7:03] what the Everglades...  [7:05] When I  [7:05] say Everglades, Everglades  [7:06] National Park is down here. This  [7:08] is the greater Everglades. This  [7:10] is typically when we're in the  [7:12] Everglade, this is what we're  [7:13] talking about when we say the  [7:14] Everglades.  [7:16] This figure is what  [7:17] the Everglades look like in the  [7:19] 1800. We had this area was the  [7:23] headwaters of the Everglades.  [7:25] I'll show you that more closely.  [7:28] That basically form this really  [7:30] nice ship flow that moved fast  [7:32] over the landscape.  [7:34] Then,  [7:36] humans came in and started all  [7:38] of these mines here. People  [7:40] started digging, so they were  [7:41] draining this ecosystem. These  [7:44] lines are levees and canals and  [7:47] other features that basically  [7:48] move water away for flood  [7:51] control or in this case,  [7:52] agriculture or whatever human  [7:54] purposes were. The system looks  [7:56] very different today than it did  [7:58] in the 1800s.  [8:00] The idea is that  [8:02] we would like with restoration,  [8:04] which will take many decades. We  [8:06] would like water to come from  [8:09] this headwaters and move south  [8:11] through the ecosystem once again.  [8:16] Why is this area important? This  [8:19] is the area that we're going to  [8:19] be focusing in on and this is  [8:22] the headwaters of the Everglades.  [8:26] It's important because in terms  [8:28] of the Fish and Wildlife Service  [8:30] desiring to have a refuge in  [8:31] this location, the headwaters of  [8:34] the Everglades, this can help to  [8:37] restore the wetlands and uplands  [8:40] in this region.  [8:42] Some of which  [8:43] are degraded from human activity.  [8:45] There's a lot of restoration  [8:46] going on in that area. If they  [8:48] already in particular can  [8:51] protect a number of federally  [8:53] listed threatened and endangered  [8:54] species, 38 specifically in 161  [8:57] state‑listed species.  [8:59] It also  [9:00] connects habitats, which you'll  [9:01] see in a map that will show you  [9:04] the refuge study area. Connect  [9:06] them to other protected areas in  [9:08] the region. Some are state parks  [9:11] or military lands in better  [9:13] under one protection or another.  [9:16] Another idea was that this would  [9:17] be a potential refuge for many  [9:19] species from impacts of climate  [9:21] change.  [9:25] Zooming in a little bit,  [9:27] maybe I can reorient you. Here  [9:28] was Lake Okeechobee. This was  [9:30] the Kissimmee River coming down.  [9:32] This is the headwaters of the  [9:33] Everglades. Here's the edge of  [9:37] Lake Okeechobee here to reorient  [9:39] you here. This is the Everglades  [9:41] headwaters.  [9:43] In black is  [9:45] bounding the study area, which  [9:49] is a 1.8 million acre area. The  [9:53] reserve will not be continuous.  [9:56] What this means is the Fish and  [9:57] Wildlife Service can select  [9:59] parcels of land from within this  [10:00] study area.  [10:02] There are two  [10:03] constraints. One is that they  [10:05] can buy up to 50,000 acres  [10:09] outright and then they can put  [10:13] easements on another 100,000  [10:15] acres. It'll be a 150,000‑acre  [10:17] refuge when it's done. The good  [10:19] news is that there are willing  [10:20] sellers. This was not the case  [10:22] at the start.  [10:23] There was a  [10:23] little bit of a turbulent start  [10:26] along the lines of the  [10:28] government coming in to take  [10:30] over people's property. That has  [10:32] subsided. There are willing  [10:34] sellers now, so that's not a  [10:36] problem. Some of the desired  [10:42] outcomes for the Fish and  [10:43] Wildlife Service in the  [10:44] conservation‑partnership area,  [10:47] and again that's where the  [10:48] easements can be, are to  [10:51] basically keep working lands  [10:52] working.  [10:53] Allow ranching  [10:54] activities to persist in private  [10:56] ownership, again if they're  [10:59] reflecting the goals of the  [11:01] refuge, which are things like  [11:02] wildlife‑viewing, hunting, and  [11:04] fishing. Then for the  [11:07] conservation focal area, which  [11:08] is where the lands will be  [11:11] purchased, some of the desired  [11:15] outcomes there are to have  [11:16] habitats for the threatened and  [11:17] endangered species.  [11:19] Inland, to  [11:20] keep common species common. They  [11:24] have an emphasis on migratory  [11:26] birds, listed species, and  [11:28] wetlands in particular. The Fish  [11:29] and Wildlife Service had thought  [11:33] through what were their  [11:34] priorities. These were some of  [11:37] their top priorities, which were  [11:39] for land acquisition. That was  [11:41] to have landscape connectivity  [11:44] or corridors through the  [11:45] habitats in the region.  [11:46] Again,  [11:47] habitats for threatened and  [11:48] endangered species. Restoration  [11:51] of wetlands and water quality  [11:55] was very important. You can hear  [11:56] that in the name of the  [11:57] Everglades Headwaters. They  [11:59] wanted opportunities for  [12:00] wildlife‑dependent recreation  [12:02] and educational opportunities.  [12:06] Some of the species of  [12:09] particular interest are the  [12:13] panther.  [12:14] You've probably heard  [12:15] about the Florida grasshopper  [12:16] sparrow. I know at one point it  [12:18] was considered the most  [12:19] endangered bird in North America.  [12:21] Things have not gotten any  [12:22] better for it. Unless another  [12:24] species has gotten worse, this  [12:25] one was considered to be the  [12:28] most endangered in North America.  [12:30] There are a number of other  [12:31] birds, mammals, and some  [12:33] habitats in need of protection ‑‑  [12:36] Florida dry prairie, scrub  [12:38] habitat, and sandhill. The Fish  [12:43] and Wildlife Service was looking  [12:45] at land‑protection options in  [12:48] order. The ideal would be if  [12:51] others would protect the land so  [12:53] that they wouldn't have to buy  [12:56] it or use the resources in those  [12:58] particular areas. The second  [13:01] option would be easements.  [13:04] Then  [13:05] the third, then they would go to  [13:07] a fee title or buying land  [13:09] outright. The focus there would  [13:12] really be for things like if  [13:14] they wanted to build boardwalks  [13:16] or whatever other visitor‑use‑  [13:21] type areas that really wouldn't  [13:24] be as suitable to land with an  [13:27] easement. Just a little bit  [13:31] about cost. It's not cheap to  [13:36] design [laughs] and to take over  [13:37] a refuge.  [13:39] If we look at the  [13:40] conservation focal area, which  [13:41] is where the land is going to be  [13:42] bought outright, the estimated  [13:45] cost is about $4,000 per acre.  [13:49] For the 50,000 acres that they  [13:51] can buy minus some expected  [13:53] donations, it's just under 200  [13:55] million.  [13:57] If we look at the  [13:58] easements, that's estimated to  [14:00] be at about 2,000 per acre times  [14:02] the hundred thousand. That's  [14:03] another two million. Just for  [14:06] the land is just under $400  [14:09] million, plus some operations  [14:13] costs, both one‑time and annual  [14:15] costs.  [14:15] Just to buy the land  [14:17] necessary for the refuge is  [14:18] about 398 million. You can see  [14:23] why I was saying that we usually  [14:25] don't have that kind of money  [14:27] all at once to just say, "OK,  [14:29] here's the best design for the  [14:30] refuge. We're going to buy all  [14:31] these lands." It's just not  [14:33] going to happen.  [14:37] The Fish and  [14:38] Wildlife Service did all of the  [14:40] hard work for us. It can be  [14:42] really challenging to really get  [14:48] at objectives, and then  [14:49] quantifying them is another  [14:51] challenge, but the Fish and  [14:52] Wildlife Service did this quite  [14:54] nicely.  [14:55] They were very clear on  [14:57] exactly how many acres of each  [14:59] habitat they wanted ‑‑ for  [15:01] example, 34,000 acres of wetland  [15:03] for particular species of  [15:06] interest, 13,000 acres of dry  [15:07] prairie. We had their objectives  [15:12] to work with while we were  [15:14] working on this refuge design.  [15:21] We ended up choosing to work  [15:24] with Marxan with Zones.  [15:34] Basically what it allowed us to  [15:35] do...It goes a little bit beyond  [15:36] Marxan to help us find an  [15:39] optimal reserve design. What we  [15:46] considered as zones were  [15:48] easement and fee titles. We were  [15:50] able to allocate land in those  [15:52] two categories to help us meet  [15:54] the overall habitat objective.  [15:57] Marxan was originally designed  [15:59] in the marine world, in  [16:01] Australia. I think it was  [16:03] University of Queensland. It's  [16:06] been widely used all over the  [16:07] world on land ‑‑ terrestrial  [16:10] uses as well as marine, at this  [16:11] point. It's been in use for  [16:12] quite some time. It's a pretty  [16:14] strong program.  [16:17] We used that  [16:18] because for me, I can't do trade‑  [16:21] offs of calculating 50,000  [16:23] parcels in my head, so we had to  [16:26] find some way to do it, and  [16:27] Marxan with Zones was a great  [16:28] way for us to be able to do that.  [16:31] We were also able to look at the  [16:33] target habitat within a parcel,  [16:36] so not just the whole parcel,  [16:37] but making sure that we were  [16:39] able to get the habitat within a  [16:41] parcel while we did these  [16:42] analyses.  [16:44] We also looked at two  [16:45] urbanization scenarios, and  [16:46] those urbanization scenarios  [16:48] went out to 2060. I'll talk a  [16:52] little bit more about those  [16:53] later. What Marxan does is it  [16:57] allows us to have a static  [17:02] solution to the problem. This  [17:04] assumes we have that 400 million  [17:05] today, and we don't. I'll talk  [17:08] about that a little bit later,  [17:10] but this was the best we could  [17:13] do for this case.  [17:15] There are a  [17:16] number of other methods that  [17:17] allow us to look a little bit  [17:18] more dynamically, but not for 50,  [17:21] 000 parcels. That method did not  [17:24] exist.  [17:27] I'll just show you the  [17:27] inputs that we used, which is  [17:29] pretty simple and  [17:30] straightforward. These were the  [17:32] five inputs we used, which were  [17:34] the parcel boundaries and cost,  [17:37] habitat...I don't know if you  [17:40] remember if I said, this was  [17:41] funded in 2012. That's where the  [17:44] numbers...You'll see these years  [17:46] of these data layers come from  [17:48] that time period.  [17:52] We did the  [17:52] work from 2012 to 2015. The  [17:55] parcel boundaries, the costs,  [17:57] and the cost we used was just  [17:58] value ‑‑ the land cover,  [18:02] existing protected areas.  [18:04] I  [18:06] think I forgot to point this out  [18:07] to you on the map ‑‑ areas not  [18:08] considered. The Fish and  [18:10] Wildlife Service...There were  [18:11] particular parcels of land that  [18:13] they knew they would not buy or  [18:16] put easements on, and so those  [18:18] they had designated as areas not  [18:20] considered. We factored that in  [18:21] to make sure we weren't  [18:22] selecting parcels in that area.  [18:26] Then, of course, we used the  [18:29] focal area boundary where we  [18:30] could buy land, and then we  [18:31] can't buy land outside of that,  [18:33] and so easements on the  [18:34] partnership area.  [18:38] This is where  [18:39] we ended up. We had four  [18:41] counties, and to start, we had  [18:47] 200,000 parcels. We removed  [18:51] those in the areas not  [18:51] considered, removed those that  [18:53] are already protected, and we  [18:53] ended up with 51,000, almost 52,  [18:56] 000 parcels to consider in the  [18:59] reserve design.  [19:02] The two  [19:04] urbanization layers that we used  [19:07] have a lot of the same  [19:09] assumptions, but there are some  [19:11] differences. There was one layer  [19:14] developed by a group called  [19:16] Geodesign, for anyone who's  [19:19] worked with Mike Flaxman. He  [19:23] used to be at MIT. He's done a  [19:25] lot of work for the Fish and  [19:27] Wildlife Service in the past.  [19:29] He did a layer for us, and it  [19:33] used a 70‑year median growth  [19:34] rate. He incorporated some local  [19:38] county regulations, and he  [19:40] develops an attractiveness index  [19:43] to look at where residential  [19:47] development might be occurring  [19:49] in the future.  [19:52] We also used a  [19:53] University of Florida  [19:55] urbanization layer. It had a lot  [19:56] of the same assumptions. It also  [19:59] used distance from existing  [20:01] development ‑‑ development based  [20:05] on projected population growth,  [20:07] and major roads that are planned.  [20:12] We ended up with nine scenarios.  [20:14] We basically had a current,  [20:16] assuming‑no‑further‑urbanization  [20:18] scenario, the Geodesign scenario,  [20:23] and the University of Florida  [20:24] scenario, and then we had three  [20:26] land allocation scenarios. If we  [20:28] had 10 percent of land in the  [20:31] fee zones and 90 percent in  [20:32] easement, 33 percent in the fee  [20:34] zone, 67 in the easement, and  [20:36] then 50‑50 split.  [20:38] We also  [20:38] looked at connectivity. We  [20:41] explored a range of values for  [20:42] the boundary length modifier, or  [20:44] the boundary cost matrix. It  [20:46] really had no influence, which  [20:48] suggested that our design was as  [20:51] compact and efficient as it  [20:53] could be.  [20:58] Just so you're aware,  [21:01] and we should have said this at  [21:02] the start, but I'm not going to  [21:03] show you through a map of our  [21:06] final design. I'll show you  [21:07] plenty of other results, but I'm  [21:09] not going to show you that, just  [21:10] because of the sensitivities of  [21:11] land acquisition. You won't see  [21:13] a final map, but you'll see some  [21:15] other grasping tables and maps  [21:18] of other parts that went into  [21:20] the design.  [21:23] This ‑‑ I just had  [21:24] to put up here the calibration,  [21:25] because I thought it was quite  [21:26] crazy. As Ryan mentioned at the  [21:29] introduction, I run a team,  [21:31] ecological modeling team. Marxan  [21:35] was new to me. That's why Fred  [21:36] and I brought in Brad Stith,  [21:38] who's worked quite a bit with  [21:38] Marxan. I'd just never seen a  [21:40] hundred trillion iterations per  [21:42] repetition [laughs] before. I  [21:44] thought that was pretty neat. I  [21:46] just thought I'd bring that up.  [21:47] It was a little crazy.  [21:50] I will  [21:50] jump into some of the results,  [21:53] and I'll first show you the  [21:54] urbanization results. Let's see.  [22:01] This is the study area, again.  [22:03] The gray are the areas that are  [22:06] existing protected areas. These  [22:10] polygons in the dark outline are  [22:12] areas where purchases can be  [22:14] made, and then the rest of this  [22:16] area is where the easements can  [22:18] be made.  [22:19] Green is showing you  [22:22] habitat that is projected to not  [22:26] become urbanized by 2060. This  [22:30] orange is showing the habitat  [22:34] that the University of Florida  [22:35] model is showing to be urbanized.  [22:37] The purple is where the  [22:41] Geodesign layer is showing to be  [22:44] urbanized, and then red is where  [22:46] they both are showing.  [22:47] There's  [22:47] most agreement between the two  [22:49] models up here in the northeast,  [22:52] and we used two because we just  [22:55] wanted a broader look. Like I  [22:57] said, they do have a lot of the  [22:58] same assumptions, but they have  [22:59] differences. We just wanted to  [23:01] be able to see, what are these  [23:03] different approaches? What might  [23:05] they be telling us about where  [23:07] we might want to focus attention  [23:10] or not?  [23:11] That's another  [23:12] conversation and something I'll  [23:13] bring up later. The good news is  [23:17] that the majority of habitats  [23:19] you can see a lot of green. The  [23:20] majority of them are not  [23:21] forecast to be developed by 2060,  [23:23] so there'll be a lot to choose  [23:24] from.  [23:25] We did see greater  [23:27] habitat loss with the Geodesign  [23:30] model, the purple compared to  [23:33] the orange ‑‑ the 31 percent  [23:35] habitat loss overall with the  [23:37] Geodesign model compared to 20  [23:38] percent for the University of  [23:39] Florida model. This scrub and  [23:44] sandhill, these xeric habitats,  [23:48] had the largest projected losses ‑‑  [23:51] from 40 to 60 percent, which is  [23:56] quite big.  [24:00] Let me orient you on  [24:01] this figure a little bit.  [24:02] There's a lot of information  [24:04] here. Here on the y axis, you  [24:06] see area in hectares, and here  [24:09] are the habitats we considered.  [24:13] What this is showing you is, in  [24:17] the fee zones, if we look at  [24:18] just the area where we might  [24:20] make purchases, if we wanted 50  [24:24] percent of the refuge to be in  [24:27] the fee area, versus 33 percent  [24:32] or 10 percent, and this is  [24:35] showing you 100 percent, just  [24:36] for reference.  [24:38] Really, this  [24:41] should be a bar across the top  [24:42] to show you this is the target ‑‑  [24:44] that's what we're aiming for ‑‑  [24:45] but because that target is  [24:46] different for each habitat type,  [24:49] it's just represented here as  [24:50] another bar, but that way you  [24:51] can see what we were shooting  [24:52] for.  [24:55] These show the future  [24:59] scenarios. In this case, current  [25:03] urbanization versus the  [25:05] University of Florida versus the  [25:06] GeoAdaptive. Really, the bottom  [25:09] line here is that we were able  [25:10] to design a refuge that met the  [25:13] Fish and Wildlife Service  [25:14] objectives for all of the  [25:16] habitat types in all of the  [25:17] scenarios, except for xeric.  [25:20] This is the one that we had the  [25:21] most trouble with.  [25:24] I'll show  [25:25] you an image here to give you an  [25:28] idea. This is, in our modeling  [25:31] world, on the left, and on the  [25:32] right is a satellite image just  [25:35] to show you. You can see all  [25:36] these tiny little blue parcels.  [25:41] This is what we had to work with  [25:44] for xeric.  [25:45] There was a point  [25:46] where we met with the Fish and  [25:50] Wildlife Service to talk about  [25:52] the outputs as we were partway  [25:54] through the process and refining  [25:56] what we were doing to help. We  [25:59] had made the decision that we  [26:01] weren't going to consider any  [26:02] parcels that were under 100  [26:05] acres. What that meant was, all  [26:09] of these little parcels that  [26:11] have xeric habitat couldn't be  [26:13] considered.  [26:15] That's something  [26:16] that, as the Fish and Wildlife  [26:17] Service is moving forward with  [26:19] the refuge design that they'll  [26:21] need to think about ‑‑ whether  [26:22] that 100,000‑acre cutoff is  [26:26] across the board a good idea.  [26:29] For a lot of reasons, it is ‑‑ I  [26:31] mean, this would be a really  [26:33] difficult habitat to manage ‑‑  [26:34] but just to show you the  [26:36] struggle that we were having  [26:37] with that xeric habitat.  [26:40] Most  [26:40] of the area where the xeric  [26:41] habitat is is already heavily  [26:43] urbanized, so it just makes it a  [26:45] real challenge. This is similar  [26:48] to the other figure that I  [26:51] showed you ‑‑ same setup, so  [26:56] trying to get 50 percent of dry  [26:58] prairie in the easement zone,  [27:00] versus 67 percent, versus 90.  [27:02] This is just in the easement  [27:03] zone. Again, we really only  [27:05] struggled to get the right  [27:08] amount of xeric habitat.  [27:10] Just  [27:11] to show you a little bit about  [27:13] cost, this is kind of obvious.  [27:19] Well, let me tell you what  [27:20] you're looking at [laughs]  [27:21] before I tell you how obvious it  [27:22] is. Here's the cost on the y  [27:24] axis. This set of columns is  [27:26] showing the scenario where we  [27:29] would buy outright 10 percent of  [27:30] the land, and then have  [27:32] easements on 90 percent.  [27:35] Then  [27:35] we move up to a 33/67 split, and  [27:37] then 50/50. The more land that  [27:42] we're buying outright, so 10  [27:43] percent versus 33 versus 50 ‑‑  [27:45] as that amount we buy goes up,  [27:47] it's going to cost more money  [27:48] compared to an easement, and  [27:51] then shows us with the current  [27:56] urbanization, so if no further  [27:58] urbanization versus the UF model  [28:00] versus the GeoAdaptive model.  [28:03] We always had a higher cost with  [28:06] the habitat that was projected  [28:08] to be left as available with  [28:11] using the GeoAdaptive model.  [28:13] We'd have to do a lot more  [28:14] digging into those models and  [28:18] understand all of the  [28:22] assumptions that go into the  [28:23] model and those projections to  [28:25] see why it's coming up that way.  [28:29] Again, this is why we use two,  [28:30] just to get a little bit of a  [28:31] broader look at a couple of  [28:33] possibilities of how things  [28:34] might play out in the future. In  [28:39] terms of the design, here again,  [28:44] we have area on the y axis. The  [28:49] limit for the reserve ‑‑ if  [28:50] you'll remember, it was 50,000  [28:52] acres in the fee simple and 100,  [28:56] 000 acres in easement. That is  [29:00] 60,000 hectares, which is up  [29:01] here.  [29:02] You can see that in every  [29:04] way that we configured it, we  [29:05] were able to get the reserve to  [29:07] be smaller than what the refuge  [29:10] was aiming for.  [29:13] That's good  [29:13] news, because we're able to meet  [29:16] the habitat targets that they  [29:17] have in a smaller area, which  [29:19] means that you spend less money  [29:21] on it. That could be a really  [29:23] good thing if it we're able to  [29:24] meet those objectives for the  [29:25] species and habitats and  [29:27] recreational and all those other  [29:28] objectives that they have. We  [29:30] were able to do that in every  [29:32] scenario.  [29:34] Again, this is the  [29:35] splits of 10 percent fee, 90  [29:38] percent easements under the  [29:40] current conditions, what's  [29:42] projected with GeoAdaptive and  [29:43] the University of Florida  [29:45] projections, and so no matter  [29:46] how you slice it, we were able  [29:48] to meet the habitat objectives  [29:51] well below the limit that the  [29:54] Fish and Wildlife Service had  [29:56] authorized.  [29:57] This is something I  [29:58] want to show you just because I  [30:00] like it. This is selection  [30:04] frequency that comes out of the  [30:06] Marxan output.  [30:09] One way that I  [30:11] like to look at this is as a  [30:13] measure of irreplaceability.  [30:17] When there are parcels that are  [30:18] coming up over and over again,  [30:22] in every design, every scenario  [30:24] coming up, all the time, these  [30:26] darker parcels, that probably  [30:28] tells you something ‑‑ that  [30:29] these should be high‑priority  [30:31] areas when they do come  [30:33] available for purchase.  [30:34] That's  [30:34] a really nice output we got.  [30:36] That was something that the Fish  [30:38] and Wildlife Service really  [30:38] liked as well.  [30:40] I mentioned that  [30:42] our approach was a static one.  [30:44] We did this design assuming that  [30:47] we have $400 million today and  [30:50] we don't. Fred Johnson, one of  [30:55] my collaborators ended up  [30:58] bringing in a postdoc while we  [31:00] were doing this work, who was a  [31:01] mathematician. His name is  [31:04] Matthew Bono and he was tasked  [31:07] with figuring out a dynamic  [31:09] approach.  [31:11] Just for  [31:12] entertainment value, I should  [31:13] have put some of his equations  [31:15] up. You can look at the paper  [31:16] and see for yourself, but it's  [31:18] quite something. I could usually  [31:19] follow him through 75 percent of  [31:21] what he was saying and then he  [31:22] would loss me. It's pretty  [31:23] intense. He was able to pull  [31:27] this off even with the amount of  [31:32] parcels that were looking at.  [31:34] This paper was published last  [31:35] year in PLOS One if anyone's  [31:37] interested. You can look it up.  [31:44] We know that parcels are  [31:45] subjected to all these threats  [31:46] that we talked about earlier  [31:47] like urban development, habitat  [31:48] degradation, climate change. He  [31:50] was able to come up with a way  [31:53] that accounts for the conversion  [31:55] rate of these habitats through  [31:56] time. It's done on an annual  [31:58] time period, so it gets that  [32:00] that dynamic aspect that I was  [32:02] talking about.  [32:06] You can see this  [32:07] trade‑off of buying parcels that  [32:11] have high ecological value. Yet,  [32:15] the highways threatened parcels,  [32:16] threatened to change accounting  [32:19] for the conversion from one of  [32:21] these factors. It's really neat  [32:24] that he was able to do that.  [32:29] It  [32:30] took a while for him to develop  [32:31] this. He finished it after the  [32:33] project had ended. Even if he  [32:35] had finished it during the  [32:36] lifespan of the project, we  [32:37] still would have needed that  [32:38] information to go into. We would  [32:41] need the information on even if  [32:45] it's expert opinion or best  [32:47] judgment on probabilities of  [32:50] those changes of habitat  [32:51] degradation and climate change.  [32:54] Those conversion rates.  [32:56] That  [32:56] would have taken more time in  [32:59] another step. I'd be interested  [33:00] to know if this method has been  [33:03] used since Matthew was able to  [33:05] put it out in the literature. I  [33:08] don't know. It'd be really neat.  [33:12] Just to summarize some of the  [33:18] takeaways for us in working with  [33:20] the Fish and Wildlife Service on  [33:21] the things that were really  [33:25] valuable. Again, we were able to  [33:28] get those designs well below the  [33:32] limit that they have authorized.  [33:36] We were also able to get the  [33:38] cost down. Compared to the 400  [33:42] million that was estimated, our  [33:45] maximum cost was 138 million. I  [33:49] will note again that we are  [33:51] using just value. There are  [33:55] problems with that type of  [33:56] database where someone can  [33:59] record the value of their  [34:02] several hundred acre ranch a  [34:04] dollar to sell to their cousin  [34:07] or whatever it might be.  [34:08] The  [34:08] values might not be great, but  [34:10] this was all we could find in  [34:13] the records. If we had more  [34:17] accurate values for the land  [34:20] than for the parcels, that would  [34:22] help us to know are we really  [34:24] getting closer to this 400  [34:26] million or how much closer would  [34:27] we be getting to that?  [34:30] Again, I  [34:30] mentioned the selection  [34:31] frequency which the Fish and  [34:33] Wildlife Service really liked in  [34:36] thinking about how to prioritize  [34:40] land as it's coming available.  [34:44] As I said they're going to have  [34:46] to think about what to do for  [34:48] the xeric habitats where they're  [34:52] smaller in urbanized areas, how  [34:54] do you manage that? Can you burn  [34:57] those habitats with all those  [34:58] houses? There's so much to think  [35:00] about for that xeric, meeting  [35:02] the xeric targets that they have.  [35:06] Really, the question of whether  [35:11] to avoid land that's projected  [35:14] to be urbanized or not. You  [35:16] could argue that you should jump  [35:17] in and protect the land that  [35:19] should be urbanized or avoid  [35:22] that conflict and go for the  [35:23] land that have that high quality  [35:25] target habitat that isn't as  [35:27] likely to be surrounded by  [35:29] condos and roads and strip malls  [35:33] and whatever else it might be.  [35:35] I think that's an interesting  [35:36] debate in their definitely  [35:37] strong views on both sides of  [35:39] that. That's it. That's it for  [35:43] me.

Katie:  [35:43] I just want to say a big thank  [35:45] you to our presenter Stephanie  [35:46] today. Thank you all for tuning  [35:49] in. So much, Stephanie. Thank  [35:50] you, Ryan.

Stephanie:  [35:52] Thank you.

Katie:  [35:53] Thank you.