1 00:00:00,280 --> 00:00:02,791 2 00:00:04,000 --> 00:00:05,335 Well, thank you very much. 3 00:00:05,335 --> 00:00:08,680 (audience applaud) 4 00:00:08,680 --> 00:00:10,420 -Thank you. -(audience laugh) 5 00:00:10,420 --> 00:00:13,390 Salt Marsh Sparrows are gonna be extinct in 40 years. 6 00:00:13,390 --> 00:00:15,190 That is the conclusion of this giant 7 00:00:15,190 --> 00:00:17,020 extinction threshold model that we've built 8 00:00:17,020 --> 00:00:19,990 using everything we know about the biology of the species, 9 00:00:19,990 --> 00:00:22,570 incorporating climate change, title projection models, 10 00:00:22,570 --> 00:00:24,670 and sea-level rise. 11 00:00:24,670 --> 00:00:26,530 Salt marshes are gonna go underwater, 12 00:00:26,530 --> 00:00:27,520 and in a couple of decades, 13 00:00:27,520 --> 00:00:30,130 the reproduction for the species is going to be impossible. 14 00:00:30,130 --> 00:00:33,220 Now salt marshes occupy this linear strip of habitat 15 00:00:33,220 --> 00:00:35,860 between the ocean and really expensive 16 00:00:35,860 --> 00:00:38,260 upland real estate and infrastructure. 17 00:00:38,260 --> 00:00:40,840 As sea level rises, salt marshes can't simply 18 00:00:40,840 --> 00:00:42,250 move up into the environment, 19 00:00:42,250 --> 00:00:45,040 they're physically blocked from doing so. 20 00:00:45,040 --> 00:00:47,710 Organisms that occur in the spruce-fir zone 21 00:00:47,710 --> 00:00:48,850 at the top of our mountains 22 00:00:48,850 --> 00:00:51,040 are in that same proverbial boat. 23 00:00:51,040 --> 00:00:52,690 They're already at the top of the mountains, 24 00:00:52,690 --> 00:00:54,190 they can't retreat any further 25 00:00:54,190 --> 00:00:56,200 as climate change or competitive pressures 26 00:00:56,200 --> 00:00:59,080 from other species below them pushes them up. 27 00:00:59,080 --> 00:01:01,570 And as they do so, because of the shape of mountains, 28 00:01:01,570 --> 00:01:03,340 the amount of habitat as they're squeezed 29 00:01:03,340 --> 00:01:05,950 to the top exponentially decreases. 30 00:01:05,950 --> 00:01:09,070 Not surprisingly, we have evidence from around the world 31 00:01:09,070 --> 00:01:12,670 from plants to butterflies, from mount birds, 32 00:01:12,670 --> 00:01:16,270 that organisms are dispersing further up-slope 33 00:01:16,270 --> 00:01:19,030 and towards the poles to retreat from climate change. 34 00:01:19,030 --> 00:01:20,590 Climate is warming in the mountains 35 00:01:20,590 --> 00:01:24,070 at a rate of two to five, five times faster 36 00:01:24,070 --> 00:01:25,960 than the average global background rate, 37 00:01:25,960 --> 00:01:27,790 faster than anywhere else on the planet, 38 00:01:27,790 --> 00:01:29,743 except for the poles. 39 00:01:31,780 --> 00:01:33,400 At this point, we have multiple independent 40 00:01:33,400 --> 00:01:36,040 modeling sets that suggest we are likely to lose 41 00:01:36,040 --> 00:01:39,940 maybe 50% of our existing spruce-fir forest here 42 00:01:39,940 --> 00:01:42,490 over the next 2-300 years. 43 00:01:42,490 --> 00:01:45,580 Not surprisingly, our spruce-fir bird specialists 44 00:01:45,580 --> 00:01:48,670 are the most vulnerable group to extinction 45 00:01:48,670 --> 00:01:52,360 from climate change of any group of birds in North America. 46 00:01:52,360 --> 00:01:54,154 Of our four dozen-some species, 47 00:01:54,154 --> 00:01:56,710 under moderate climate-change scenarios, 48 00:01:56,710 --> 00:01:59,680 those species are likely to entirely become 49 00:01:59,680 --> 00:02:03,010 Canadian endemic species by the end of this century. 50 00:02:03,010 --> 00:02:05,110 Those are your crossbills, and your grosbeaks, 51 00:02:05,110 --> 00:02:06,430 and your white throated sparrows, 52 00:02:06,430 --> 00:02:08,860 and yellow-bellied flycatchers, and spruce grouse, 53 00:02:08,860 --> 00:02:12,130 they will all be relegated to Canadian endemic species 54 00:02:12,130 --> 00:02:14,860 under moderate climate-change scenarios. 55 00:02:14,860 --> 00:02:16,990 Here you see the breeding range 56 00:02:16,990 --> 00:02:20,080 for black warblers in red and yellow. 57 00:02:20,080 --> 00:02:21,830 Anywhere you see red is where these 58 00:02:22,810 --> 00:02:24,403 climate projection models suggest 59 00:02:24,403 --> 00:02:28,393 that that black warbler breeding range will be lost. 60 00:02:30,100 --> 00:02:30,990 So... 61 00:02:33,250 --> 00:02:34,240 these are available 62 00:02:34,240 --> 00:02:37,300 on the Audubon Survival by Degrees website. 63 00:02:37,300 --> 00:02:41,800 So we don't have a great understanding of the... 64 00:02:41,800 --> 00:02:43,420 Oh, I should say it's really important 65 00:02:43,420 --> 00:02:46,600 that we've begun moving forward into this century, 66 00:02:46,600 --> 00:02:49,330 it's likely that even if these models are wrong, 67 00:02:49,330 --> 00:02:51,010 that these populations of boreal birds 68 00:02:51,010 --> 00:02:52,690 are going to be further reduced. 69 00:02:52,690 --> 00:02:54,520 They're likely gonna be more fragmented 70 00:02:54,520 --> 00:02:56,680 and isolated as they already are. 71 00:02:56,680 --> 00:02:58,030 It's going to be extremely important, 72 00:02:58,030 --> 00:02:59,560 from a management perspective, 73 00:02:59,560 --> 00:03:01,870 that we're able to understand the population dynamics 74 00:03:01,870 --> 00:03:03,700 of the species and what's driving them, 75 00:03:03,700 --> 00:03:05,650 in addition to climate change, of course, 76 00:03:05,650 --> 00:03:09,310 and food availability and abundance certainly comes to mind. 77 00:03:09,310 --> 00:03:11,660 I'm sure you've heard of the insect apocalypse. 78 00:03:12,670 --> 00:03:14,410 We are still debating and figuring out 79 00:03:14,410 --> 00:03:17,650 what groups of invertebrates are declining or increasing, 80 00:03:17,650 --> 00:03:21,070 where and when and how severe those changes are happening. 81 00:03:21,070 --> 00:03:23,320 But I think it's very likely that the same forces 82 00:03:23,320 --> 00:03:25,843 that are driving global declines in vertebrates, 83 00:03:26,920 --> 00:03:29,560 industrial agricultural practices, climate change, 84 00:03:29,560 --> 00:03:33,280 habitat loss and degradation, tropical forest loss. 85 00:03:33,280 --> 00:03:34,720 Those same forces that are driving 86 00:03:34,720 --> 00:03:37,360 vertebrate population declines and range reductions 87 00:03:37,360 --> 00:03:40,138 are also likely influencing invertebrate populations 88 00:03:40,138 --> 00:03:41,950 in the same way. 89 00:03:41,950 --> 00:03:44,860 It's really important at this time, more than ever, 90 00:03:44,860 --> 00:03:46,930 that we begin to understand those population 91 00:03:46,930 --> 00:03:48,820 and annex of those invertebrate groups. 92 00:03:48,820 --> 00:03:51,130 Our knowledge for these groups is very limited 93 00:03:51,130 --> 00:03:53,140 to begin with, and in montane environments, 94 00:03:53,140 --> 00:03:54,553 it's basically non-existent. 95 00:03:55,570 --> 00:03:58,210 So I'm not an ecologist that cares about invertebrates 96 00:03:58,210 --> 00:04:00,190 because that happens to be what birds eat 97 00:04:00,190 --> 00:04:01,660 during the summer months. 98 00:04:01,660 --> 00:04:04,570 I'm an ecologist that works with invertebrate populations 99 00:04:04,570 --> 00:04:05,950 and bird populations. 100 00:04:05,950 --> 00:04:07,210 And this is a nice opportunity 101 00:04:07,210 --> 00:04:09,370 to put them together into one project. 102 00:04:09,370 --> 00:04:11,830 So I'm a coordinator for a community science project 103 00:04:11,830 --> 00:04:14,590 called Mountain Birdwatch, where each June 104 00:04:14,590 --> 00:04:16,900 hundreds of community scientists trek out 105 00:04:16,900 --> 00:04:19,660 to more than 700 sampling locations, 106 00:04:19,660 --> 00:04:21,700 from the Catskills to (indistinct), 107 00:04:21,700 --> 00:04:24,190 and that is a high-elevation spruce for habitat 108 00:04:24,190 --> 00:04:26,380 on established hiking trails. 109 00:04:26,380 --> 00:04:29,890 They do repeated point-counts for 10 species of birds 110 00:04:29,890 --> 00:04:32,590 and red squirrel, which is the primary nest predator 111 00:04:32,590 --> 00:04:35,683 for these species of birds; that's each June. 112 00:04:36,970 --> 00:04:39,430 Last year at 42 of those sampling locations 113 00:04:39,430 --> 00:04:41,380 on Mount Mansfield, Bolton Mountain, 114 00:04:41,380 --> 00:04:43,630 in the dual peaks of Cherry, and Martha, 115 00:04:43,630 --> 00:04:44,800 and the White Mountains, 116 00:04:44,800 --> 00:04:47,020 I deployed invertebrate sampling equipment 117 00:04:47,020 --> 00:04:49,300 at those 42 locations where community scientists 118 00:04:49,300 --> 00:04:53,380 also went out and did repeated point-counts for those birds. 119 00:04:53,380 --> 00:04:56,950 I'm gonna be focusing on just the six birds in orange there. 120 00:04:56,950 --> 00:04:58,300 Those are the most common birds 121 00:04:58,300 --> 00:05:00,400 at these 42 sampling locations. 122 00:05:00,400 --> 00:05:01,750 And they're present at about, 123 00:05:01,750 --> 00:05:03,610 each one of those species is present 124 00:05:03,610 --> 00:05:07,030 at about 40%-50% of those 42 stations, 125 00:05:07,030 --> 00:05:08,680 so pretty common in the data set. 126 00:05:09,640 --> 00:05:12,400 Our knowledge of the diet and invertebrate associations 127 00:05:12,400 --> 00:05:14,740 for these six birds is pretty poor. 128 00:05:14,740 --> 00:05:16,690 Some of our best knowledge of the diet 129 00:05:16,690 --> 00:05:20,080 of these six bird species comes from 100 years ago 130 00:05:20,080 --> 00:05:21,730 when somebody walked out into the woods 131 00:05:21,730 --> 00:05:23,350 and shot half-a-dozen in the spurs, 132 00:05:23,350 --> 00:05:25,000 and cut open their stomach. 133 00:05:25,000 --> 00:05:27,460 For other species like Swainson's Thrush, 134 00:05:27,460 --> 00:05:29,230 our knowledge of their diet comes from 135 00:05:29,230 --> 00:05:32,860 lower-elevation habitats, not in the spruce-fir zone. 136 00:05:32,860 --> 00:05:35,410 And most of our knowledge about their diet 137 00:05:35,410 --> 00:05:38,382 is loosely inferred from correlations of invertebrates 138 00:05:38,382 --> 00:05:41,200 in their territories, not of these birds 139 00:05:41,200 --> 00:05:44,440 directly eating those invertebrate groups. 140 00:05:44,440 --> 00:05:47,230 But if I had to make some guesses putting it all together, 141 00:05:47,230 --> 00:05:50,320 I would predict that the occupancy of these birds 142 00:05:50,320 --> 00:05:53,440 is positively related to increases in the biomass 143 00:05:53,440 --> 00:05:57,763 of beetles, true bugs, hymenoptera, and flies probably. 144 00:05:59,186 --> 00:06:02,020 So at those 42 sampling locations last June, 145 00:06:02,020 --> 00:06:04,660 I deployed some invertebrate-sampling equipment, 146 00:06:04,660 --> 00:06:07,780 they were deployed for a week, 70 consecutive dates. 147 00:06:07,780 --> 00:06:10,303 At each Mountain Birdwatch sampling location 148 00:06:10,303 --> 00:06:12,190 20 meters off the trail, 149 00:06:12,190 --> 00:06:15,103 I deployed four pitfall-sampling traps for each standard, 150 00:06:15,943 --> 00:06:17,530 spaced out in a 25-square-meter area. 151 00:06:17,530 --> 00:06:19,450 They're covered with these plates 152 00:06:19,450 --> 00:06:22,060 to prevent them from filling up with water. 153 00:06:22,060 --> 00:06:24,550 We also deploy these modified window traps, 154 00:06:24,550 --> 00:06:26,590 which I think are pretty cool, because they're largely built 155 00:06:26,590 --> 00:06:30,070 with recycled materials for about $5 a piece. 156 00:06:30,070 --> 00:06:32,200 That's a soda bottle in the middle there. 157 00:06:32,200 --> 00:06:34,810 Insects can hit the clear plastic in the middle, 158 00:06:34,810 --> 00:06:36,670 fly downwards and be captured. 159 00:06:36,670 --> 00:06:39,340 Other groups of insects some are likely to flee upwards 160 00:06:39,340 --> 00:06:42,400 and they're captured on a separate container up there. 161 00:06:42,400 --> 00:06:44,770 We tested side-by-side in the models, 162 00:06:44,770 --> 00:06:47,590 with the yellow-attractant paper in the inside, 163 00:06:47,590 --> 00:06:48,970 and models without. 164 00:06:48,970 --> 00:06:51,640 The models without, probably because of the small 165 00:06:51,640 --> 00:06:53,200 surface area of these traps to begin with, 166 00:06:53,200 --> 00:06:55,330 captured very few insects. 167 00:06:55,330 --> 00:06:58,041 So last June, our modified window sampling traps 168 00:06:58,041 --> 00:07:00,823 had that yellow-attracting sheet inside. 169 00:07:01,960 --> 00:07:05,260 All the specimens we capture in the field after one week 170 00:07:05,260 --> 00:07:08,320 are brought back to the lab, identified under a microscope, 171 00:07:08,320 --> 00:07:10,540 to at least order. 172 00:07:10,540 --> 00:07:14,980 And then dried in a specimen oven for 30 hours, 173 00:07:14,980 --> 00:07:17,050 and then dry biomass is measured 174 00:07:17,050 --> 00:07:19,993 to the nearest 0.0001 grams. 175 00:07:22,060 --> 00:07:25,510 We captured about 5,000 individual invertebrates last June 176 00:07:25,510 --> 00:07:28,270 at these 42 Mountain Birdwatch sampling stations. 177 00:07:28,270 --> 00:07:30,970 Here you see the brown, the percentage of the biomass 178 00:07:30,970 --> 00:07:32,500 of each group, and in green, 179 00:07:32,500 --> 00:07:35,530 the percentage of the number of specimens of each group. 180 00:07:35,530 --> 00:07:37,390 Pretty congruent for most groups, 181 00:07:37,390 --> 00:07:39,820 except you'll notice halfway down at the bottom, 182 00:07:39,820 --> 00:07:43,420 flies constituted about 60% of the individuals 183 00:07:43,420 --> 00:07:44,950 on our sampling data set. 184 00:07:44,950 --> 00:07:46,540 But they're so small and weigh so little, 185 00:07:46,540 --> 00:07:49,510 they only represent about 5% of the biomass 186 00:07:49,510 --> 00:07:50,740 in our total sample. 187 00:07:50,740 --> 00:07:53,890 And that's an inverse relationship for common land slugs. 188 00:07:53,890 --> 00:07:55,360 I will just point out that, obviously, 189 00:07:55,360 --> 00:07:56,740 camel crickets are not an order, 190 00:07:56,740 --> 00:07:58,780 but that's the only orthopteran that we captured 191 00:07:58,780 --> 00:08:00,280 in our data set. 192 00:08:00,280 --> 00:08:05,200 And likewise that this community is pretty simple, really, 193 00:08:05,200 --> 00:08:08,680 for 5,000 individuals and over 1,000 trapping nights. 194 00:08:08,680 --> 00:08:10,990 There's a lot of very common groups of invertebrates 195 00:08:10,990 --> 00:08:12,370 up there, like scale insects, 196 00:08:12,370 --> 00:08:14,260 that are not represented here at all, 197 00:08:14,260 --> 00:08:15,910 I entirely acknowledge that. 198 00:08:15,910 --> 00:08:19,000 Also, most of the mites we capture are development mites. 199 00:08:19,000 --> 00:08:20,980 There were some mites that we captured, you know, 200 00:08:20,980 --> 00:08:22,930 one millimeter or so that are difficult 201 00:08:22,930 --> 00:08:25,210 to put into a specific mite order. 202 00:08:25,210 --> 00:08:27,340 And as common in these types of studies, 203 00:08:27,340 --> 00:08:30,490 those orders of mites are lumped together for analysis, 204 00:08:30,490 --> 00:08:32,143 ditto with millipedes. 205 00:08:33,640 --> 00:08:36,100 So I'm interested in taking measurements 206 00:08:36,100 --> 00:08:38,230 of the invertebrate community biomass, 207 00:08:38,230 --> 00:08:40,450 and using that to predict the occupancy pattern 208 00:08:40,450 --> 00:08:41,890 for those six species of birds. 209 00:08:41,890 --> 00:08:44,470 The first thing you do is build an occupancy model 210 00:08:44,470 --> 00:08:45,760 that accounts for variation, 211 00:08:45,760 --> 00:08:47,617 observer ability on the detection side, 212 00:08:47,617 --> 00:08:49,780 and time of day, and day of the month. 213 00:08:49,780 --> 00:08:52,180 And on on the occupancy side of the model, 214 00:08:52,180 --> 00:08:56,620 I model occupancy via quadratic relationship to elevation. 215 00:08:56,620 --> 00:08:58,450 And as you can imagine for montane birds, 216 00:08:58,450 --> 00:09:00,760 that's a really powerful predictor 217 00:09:00,760 --> 00:09:02,200 of where we expect to see Bicknell, 218 00:09:02,200 --> 00:09:04,420 Swainson, and White-Throated Sparrow. 219 00:09:04,420 --> 00:09:06,520 And then I add various covariates, 220 00:09:06,520 --> 00:09:08,950 describing the biomass of the invertebrate community 221 00:09:08,950 --> 00:09:11,560 to that same model, and see if it improves 222 00:09:11,560 --> 00:09:15,010 the predictability of that baseline occupancy model 223 00:09:15,010 --> 00:09:17,230 that just includes elevation. 224 00:09:17,230 --> 00:09:20,440 Oh, I measure, I calculate species richness, 225 00:09:20,440 --> 00:09:23,789 Simpson's measurements of diversity and evenness, 226 00:09:23,789 --> 00:09:27,220 using individual specimens and biomass. 227 00:09:27,220 --> 00:09:29,230 And I'll just tell you that those indices 228 00:09:29,230 --> 00:09:32,230 had no predicted ability to describe patterns 229 00:09:32,230 --> 00:09:35,230 of occupancy of those six bird species. 230 00:09:35,230 --> 00:09:36,790 So proportional representation 231 00:09:36,790 --> 00:09:39,070 of that invertebrate community's not a good predictor. 232 00:09:39,070 --> 00:09:40,900 Maybe you think it's total biomass, right? 233 00:09:40,900 --> 00:09:42,250 Maybe you think just locations 234 00:09:42,250 --> 00:09:45,100 that have lots of stuff, period. 235 00:09:45,100 --> 00:09:48,040 Some of it's gonna be available for food for those birds, 236 00:09:48,040 --> 00:09:49,960 and that locations with lots of biomass 237 00:09:49,960 --> 00:09:52,870 might be predictive of high-occupancy rates. 238 00:09:52,870 --> 00:09:54,430 And that is also not the case 239 00:09:54,430 --> 00:09:57,160 for any of those six species of birds. 240 00:09:57,160 --> 00:10:00,250 Okay, maybe you think that it means 241 00:10:00,250 --> 00:10:02,590 taking a chunk of beetle biomass, 242 00:10:02,590 --> 00:10:04,810 and a chunk of hymenoptera beetle mass, 243 00:10:04,810 --> 00:10:07,240 mixed in with a little bit of true bug biomass 244 00:10:07,240 --> 00:10:09,100 in some complex relationship. 245 00:10:09,100 --> 00:10:12,070 If you think like that, you think more methodically, 246 00:10:12,070 --> 00:10:14,620 and you probably think PCA is a great thing to do 247 00:10:14,620 --> 00:10:16,990 -on a Friday night. -(audience laugh) 248 00:10:16,990 --> 00:10:18,730 If we do that and we break down 249 00:10:18,730 --> 00:10:21,880 the biomass invertebrate dataset into principle components 250 00:10:21,880 --> 00:10:24,220 you see the first principle component on the right there 251 00:10:24,220 --> 00:10:28,570 explains a lousy 18% of the variation of the dataset. 252 00:10:28,570 --> 00:10:30,460 And you probably know where I'm going with this, 253 00:10:30,460 --> 00:10:32,610 but principle components of the biomass 254 00:10:32,610 --> 00:10:35,980 of these invertebrate communities are very poor predictors 255 00:10:35,980 --> 00:10:38,470 of the occupancy pattern of those six birds. 256 00:10:38,470 --> 00:10:39,370 They don't do any better 257 00:10:39,370 --> 00:10:42,400 than a simple model that just includes elevation. 258 00:10:42,400 --> 00:10:43,690 What have we left ourselves with, 259 00:10:43,690 --> 00:10:46,490 since we've ruled out most of the things I can think of? 260 00:10:47,650 --> 00:10:51,580 And the only thing that adequately predicts occupancy, 261 00:10:51,580 --> 00:10:55,150 for any of these species, is the abundance or scarcity 262 00:10:55,150 --> 00:10:58,330 of biomass of specific invertebrate groups. 263 00:10:58,330 --> 00:11:01,450 So for Bicknell's Thrush there at the top and middle, 264 00:11:01,450 --> 00:11:03,730 you see there's a positive association 265 00:11:03,730 --> 00:11:05,562 of the occupancy of Bicknell's Thrush 266 00:11:05,562 --> 00:11:09,130 with increasing biomass of true bugs 267 00:11:09,130 --> 00:11:11,470 around that Mountain Birdwatch sampling station, 268 00:11:11,470 --> 00:11:15,580 and an inverse relationship to the biomass of beetles. 269 00:11:15,580 --> 00:11:18,130 Some of these results are surprising to me. 270 00:11:18,130 --> 00:11:20,470 For Swainson's Thrush, that makes sense to me, 271 00:11:20,470 --> 00:11:22,960 for Yellow-Bellied Flycatcher, okay. 272 00:11:22,960 --> 00:11:24,550 For some of these other groups, 273 00:11:24,550 --> 00:11:27,073 I'm struggling to understand those relationships. 274 00:11:28,030 --> 00:11:30,460 It could be very much because we have very limited 275 00:11:30,460 --> 00:11:32,830 knowledge of the diet of these species, 276 00:11:32,830 --> 00:11:35,860 that we have discovered some important new associations 277 00:11:35,860 --> 00:11:38,470 of their diet for these groups, that's possible. 278 00:11:38,470 --> 00:11:41,800 It's also possible with just 5,000 specimens 279 00:11:41,800 --> 00:11:43,450 in one year of the study that some of these 280 00:11:43,450 --> 00:11:45,130 are spurious relationships. 281 00:11:45,130 --> 00:11:46,990 If that was the case, I also would've expected 282 00:11:46,990 --> 00:11:49,060 some of those principle components to spuriously 283 00:11:49,060 --> 00:11:52,270 show up for some of those species as well. 284 00:11:52,270 --> 00:11:54,790 It may tell us something more complex though, 285 00:11:54,790 --> 00:11:57,580 mites is pretty surprising to me, right? 286 00:11:57,580 --> 00:12:00,460 Except that velvet mites in their juvenile form 287 00:12:00,460 --> 00:12:03,790 are ecto-parasites on adult insects and plants. 288 00:12:03,790 --> 00:12:06,250 And in their adult form, they're direct predators 289 00:12:06,250 --> 00:12:08,863 on insects and on insect eggs. 290 00:12:10,152 --> 00:12:11,290 Though the mites have these incredibly 291 00:12:11,290 --> 00:12:15,100 complex relationship-enforced ecosystems with fungi 292 00:12:15,100 --> 00:12:18,340 and soil formation and structure and decomposition, 293 00:12:18,340 --> 00:12:20,980 it could be that the presence of some of these groups 294 00:12:20,980 --> 00:12:23,170 for these species isn't telling us directly 295 00:12:23,170 --> 00:12:25,750 about what they're eating, so much as it's telling us 296 00:12:25,750 --> 00:12:28,480 about the health of the forest ecosystem there, 297 00:12:28,480 --> 00:12:31,300 or the strength of those ecological relationships. 298 00:12:31,300 --> 00:12:32,713 I don't know that, honestly. 299 00:12:34,840 --> 00:12:36,850 So in conclusion, again, it's the abundance 300 00:12:36,850 --> 00:12:38,770 and scarcity of individual groups of invertebrates, 301 00:12:38,770 --> 00:12:40,480 that's the only thing that adequately predicts 302 00:12:40,480 --> 00:12:42,813 occupancy for these species. 303 00:12:42,813 --> 00:12:46,450 One year is a short study, is a short timeframe. 304 00:12:46,450 --> 00:12:48,880 Lots of studies have correlated invertebrates 305 00:12:48,880 --> 00:12:51,910 in the environment to birds' nesting success, 306 00:12:51,910 --> 00:12:54,160 and found relationships between nesting success 307 00:12:54,160 --> 00:12:57,790 and beetles in one year, but not the next year. 308 00:12:57,790 --> 00:12:59,890 It's possible some of these results are spurious, 309 00:12:59,890 --> 00:13:02,410 it's possible these relationships change, 310 00:13:02,410 --> 00:13:05,560 based on overall food availability in a given year. 311 00:13:05,560 --> 00:13:06,790 In years where there's lots of food, 312 00:13:06,790 --> 00:13:08,500 maybe those preferences break down 313 00:13:08,500 --> 00:13:10,330 and we don't see those relationships. 314 00:13:10,330 --> 00:13:12,700 It might only be when food is really scarce 315 00:13:12,700 --> 00:13:15,760 that we're able to detect those relationships. 316 00:13:15,760 --> 00:13:18,100 Also you know, we're limited 317 00:13:18,100 --> 00:13:19,750 in our sampling approaches, absolutely. 318 00:13:19,750 --> 00:13:21,400 I think I captured a single scorpionfly 319 00:13:21,400 --> 00:13:24,370 of the 5,000 in invertebrates last summer, 320 00:13:24,370 --> 00:13:26,162 of the panorpa species. 321 00:13:26,162 --> 00:13:30,340 If scorpionflies are important for Blackpoil Warblers, 322 00:13:30,340 --> 00:13:32,408 there's no way I'm gonna be able to detect that 323 00:13:32,408 --> 00:13:34,810 in this data set, because they're just not captured 324 00:13:34,810 --> 00:13:38,080 with enough frequency, so I freely acknowledge that. 325 00:13:38,080 --> 00:13:40,930 All sampling methods have biases. 326 00:13:40,930 --> 00:13:44,680 And lastly, I'd say identifying things to order. 327 00:13:44,680 --> 00:13:46,300 It takes a lot of time obviously, 328 00:13:46,300 --> 00:13:50,500 but it may be that that high-order of identification 329 00:13:50,500 --> 00:13:53,830 prevents us from teasing out lower-order relationships. 330 00:13:53,830 --> 00:13:58,830 So rogue beetles are important to White-Throated Sparrows, 331 00:13:58,900 --> 00:14:01,090 but we are lumping in rogue beetles 332 00:14:01,090 --> 00:14:03,400 with tumbling flower beetles, and lightning beetles, 333 00:14:03,400 --> 00:14:05,800 and ground beetles, that relationship may become 334 00:14:05,800 --> 00:14:09,430 obscured for that individual family of beetles 335 00:14:09,430 --> 00:14:11,920 and White-Throated Sparrow, Black Warbler. 336 00:14:11,920 --> 00:14:13,810 And we may be failing to detect that, 337 00:14:13,810 --> 00:14:15,790 because we are mixing data together 338 00:14:15,790 --> 00:14:17,669 in a way that's uninformative. 339 00:14:17,669 --> 00:14:21,490 The alternative is identifying everything to family level, 340 00:14:21,490 --> 00:14:24,070 except maybe for beetles, is impossible. 341 00:14:24,070 --> 00:14:28,068 I mean, it's not logistically or expertise-wise possible. 342 00:14:28,068 --> 00:14:30,520 And lastly, I'd say I look forward to having 343 00:14:30,520 --> 00:14:32,020 multiple years of this data. 344 00:14:32,020 --> 00:14:33,910 We'll be able to look at trends 345 00:14:33,910 --> 00:14:35,770 in the invertebrate communities themselves, 346 00:14:35,770 --> 00:14:37,840 look at how much variation we see 347 00:14:37,840 --> 00:14:39,880 in our capture rates from year-to-year, 348 00:14:39,880 --> 00:14:42,820 so we can start to tease out signal from noise. 349 00:14:42,820 --> 00:14:45,100 And look at the trends of the birds of these locations 350 00:14:45,100 --> 00:14:47,350 compared to the trends of the invertebrates. 351 00:14:47,350 --> 00:14:50,260 And in subsequent summers, we'll be taking measurements 352 00:14:50,260 --> 00:14:51,940 of forced health and vegetation. 353 00:14:51,940 --> 00:14:55,540 So this is just the first complete year of this study. 354 00:14:55,540 --> 00:14:58,570 I'd like to thank FEMC for partially funding this study, 355 00:14:58,570 --> 00:15:00,490 for the entire Mountain Birdwatch family, 356 00:15:00,490 --> 00:15:03,021 and for all of you being here today, thank you. 357 00:15:03,021 --> 00:15:06,104 (audience applauds) 358 00:15:12,580 --> 00:15:13,413 Yeah. 359 00:15:14,740 --> 00:15:17,320 [Female Speaker] Did you parse out 360 00:15:17,320 --> 00:15:18,790 different populations at all? 361 00:15:18,790 --> 00:15:22,360 Because I'm wondering if maybe as we see 362 00:15:22,360 --> 00:15:27,360 the birds' habitat selection varying for a given species 363 00:15:27,400 --> 00:15:29,980 across their range, maybe there's differences 364 00:15:29,980 --> 00:15:33,880 in food-selection choices that are being obscured 365 00:15:33,880 --> 00:15:35,800 when you lump all of them together. 366 00:15:35,800 --> 00:15:40,800 Yes, so I did include mountain identity as a component 367 00:15:42,580 --> 00:15:43,750 of the principle components. 368 00:15:43,750 --> 00:15:45,940 And I didn't show that, I didn't show that by-plot, 369 00:15:45,940 --> 00:15:46,960 just to be simple. 370 00:15:46,960 --> 00:15:50,890 But yes, that comes across very strongly in that by-plot. 371 00:15:50,890 --> 00:15:52,840 If you show the classic principle components 372 00:15:52,840 --> 00:15:54,580 with the arrows going every direction, 373 00:15:54,580 --> 00:15:58,180 and you color-coded sample locations by mountain, 374 00:15:58,180 --> 00:15:59,860 you'd see some real clustering. 375 00:15:59,860 --> 00:16:02,110 And it does suggest, I'm not sure it suggests 376 00:16:02,110 --> 00:16:05,834 preferences in the birds, as much as it might suggest 377 00:16:05,834 --> 00:16:09,160 localness of the invertebrate community. 378 00:16:09,160 --> 00:16:10,810 But that definitely comes through 379 00:16:10,810 --> 00:16:12,660 in the principle components analysis. 380 00:16:19,390 --> 00:16:21,730 -These T-shirts are available, -(audience laughs) 381 00:16:21,730 --> 00:16:24,550 well habitats (indistinct), always great gifts, 382 00:16:24,550 --> 00:16:25,750 great stocking stuffers. 383 00:16:26,740 --> 00:16:27,758 Rosalind. 384 00:16:27,758 --> 00:16:30,250 [Rosalind] Are you aware, I mean, would you expect 385 00:16:30,250 --> 00:16:34,450 to maybe see this complexity, where the insects 386 00:16:34,450 --> 00:16:36,474 may be just indicating something else 387 00:16:36,474 --> 00:16:38,020 that we're not measuring directly? 388 00:16:38,020 --> 00:16:40,000 Would you expect to see that in other habitats, 389 00:16:40,000 --> 00:16:42,940 or have other people done research that have found that? 390 00:16:42,940 --> 00:16:46,480 Yes, and it takes a lot more direct observations of those, 391 00:16:46,480 --> 00:16:49,570 well actually, you can build pathway models 392 00:16:49,570 --> 00:16:52,630 that make predictions, that you say, you know, 393 00:16:52,630 --> 00:16:54,490 there's a bunch of invertebrates and a bunch of birds, 394 00:16:54,490 --> 00:16:57,400 and there's some reproductive success and adult survival. 395 00:16:57,400 --> 00:16:59,050 I don't know how these things are all connected, 396 00:16:59,050 --> 00:17:00,040 you tell me. 397 00:17:00,040 --> 00:17:01,513 And I've built those models, 398 00:17:02,644 --> 00:17:03,477 and absolutely you can build 399 00:17:03,477 --> 00:17:04,663 those types of pathway analyses. 400 00:17:05,770 --> 00:17:08,380 The vast majority of bird papers 401 00:17:08,380 --> 00:17:11,803 that look at relationships between invertebrate abundance, 402 00:17:13,362 --> 00:17:15,370 or biomass, and some measurement of breeding success 403 00:17:15,370 --> 00:17:16,603 or just adult density. 404 00:17:17,770 --> 00:17:19,840 Don't make the leap. 405 00:17:19,840 --> 00:17:22,510 Don't go to the actual mechanisms, 406 00:17:22,510 --> 00:17:24,192 and don't try to say that... 407 00:17:24,192 --> 00:17:25,930 You know, is the stuff that we capture in their territory 408 00:17:25,930 --> 00:17:27,970 necessarily is the stuff they're eating? 409 00:17:27,970 --> 00:17:29,770 They tend to leave it ambiguous, 410 00:17:29,770 --> 00:17:31,960 'cause it's amazing how little 411 00:17:31,960 --> 00:17:35,080 direct-observations we have of foraging. 412 00:17:35,080 --> 00:17:36,799 You know, for Bicknell's Thrush, 413 00:17:36,799 --> 00:17:39,700 we have six dead Bicknell's Thrush, basically, 414 00:17:39,700 --> 00:17:42,163 that tells us about the diet of those species, 415 00:17:43,210 --> 00:17:44,110 and that's the limit. 416 00:17:44,110 --> 00:17:46,750 So we definitely need some very basic, 417 00:17:46,750 --> 00:17:49,780 natural history studies, documenting the actual 418 00:17:49,780 --> 00:17:52,090 food items that are eaten by these species. 419 00:17:52,090 --> 00:17:54,520 And then we probably need to do some more 420 00:17:54,520 --> 00:17:56,350 creative modeling to try to get it, 421 00:17:56,350 --> 00:17:59,020 how mites are actually informing occupancy 422 00:17:59,020 --> 00:18:00,340 of White-Tailed Sparrows. 423 00:18:00,340 --> 00:18:02,808 Probably not, because they're being eaten a lot 424 00:18:02,808 --> 00:18:04,483 for those birds. 425 00:18:09,320 --> 00:18:10,930 Okay, yeah? 426 00:18:10,930 --> 00:18:13,750 [Male Speaker] Are there any, like, habitat interventions 427 00:18:13,750 --> 00:18:16,026 that you could do to help mountain birds, 428 00:18:16,026 --> 00:18:18,282 like, up in the outlying north? 429 00:18:18,282 --> 00:18:19,513 That's a great question. 430 00:18:19,513 --> 00:18:23,500 Here in Vermont, Act 250 would make that challenging. 431 00:18:23,500 --> 00:18:24,333 [Male Speaker] All right. 432 00:18:24,333 --> 00:18:27,310 But actually on Monday, I was in a 2-hour Zoom call 433 00:18:27,310 --> 00:18:30,700 with 30 people, people that studied Bicknell's Thrush, 434 00:18:30,700 --> 00:18:34,390 and 15 commercial foresters from Wagner, 435 00:18:34,390 --> 00:18:35,890 and all of the other big companies, 436 00:18:35,890 --> 00:18:38,380 talking about how we can manage industrial 437 00:18:38,380 --> 00:18:40,330 commercial forest lands in Northern New Hampshire 438 00:18:40,330 --> 00:18:44,173 and Maine, to promote Bicknell's Thrush. 439 00:18:45,160 --> 00:18:47,407 And it likely means promoting food 440 00:18:47,407 --> 00:18:49,511 and availability of food for them, 441 00:18:49,511 --> 00:18:52,090 for those bird species. 442 00:18:52,090 --> 00:18:56,230 Yeah, so the answer at the moment is no, don't know, 443 00:18:56,230 --> 00:18:58,900 but a lot of us are thinking about that. 444 00:18:58,900 --> 00:19:02,380 A lot of these species are disturbance-prone species, 445 00:19:02,380 --> 00:19:05,620 and in Maine they are most abundant, adjacent to fir waves. 446 00:19:05,620 --> 00:19:07,420 I did analysis for the state of Vermont, 447 00:19:07,420 --> 00:19:11,530 looking at abundance of these birds compared to ski slopes. 448 00:19:11,530 --> 00:19:14,230 And most of these bird species are most abundant, 449 00:19:14,230 --> 00:19:17,620 greatest density, immediately adjacent to ski slopes. 450 00:19:17,620 --> 00:19:18,453 [Male Speaker] Huh. 451 00:19:18,453 --> 00:19:21,550 It may be that those ski slopes artificially mimic 452 00:19:21,550 --> 00:19:23,470 the natural disturbance these species would see 453 00:19:23,470 --> 00:19:25,420 in fir waves or something. 454 00:19:25,420 --> 00:19:28,060 I would point out that Kent McFarland and Chris Rimmer, 455 00:19:28,060 --> 00:19:32,410 20 years ago, studied the birds at the edges of ski slopes 456 00:19:32,410 --> 00:19:36,336 versus the interior, and couldn't find any differences 457 00:19:36,336 --> 00:19:39,730 in survival of the adults, and nesting productivity, 458 00:19:39,730 --> 00:19:42,160 survival of the young, clutch size. 459 00:19:42,160 --> 00:19:43,960 So it appears, as far as we know, 460 00:19:43,960 --> 00:19:46,000 that adults cue in on those disturbances, 461 00:19:46,000 --> 00:19:48,370 but we can't actually detect any improvements 462 00:19:48,370 --> 00:19:51,400 in fitness for birds that choose to nest there. 463 00:19:51,400 --> 00:19:52,960 But all that's on our mind when we're working 464 00:19:52,960 --> 00:19:55,270 with commercial foresters in the next 20 years here, 465 00:19:55,270 --> 00:19:57,610 to think about how to manage the remaining forest 466 00:19:57,610 --> 00:19:59,610 for Bicknell's Thrush and other species. 467 00:20:02,290 --> 00:20:03,123 Thank you. 468 00:20:04,044 --> 00:20:04,877 [Female Speaker] Great, thank you. 469 00:20:04,877 --> 00:20:07,960 (audience applauds)