1
00:00:02,890 --> 00:00:03,723
- [Kate] Okay

2
00:00:03,723 --> 00:00:05,810
Thanks to the Forest Ecosystem
Monitoring Cooperative

3
00:00:05,810 --> 00:00:08,990
for organizing this meeting and
for giving me an opportunity

4
00:00:08,990 --> 00:00:11,030
to talk about our program.

5
00:00:11,030 --> 00:00:13,290
I, of course, wish we
could do this in person,

6
00:00:13,290 --> 00:00:14,690
but I'm glad we're at least able

7
00:00:14,690 --> 00:00:16,360
to have this virtual meeting.

8
00:00:16,360 --> 00:00:17,730
And so, I'm Kate Miller.

9
00:00:17,730 --> 00:00:19,410
And today I'm gonna talk about

10
00:00:19,410 --> 00:00:21,380
what we've learned over the last 14 years

11
00:00:21,380 --> 00:00:23,820
of monitoring in eastern national parks,

12
00:00:23,820 --> 00:00:25,970
as well as concerns we
have for the future.

13
00:00:27,320 --> 00:00:29,560
Before we get to that, I
want to remind everyone

14
00:00:29,560 --> 00:00:31,620
about the mission of the
National Park Service,

15
00:00:31,620 --> 00:00:34,330
which is to protect natural
and cultural resources

16
00:00:34,330 --> 00:00:36,600
unimpaired for future generations.

17
00:00:36,600 --> 00:00:39,300
So parks really have the
highest level of protection

18
00:00:39,300 --> 00:00:41,680
of any federal land management agency.

19
00:00:41,680 --> 00:00:43,070
And in all but a few cases,

20
00:00:43,070 --> 00:00:45,477
it also includes protection from logging.

21
00:00:45,477 --> 00:00:47,730
But as late as the 1990s, many parks

22
00:00:47,730 --> 00:00:50,220
had very limited knowledge
of their natural resources

23
00:00:50,220 --> 00:00:52,930
and whether management
actions were effective.

24
00:00:52,930 --> 00:00:53,910
And it's hard to know

25
00:00:53,910 --> 00:00:55,940
whether you're meeting the
mission of the Parks Service

26
00:00:55,940 --> 00:00:58,170
if you don't know what resources you have

27
00:00:58,170 --> 00:01:00,420
and you don't know what
condition they're in.

28
00:01:02,190 --> 00:01:03,790
The Inventory and Monitoring Program

29
00:01:03,790 --> 00:01:05,160
of the National Park Service

30
00:01:05,160 --> 00:01:07,770
was established to fill this need.

31
00:01:07,770 --> 00:01:10,670
Specifically, the I and M
Program was established to assess

32
00:01:10,670 --> 00:01:13,920
the status and trends of
important ecosystems in parks

33
00:01:13,920 --> 00:01:16,340
to help inform management decisions.

34
00:01:16,340 --> 00:01:19,140
The I and M program started
with baseline inventories,

35
00:01:19,140 --> 00:01:21,930
like vegetation maps and
small mammal inventories,

36
00:01:21,930 --> 00:01:24,913
and has since transitioned
into long-term monitoring.

37
00:01:26,000 --> 00:01:28,840
I worked primarily for the
Northeast Temperate Network,

38
00:01:28,840 --> 00:01:32,270
which covers parks from Morristown
National Historical Park

39
00:01:32,270 --> 00:01:35,650
in New Jersey, up to Acadia
National Park in Maine.

40
00:01:35,650 --> 00:01:38,660
We also coordinate monitoring
along the Appalachian Trail.

41
00:01:38,660 --> 00:01:41,470
And we implement multiple
long-term monitoring programs

42
00:01:41,470 --> 00:01:43,020
in these parks, which
I'm gonna walk through

43
00:01:43,020 --> 00:01:44,623
in the next couple of slides.

44
00:01:46,070 --> 00:01:48,910
We monitor water quality
and quantity annually

45
00:01:48,910 --> 00:01:52,210
in all but Boston Harbor
Islands and Appalachian Trail,

46
00:01:52,210 --> 00:01:54,643
and we've been doing that since 2006.

47
00:01:56,200 --> 00:01:58,070
We've also been monitoring

48
00:01:58,070 --> 00:02:00,741
breeding land birds annually since 2006.

49
00:02:00,741 --> 00:02:04,110
And we mostly use volunteers
to collect the data.

50
00:02:04,110 --> 00:02:06,174
If you want to know
more about this program,

51
00:02:06,174 --> 00:02:07,770
I encourage you to check out

52
00:02:07,770 --> 00:02:10,730
our program manager Aaron
Weed's talk tomorrow,

53
00:02:10,730 --> 00:02:12,990
where he's gonna be discussing
some recent findings

54
00:02:12,990 --> 00:02:15,240
from a trend analysis for this program.

55
00:02:15,240 --> 00:02:17,773
And the info on his talk is below.

56
00:02:19,830 --> 00:02:23,490
Okay, next we have a number
of park-specific protocols

57
00:02:23,490 --> 00:02:25,700
based on park resources and needs.

58
00:02:25,700 --> 00:02:29,560
So we have freshwater
wetland monitoring in Acadia,

59
00:02:29,560 --> 00:02:31,820
and rocky intertidal habitat monitoring

60
00:02:31,820 --> 00:02:33,920
in Acadia and Boston Harbor Islands,

61
00:02:33,920 --> 00:02:36,970
and then coastal bird monitoring
in Boston Harbor Islands.

62
00:02:36,970 --> 00:02:39,970
And wherever possible,
we've adapted our protocols

63
00:02:39,970 --> 00:02:42,180
from existing national protocols.

64
00:02:42,180 --> 00:02:43,659
So for wetland monitoring,

65
00:02:43,659 --> 00:02:47,830
we use EPA's National Wetland
Condition Assessment Protocol.

66
00:02:47,830 --> 00:02:50,130
Our rocky intertidal protocol was adapted

67
00:02:50,130 --> 00:02:52,290
from the Marine Protocol,
which was developed

68
00:02:52,290 --> 00:02:55,340
for rocky intertidal
habitats on the West Coast.

69
00:02:55,340 --> 00:02:57,500
And there really are a ton of benefits

70
00:02:57,500 --> 00:03:00,690
to adapting existing and
nationwide protocols,

71
00:03:00,690 --> 00:03:03,640
which I'm gonna kind of talk
about throughout this plenary.

72
00:03:06,690 --> 00:03:09,570
Okay, finally, we have our
forest health protocol.

73
00:03:09,570 --> 00:03:12,350
So the parks with the
green tree symbol are part

74
00:03:12,350 --> 00:03:14,970
of our field-based forest
monitoring program,

75
00:03:14,970 --> 00:03:16,380
where we send crews out every year

76
00:03:16,380 --> 00:03:18,450
to sample permanent forest plots.

77
00:03:18,450 --> 00:03:21,300
Our protocol was adapted from
the U.S. Forest Service's

78
00:03:21,300 --> 00:03:23,260
Forest Inventory and Analysis Program,

79
00:03:23,260 --> 00:03:27,040
which is a nationwide survey
of permanent forest plots.

80
00:03:27,040 --> 00:03:30,600
And we've been monitoring
in these parks since 2006.

81
00:03:30,600 --> 00:03:32,270
We sample on a four-year cycle.

82
00:03:32,270 --> 00:03:35,470
And so we're midway through
our fourth monitoring cycle.

83
00:03:35,470 --> 00:03:38,000
And I'll note that the
Appalachian Trail monitoring,

84
00:03:38,000 --> 00:03:40,090
the reason why it's a different symbol,

85
00:03:40,090 --> 00:03:44,080
is that we rely on summarizing
data from FIA plots

86
00:03:44,080 --> 00:03:46,160
within a specified buffer along the trail

87
00:03:46,160 --> 00:03:48,173
instead of sending out field crews.

88
00:03:50,440 --> 00:03:52,110
Okay, for those of you who are interested,

89
00:03:52,110 --> 00:03:54,520
I'm gonna walk through
our forest plot design.

90
00:03:54,520 --> 00:03:57,310
So our plots are 20 by 20 meters,

91
00:03:57,310 --> 00:04:00,310
except in Acadia where
they're 15 by 15 meters.

92
00:04:00,310 --> 00:04:03,660
So our plots are slightly
larger than one FIA subplot.

93
00:04:03,660 --> 00:04:06,550
And in each plot we monitor
tree diameter at breast height

94
00:04:06,550 --> 00:04:07,980
and condition of all trees that are

95
00:04:07,980 --> 00:04:10,560
at least 10 centimeters diameter.

96
00:04:10,560 --> 00:04:12,634
We also monitor seedlings and saplings

97
00:04:12,634 --> 00:04:15,760
in three, two meter radius micro plots.

98
00:04:15,760 --> 00:04:17,820
We periodically collect soil samples

99
00:04:17,820 --> 00:04:20,750
for chemical analysis
outside of our plots.

100
00:04:20,750 --> 00:04:22,170
We measure coarse woody debris

101
00:04:22,170 --> 00:04:25,180
along three, 15 meter long transects.

102
00:04:25,180 --> 00:04:29,170
We estimate vascular percent
cover for all species

103
00:04:29,170 --> 00:04:31,610
within eight, one meter square quadrats.

104
00:04:31,610 --> 00:04:35,143
And finally, we take photo
points of sixteens of the plot.

105
00:04:36,341 --> 00:04:40,660
Even though our plot diagram
isn't the same as FIA,

106
00:04:40,660 --> 00:04:44,700
wherever possible our methods
match those used by FIA.

107
00:04:44,700 --> 00:04:47,240
And major differences
really are that we opted

108
00:04:47,240 --> 00:04:51,940
for establishing many smaller
plots versus fewer large plots

109
00:04:51,940 --> 00:04:55,070
to better capture variability
in forests across each park.

110
00:04:55,070 --> 00:04:56,450
So our plots are smaller,

111
00:04:56,450 --> 00:04:58,950
but we're sampling a lot
of the same stuff as FIA.

112
00:05:01,290 --> 00:05:02,650
In addition to any NETN,

113
00:05:02,650 --> 00:05:04,490
there are over 90 parks at this point

114
00:05:04,490 --> 00:05:05,970
that are monitoring forest health

115
00:05:05,970 --> 00:05:10,100
using similar protocols
adapted from the FIA program.

116
00:05:10,100 --> 00:05:13,040
And particularly, in this
northeastern cluster,

117
00:05:13,040 --> 00:05:14,460
we've been monitoring forest health

118
00:05:14,460 --> 00:05:16,800
using similar protocols for over 12 years.

119
00:05:16,800 --> 00:05:19,300
And that's allowed us to
do more powerful analyses

120
00:05:19,300 --> 00:05:21,410
at the regional level
and to compare patterns

121
00:05:21,410 --> 00:05:24,920
in parks with surrounding
forest using FIA data.

122
00:05:24,920 --> 00:05:26,850
And I'll say that when
the I and M networks

123
00:05:26,850 --> 00:05:29,420
are first getting started,
there was really no incentive

124
00:05:29,420 --> 00:05:32,695
or pressure from above to
develop compatible protocols.

125
00:05:32,695 --> 00:05:35,580
Fortunately, there were a few
Eastern folks with insight

126
00:05:35,580 --> 00:05:37,170
to coordinate protocol development,

127
00:05:37,170 --> 00:05:39,550
and we're in a much better
place because of that.

128
00:05:39,550 --> 00:05:41,390
And most of my plenary is gonna focus

129
00:05:41,390 --> 00:05:43,190
on the things we've learned
because we were able

130
00:05:43,190 --> 00:05:45,930
to combine our data sets
for regional analyses

131
00:05:45,930 --> 00:05:47,927
and or compare our data with FIA.

132
00:05:50,551 --> 00:05:52,800
One of the examples of
how we've taken advantage

133
00:05:52,800 --> 00:05:55,690
of having similar
protocols adapted from FIA

134
00:05:55,690 --> 00:05:57,810
includes a series of
studies where we compared

135
00:05:57,810 --> 00:06:01,640
forests and parks with surrounding
forest using FIA plots.

136
00:06:01,640 --> 00:06:04,860
And we did this by comparing
forest data in parks

137
00:06:04,860 --> 00:06:07,590
with FIA plots in the
same ecological subsection

138
00:06:07,590 --> 00:06:09,600
containing each park.

139
00:06:09,600 --> 00:06:11,765
Ecological subsections, by definition,

140
00:06:11,765 --> 00:06:14,910
share similar climate,
topography, and geology,

141
00:06:14,910 --> 00:06:18,290
such that you expect the same
potential natural vegetation.

142
00:06:18,290 --> 00:06:21,170
And so differences between
park and surrounding forest

143
00:06:21,170 --> 00:06:23,363
should be due to management differences.

144
00:06:24,330 --> 00:06:27,210
So we looked at 10 metrics
of forest structure.

145
00:06:27,210 --> 00:06:29,750
We looked at tree growth
and mortality rates,

146
00:06:29,750 --> 00:06:31,370
five metrics of tree diversity,

147
00:06:31,370 --> 00:06:33,020
and then we also assessed land use

148
00:06:33,020 --> 00:06:35,360
and ownership patterns surrounding parks.

149
00:06:35,360 --> 00:06:36,920
So today I'm just gonna talk about

150
00:06:36,920 --> 00:06:39,820
the forest structure
and diversity analyses,

151
00:06:39,820 --> 00:06:42,230
but the papers are here
if you're interested

152
00:06:42,230 --> 00:06:44,023
in knowing more about the others.

153
00:06:46,100 --> 00:06:48,180
First, I will cover the highlights

154
00:06:48,180 --> 00:06:50,970
from our forest structure analysis.

155
00:06:50,970 --> 00:06:54,500
So on this figure, the
X axis has the 50 parks

156
00:06:54,500 --> 00:06:58,430
that were in the analysis,
sorted from low to high latitude.

157
00:06:58,430 --> 00:07:02,070
An asterisk indicates a
significance between the park

158
00:07:02,070 --> 00:07:05,550
and surrounding forest,
of P less than 0.05.

159
00:07:05,550 --> 00:07:08,340
Each park here has a pair
of points, with the green

160
00:07:08,340 --> 00:07:11,590
representing the parks
average and standard error,

161
00:07:11,590 --> 00:07:14,850
and the yellow representing
the surrounding forest average

162
00:07:14,850 --> 00:07:16,530
for that metric of interest.

163
00:07:16,530 --> 00:07:19,490
And what I want you to
take away from this is that

164
00:07:19,490 --> 00:07:21,500
in just about every case, the park has

165
00:07:21,500 --> 00:07:25,350
a higher live basal area
than surrounding forest.

166
00:07:25,350 --> 00:07:28,380
So the green dots are
above the yellow dots

167
00:07:28,380 --> 00:07:30,063
in just about every case.

168
00:07:31,200 --> 00:07:33,810
What's also interesting
here is that, on average,

169
00:07:33,810 --> 00:07:36,490
live basal area for
eastern old growth forest

170
00:07:36,490 --> 00:07:39,170
is 29 meters squared per hectare.

171
00:07:39,170 --> 00:07:42,290
And while I wouldn't say that
parks have old growth forest,

172
00:07:42,290 --> 00:07:44,550
it does appear that many
are at least approaching

173
00:07:44,550 --> 00:07:46,500
old growth levels of basal area,

174
00:07:46,500 --> 00:07:48,593
where surrounding forests are not.

175
00:07:50,320 --> 00:07:54,180
For density of live trees
over 30 centimeters DBH,

176
00:07:54,180 --> 00:07:56,440
again, most parks' average density

177
00:07:56,440 --> 00:07:58,650
is higher than the surrounding forest.

178
00:07:58,650 --> 00:08:00,970
And ecological thresholds from two studies

179
00:08:00,970 --> 00:08:03,560
on late-successional bird species indicate

180
00:08:03,560 --> 00:08:06,870
that these differences are
likely ecologically important,

181
00:08:06,870 --> 00:08:09,580
as all but a few of the
parks in the study meet

182
00:08:09,580 --> 00:08:12,350
the lower threshold and many
are above the higher threshold,

183
00:08:12,350 --> 00:08:14,133
where surrounding forests are not.

184
00:08:15,460 --> 00:08:17,720
Looking at coarse woody debris volume,

185
00:08:17,720 --> 00:08:20,350
we find that our parks
have considerably higher

186
00:08:20,350 --> 00:08:23,390
coarse woody debris volume
than their surrounding forests.

187
00:08:23,390 --> 00:08:26,153
And based on ecological
thresholds in the literature,

188
00:08:26,153 --> 00:08:29,250
many of our parks are meeting
at least a lower threshold,

189
00:08:29,250 --> 00:08:30,850
some meeting the higher threshold,

190
00:08:30,850 --> 00:08:33,143
where surrounding forests
generally are not.

191
00:08:35,420 --> 00:08:37,370
Okay, so now we're onto the results

192
00:08:37,370 --> 00:08:39,480
from the tree diversity study.

193
00:08:39,480 --> 00:08:41,740
So this figure is showing
percent difference

194
00:08:41,740 --> 00:08:43,670
between park and its surrounding forest

195
00:08:43,670 --> 00:08:47,070
for the five trees closest
to the center of each plot.

196
00:08:47,070 --> 00:08:48,350
If the symbol is above the line,

197
00:08:48,350 --> 00:08:50,060
the park average metric is higher.

198
00:08:50,060 --> 00:08:52,920
If it's below the line, its
surrounding forests were higher.

199
00:08:52,920 --> 00:08:55,630
An asterisk indicates
significant difference.

200
00:08:55,630 --> 00:08:57,690
So first looking at species richness,

201
00:08:57,690 --> 00:09:01,250
we found that 77% of parks had
higher tree species richness

202
00:09:01,250 --> 00:09:02,840
than surrounding forests.

203
00:09:02,840 --> 00:09:04,550
For Shannon evenness, which is a measure

204
00:09:04,550 --> 00:09:07,440
of how evenly distributed
species abundances are,

205
00:09:07,440 --> 00:09:11,010
61% of parks had higher evenness
than surrounding forest,

206
00:09:11,010 --> 00:09:12,640
so more diverse.

207
00:09:12,640 --> 00:09:14,690
For McNaughton dominance,
which is the measure

208
00:09:14,690 --> 00:09:15,930
of the relative proportion

209
00:09:15,930 --> 00:09:18,230
that the two most
abundant species take up,

210
00:09:18,230 --> 00:09:21,250
in this case lower McNaughton
dominance means more diverse,

211
00:09:21,250 --> 00:09:23,860
we had 74% of parks having lower dominance

212
00:09:23,860 --> 00:09:25,470
than surrounding forests.

213
00:09:25,470 --> 00:09:28,360
And then finally, looking
at percent rare species,

214
00:09:28,360 --> 00:09:32,100
in this case, rare species
is defined as species

215
00:09:32,100 --> 00:09:35,600
with low abundance, not rare,
threatened, and endangered.

216
00:09:35,600 --> 00:09:38,952
And for this study, we
found that 61% of parks

217
00:09:38,952 --> 00:09:41,863
had more rare species
than surrounding forests.

218
00:09:43,330 --> 00:09:46,800
Okay, so to put the results
of those two studies together,

219
00:09:46,800 --> 00:09:49,390
we found park forests were
consistently different

220
00:09:49,390 --> 00:09:51,210
from their surrounding forests

221
00:09:51,210 --> 00:09:54,030
in having higher proportion
of late successional forests,

222
00:09:54,030 --> 00:09:56,090
higher live tree basal area,

223
00:09:56,090 --> 00:09:58,080
higher density of large trees

224
00:09:58,080 --> 00:10:00,780
considerably higher coarse
woody debris volume,

225
00:10:00,780 --> 00:10:03,420
and also higher
stand-level tree diversity.

226
00:10:03,420 --> 00:10:05,620
And based on ecological thresholds

227
00:10:05,620 --> 00:10:08,280
that many of our parks met and
surrounding forests did not,

228
00:10:08,280 --> 00:10:11,420
the results are likely
ecologically significant.

229
00:10:11,420 --> 00:10:14,357
And so this was an important
finding in our parks.

230
00:10:14,357 --> 00:10:17,090
A lot of the eastern parks are cultural

231
00:10:17,090 --> 00:10:19,760
or battlefield parks,
where forests are generally

232
00:10:19,760 --> 00:10:22,080
not the focus to managers
or at least managers

233
00:10:22,080 --> 00:10:24,060
have a hard time justifying resources

234
00:10:24,060 --> 00:10:25,860
going towards forest management,

235
00:10:25,860 --> 00:10:28,480
because they're not central
to the parks' objectives.

236
00:10:28,480 --> 00:10:31,070
And by showing that these
parks are protecting

237
00:10:31,070 --> 00:10:33,730
important forest habitat,
managers have been able

238
00:10:33,730 --> 00:10:35,480
to make a stronger case for protecting

239
00:10:35,480 --> 00:10:38,330
and trying to improve the
condition of their park forest.

240
00:10:39,910 --> 00:10:41,200
That was the good news.

241
00:10:41,200 --> 00:10:42,640
Now for the bad news.

242
00:10:42,640 --> 00:10:46,688
There are some major issues
facing many of our park forests.

243
00:10:46,688 --> 00:10:51,410
The main ones being invasive
plants, overabundant deer,

244
00:10:51,410 --> 00:10:54,080
lacking or suboptimal regeneration,

245
00:10:54,080 --> 00:10:56,850
and forest pests and canopy disturbances.

246
00:10:56,850 --> 00:10:59,480
And many of these issues
are also interrelated

247
00:10:59,480 --> 00:11:01,870
and are together causing
more severe impacts

248
00:11:01,870 --> 00:11:03,453
than they would on their own.

249
00:11:05,890 --> 00:11:07,830
The first example I'll give of issues

250
00:11:07,830 --> 00:11:10,610
facing park forests is invasive plants.

251
00:11:10,610 --> 00:11:14,250
So we just published an analysis
of invasive plant trends

252
00:11:14,250 --> 00:11:16,370
in ecological applications.

253
00:11:16,370 --> 00:11:18,300
And this invasive trend analysis

254
00:11:18,300 --> 00:11:21,180
included 39 parks from five networks.

255
00:11:21,180 --> 00:11:23,760
So basically the parks you see in the map,

256
00:11:23,760 --> 00:11:25,810
we looked at trends over 12 years,

257
00:11:25,810 --> 00:11:28,090
which covered three cycles of monitoring,

258
00:11:28,090 --> 00:11:31,403
and included over 1,400 forest plots.

259
00:11:33,790 --> 00:11:37,150
For this analysis, we assessed
trends in multiple metrics

260
00:11:37,150 --> 00:11:39,660
of abundance and grouping
levels to determine

261
00:11:39,660 --> 00:11:43,080
how invasive plants are
changing over time in our parks.

262
00:11:43,080 --> 00:11:45,280
So first we looked at
plot percent frequency,

263
00:11:45,280 --> 00:11:47,280
which was the percent
of plots within a park

264
00:11:47,280 --> 00:11:49,670
where at least one invasive was found.

265
00:11:49,670 --> 00:11:51,090
We looked at quadrat frequency,

266
00:11:51,090 --> 00:11:52,500
which was the percent of quadrats

267
00:11:52,500 --> 00:11:54,037
where at one invasive was found.

268
00:11:54,037 --> 00:11:56,860
And so if you look at the NETN
plot diagram on the right,

269
00:11:56,860 --> 00:11:58,900
those are the little blue squares.

270
00:11:58,900 --> 00:12:00,560
We also looked at quadrat percent cover,

271
00:12:00,560 --> 00:12:01,700
which was the average cover

272
00:12:01,700 --> 00:12:04,585
of invasive in the quadrats, again.

273
00:12:04,585 --> 00:12:06,540
We looked at different grouping levels

274
00:12:06,540 --> 00:12:07,870
and how they changed over time.

275
00:12:07,870 --> 00:12:09,730
So that included total invasives,

276
00:12:09,730 --> 00:12:12,260
which were all invasive species combined.

277
00:12:12,260 --> 00:12:13,770
We also looked at guilds.

278
00:12:13,770 --> 00:12:16,260
So we looked at trees, shrubs/vines,

279
00:12:16,260 --> 00:12:18,130
graminoids, which are grasses and sedges,

280
00:12:18,130 --> 00:12:19,360
and then herbaceous species,

281
00:12:19,360 --> 00:12:21,670
which are non-woody species
that aren't graminoids.

282
00:12:21,670 --> 00:12:24,323
And then we also looked
at species-level trends.

283
00:12:26,140 --> 00:12:29,282
So we first ranked parks
by level of invasion

284
00:12:29,282 --> 00:12:31,130
based on the abundance metrics,

285
00:12:31,130 --> 00:12:33,730
sorted first by average percent cover.

286
00:12:33,730 --> 00:12:36,660
And we used the most
recent four years of data,

287
00:12:36,660 --> 00:12:38,970
so 2015 to 2018.

288
00:12:38,970 --> 00:12:40,400
So the table here includes

289
00:12:40,400 --> 00:12:42,700
the 10 most and 10 least invaded parks

290
00:12:42,700 --> 00:12:45,120
out of the 39 in our analysis.

291
00:12:45,120 --> 00:12:47,543
So this is not where you
want to be number one.

292
00:12:47,543 --> 00:12:49,560
So the two most invaded parks were

293
00:12:49,560 --> 00:12:53,230
Monocacy and Antietam National
Battlefields in Maryland.

294
00:12:53,230 --> 00:12:55,120
And the 10 most invaded parks,

295
00:12:55,120 --> 00:12:57,320
you can see that their average cover was

296
00:12:57,320 --> 00:13:00,200
between 20 and 40% invasive cover.

297
00:13:00,200 --> 00:13:03,650
That doesn't mean the other
60 to 80% was native cover.

298
00:13:03,650 --> 00:13:06,100
More often, where deer are a problem,

299
00:13:06,100 --> 00:13:08,563
it's bare grounds because
of soil compaction.

300
00:13:09,690 --> 00:13:12,900
And even looking at the
10 least invaded parks

301
00:13:12,900 --> 00:13:15,700
in this study, a number of them still have

302
00:13:15,700 --> 00:13:17,966
pretty high quadrat and plot frequency.

303
00:13:17,966 --> 00:13:22,150
And it's just that their
actual cover is pretty low.

304
00:13:22,150 --> 00:13:24,150
So if you look at
Marsh-Billings, for example,

305
00:13:24,150 --> 00:13:26,100
it's the second one from the bottom,

306
00:13:26,100 --> 00:13:28,478
the average percent cover is less than

307
00:13:28,478 --> 00:13:31,969
1/10 of a percent in the
quadrats, but yet we found

308
00:13:31,969 --> 00:13:36,170
at least one invasive species
in over 80% of the plots.

309
00:13:36,170 --> 00:13:39,880
And Acadia is really
the main exception here.

310
00:13:39,880 --> 00:13:43,250
And they had low abundance
across all metrics in Acadia.

311
00:13:43,250 --> 00:13:46,483
And so Acadia was the least
invaded park in this analysis.

312
00:13:48,470 --> 00:13:50,950
We also ranks species based
on how common they were

313
00:13:50,950 --> 00:13:52,510
across parks and plots.

314
00:13:52,510 --> 00:13:56,240
So Japanese stilt grass
here was the big winner.

315
00:13:56,240 --> 00:13:58,850
Japanese stilt grass was
the most widespread species

316
00:13:58,850 --> 00:14:03,850
found in over 80% of the plot
parks, over 30% of the plots.

317
00:14:03,930 --> 00:14:07,740
And the range of microstegium
is still fairly limited

318
00:14:07,740 --> 00:14:10,810
to Southern New England
into the mid-Atlantic,

319
00:14:10,810 --> 00:14:12,840
but it is moving its way forward

320
00:14:12,840 --> 00:14:15,393
into the northern states in this analysis.

321
00:14:16,820 --> 00:14:18,580
The next most common species

322
00:14:18,580 --> 00:14:21,910
are primarily woody shrubs and vines.

323
00:14:21,910 --> 00:14:24,260
So multiflora rose, Japanese honeysuckle,

324
00:14:24,260 --> 00:14:26,490
Japanese barberry, Asian bittersweet.

325
00:14:26,490 --> 00:14:28,780
These are all species that were common

326
00:14:28,780 --> 00:14:32,290
throughout the range that
we looked at in our study

327
00:14:32,290 --> 00:14:34,913
and were pretty common
throughout our parks.

328
00:14:36,700 --> 00:14:38,640
Now we're gonna look at trends.

329
00:14:38,640 --> 00:14:40,710
So in the figures I'm gonna show,

330
00:14:40,710 --> 00:14:43,120
each park is gonna be
represented by a line.

331
00:14:43,120 --> 00:14:45,640
So there'll be 39 lines.

332
00:14:45,640 --> 00:14:48,290
Red is a significant increase over time.

333
00:14:48,290 --> 00:14:50,877
Blue is a significant decrease over time.

334
00:14:50,877 --> 00:14:53,840
And gray is nonsignificant trends.

335
00:14:53,840 --> 00:14:56,760
And here we're looking at
how quadrat percent frequency

336
00:14:56,760 --> 00:14:59,053
of total invasives changes over time.

337
00:15:01,531 --> 00:15:02,930
And so you should notice there are

338
00:15:02,930 --> 00:15:06,200
quite a few significant
increases in quadrat frequency,

339
00:15:06,200 --> 00:15:07,726
not a lot of declines.

340
00:15:07,726 --> 00:15:09,820
And even the nonsignificant trends

341
00:15:09,820 --> 00:15:11,580
were mostly trending upward.

342
00:15:11,580 --> 00:15:15,214
The only significant decline
we found in quadrat frequency

343
00:15:15,214 --> 00:15:18,240
was Prince William Forest
Park, that's the blue line.

344
00:15:18,240 --> 00:15:20,023
And that park is in Virginia.

345
00:15:22,550 --> 00:15:25,270
Now looking at average
cover of total invasives,

346
00:15:25,270 --> 00:15:27,670
we have even more significant increases

347
00:15:27,670 --> 00:15:31,130
and steeper increases in
quadrat percent cover.

348
00:15:31,130 --> 00:15:33,700
And again, only one significant decline,

349
00:15:33,700 --> 00:15:36,320
that was Roosevelt-Vanderbilt
National Historical Sites

350
00:15:36,320 --> 00:15:37,603
in New York.

351
00:15:40,210 --> 00:15:41,770
Okay, so the next figures are gonna be

352
00:15:41,770 --> 00:15:44,140
looking at species-specific trends.

353
00:15:44,140 --> 00:15:47,490
So in this plot, the
circle represents the slope

354
00:15:47,490 --> 00:15:49,130
or percent change over time

355
00:15:49,130 --> 00:15:52,120
in species-level quadrat frequency.

356
00:15:52,120 --> 00:15:56,040
The error bars are 95%
empirical confidence intervals.

357
00:15:56,040 --> 00:16:00,180
Values that are to the right
of the zero vertical line

358
00:16:00,180 --> 00:16:02,180
indicate significant increases.

359
00:16:02,180 --> 00:16:05,090
Values to the left are
significant declines.

360
00:16:05,090 --> 00:16:06,760
And this is sorted by species,

361
00:16:06,760 --> 00:16:09,200
with the most number
of significant trends.

362
00:16:09,200 --> 00:16:12,390
And in case it helps,
they're color-coded by guild.

363
00:16:12,390 --> 00:16:16,540
So graminoids are yellow, herbs
are green, trees are blue,

364
00:16:16,540 --> 00:16:19,330
and then shrubs are
shades of red and orange.

365
00:16:19,330 --> 00:16:21,030
And looking at this figure,

366
00:16:21,030 --> 00:16:24,050
you can see there are a lot
more significant increases

367
00:16:24,050 --> 00:16:26,020
in quadrat frequency at the species-level.

368
00:16:26,020 --> 00:16:29,440
So many more on the right
side of the vertical line.

369
00:16:29,440 --> 00:16:31,540
Not so many on the left side.

370
00:16:31,540 --> 00:16:35,688
And Japanese stilt grass,
microstegium vinineum,

371
00:16:35,688 --> 00:16:39,400
had the most number of
increases among parks

372
00:16:39,400 --> 00:16:43,090
and had the highest recorded increase.

373
00:16:43,090 --> 00:16:46,330
And there's also an
interesting interaction

374
00:16:46,330 --> 00:16:48,270
that we caught in a couple of our parks

375
00:16:48,270 --> 00:16:50,540
between herbaceous and woody species.

376
00:16:50,540 --> 00:16:52,564
So in Morristown, for example,

377
00:16:52,564 --> 00:16:56,240
we have garlic mustard,
which is alliaria petiolata,

378
00:16:56,240 --> 00:16:59,170
and cardamine impatiens,
which is narrowly bittercress.

379
00:16:59,170 --> 00:17:02,990
Both of those had a decline in
quadrat frequency over time.

380
00:17:02,990 --> 00:17:06,640
But it was countered by
an increase in wineberry,

381
00:17:06,640 --> 00:17:09,160
which is rubus phoenicolasius.

382
00:17:09,160 --> 00:17:11,940
In Roosevelt-Vanderbilt,
we had a similar pattern

383
00:17:11,940 --> 00:17:14,210
where garlic mustard went down,

384
00:17:14,210 --> 00:17:19,210
but Norway maple, acer
platanoides, increased over time.

385
00:17:19,730 --> 00:17:21,920
So there wasn't an actual decline

386
00:17:23,379 --> 00:17:26,393
overall of invasives
because of this interaction.

387
00:17:27,310 --> 00:17:31,810
And in average cover, we
find similar patterns.

388
00:17:31,810 --> 00:17:33,926
So a lot more significant increases.

389
00:17:33,926 --> 00:17:35,890
Japanese stilt grass, again,

390
00:17:35,890 --> 00:17:38,890
has the most number of
significant increases.

391
00:17:38,890 --> 00:17:42,391
And again, we're seeing interaction

392
00:17:42,391 --> 00:17:45,290
between herbaceous and woody species.

393
00:17:45,290 --> 00:17:50,010
So in Morristown, for
example, Japanese stilt grass

394
00:17:50,010 --> 00:17:52,710
and garlic mustard declined significantly

395
00:17:52,710 --> 00:17:54,410
in average cover over time.

396
00:17:54,410 --> 00:17:57,200
But it was met with an
almost equal increase

397
00:17:57,200 --> 00:18:00,920
in Japanese barberry, which
is berberis thunbergii.

398
00:18:00,920 --> 00:18:03,820
And Valley Forge, which
is in Pennsylvania,

399
00:18:03,820 --> 00:18:05,330
we found a similar interaction,

400
00:18:05,330 --> 00:18:07,450
where Japanese stilt grass went down,

401
00:18:07,450 --> 00:18:10,780
but then we saw an increase
in actually garlic mustard,

402
00:18:10,780 --> 00:18:12,710
but then also a bunch of woody species.

403
00:18:12,710 --> 00:18:15,863
So Asian bittersweet,
wineberry, and tree of heaven.

404
00:18:17,900 --> 00:18:20,550
so when we first started seeing
these patterns in the data

405
00:18:20,550 --> 00:18:23,100
we wanted to know whether
this was some kind of artifact

406
00:18:23,100 --> 00:18:24,550
of how we're collecting the data?

407
00:18:24,550 --> 00:18:28,288
Or are we getting better at
identifying invasive species?

408
00:18:28,288 --> 00:18:31,440
Or if these reflected real
patterns on the ground?

409
00:18:31,440 --> 00:18:32,980
So we looked at our photo points

410
00:18:32,980 --> 00:18:34,330
from our plots for evidence,

411
00:18:34,330 --> 00:18:36,580
which this is an example of.

412
00:18:36,580 --> 00:18:39,210
So this is a plot in
Morristown, in New Jersey,

413
00:18:39,210 --> 00:18:41,390
that we established in 2009.

414
00:18:41,390 --> 00:18:43,230
This is what one of the scenes looked like

415
00:18:43,230 --> 00:18:45,490
the first time we sampled this plot.

416
00:18:45,490 --> 00:18:47,770
So you can see that
there's Japanese barberry

417
00:18:47,770 --> 00:18:50,130
in kind of the back half of the plot.

418
00:18:50,130 --> 00:18:51,960
The foreground of the plot,

419
00:18:51,960 --> 00:18:53,789
there's not a whole lot going on.

420
00:18:53,789 --> 00:18:57,650
The little green sprigs are
actually Japanese stilt grass,

421
00:18:57,650 --> 00:18:59,360
kind of early in the season.

422
00:18:59,360 --> 00:19:01,030
And then you do see a little bit

423
00:19:01,030 --> 00:19:02,580
of barberry in the foreground,

424
00:19:02,580 --> 00:19:06,763
and some wineberry in the right
of the screen, bottom right.

425
00:19:09,450 --> 00:19:12,083
Four years later, when
we sampled this plot,

426
00:19:13,100 --> 00:19:16,560
the barberry has filled in
a lot in the foreground.

427
00:19:16,560 --> 00:19:18,560
The wineberry in the bottom right

428
00:19:18,560 --> 00:19:20,760
has more than doubled in size.

429
00:19:20,760 --> 00:19:22,570
And there's really not a whole lot

430
00:19:22,570 --> 00:19:25,390
of bare grounds in the picture.

431
00:19:25,390 --> 00:19:28,730
And then when we last
sampled this plot in 2017,

432
00:19:28,730 --> 00:19:30,730
it's almost entirely covered

433
00:19:30,730 --> 00:19:33,730
by Japanese barberry and wineberry.

434
00:19:33,730 --> 00:19:36,870
And there's almost no Japanese
stilt grass to be found

435
00:19:36,870 --> 00:19:38,540
and very little bare ground.

436
00:19:38,540 --> 00:19:41,830
And I should mention that this
is a closed canopy forest.

437
00:19:41,830 --> 00:19:44,200
So this is not like a
forest gap was created.

438
00:19:44,200 --> 00:19:48,380
And so the invasive shrubs
responded to increased light.

439
00:19:48,380 --> 00:19:51,053
This is just what happened
in a closed canopy forest.

440
00:19:55,080 --> 00:19:57,040
Moving onto Valley Forge.

441
00:19:57,040 --> 00:20:01,790
This is a picture from a plot
that we established in 2007.

442
00:20:01,790 --> 00:20:03,410
If you remember in Valley Forge,

443
00:20:03,410 --> 00:20:05,360
we saw a decline in Japanese stilt grass

444
00:20:05,360 --> 00:20:07,280
and an increase of tree of heaven.

445
00:20:07,280 --> 00:20:10,290
And so here, in 2007, the understory

446
00:20:10,290 --> 00:20:13,480
is almost entirely dominated
by Japanese stilt grass.

447
00:20:13,480 --> 00:20:15,380
So if you've never seen
Japanese stilt grass,

448
00:20:15,380 --> 00:20:16,880
this is what it can look like.

449
00:20:18,300 --> 00:20:21,170
So four years later, in 2011,

450
00:20:21,170 --> 00:20:24,500
we see there's some seedlings
coming in in the understory.

451
00:20:24,500 --> 00:20:26,960
There's still a lot of
Japanese stilt grass.

452
00:20:26,960 --> 00:20:31,560
And I should mention that deer
management started in 2010.

453
00:20:31,560 --> 00:20:34,780
So in 2011, when we sampled this plot,

454
00:20:34,780 --> 00:20:37,630
we're starting to see some
response in the understory.

455
00:20:37,630 --> 00:20:40,350
But unfortunately it's
mostly tree of heaven.

456
00:20:40,350 --> 00:20:43,390
Almost all of these
seedlings are tree of heaven.

457
00:20:43,390 --> 00:20:46,271
Four years later, the tree
of heaven is much taller

458
00:20:46,271 --> 00:20:49,083
and taking up a lot more space.

459
00:20:50,860 --> 00:20:52,880
Four years later, which is the last time

460
00:20:52,880 --> 00:20:55,470
we sampled this plot, in 2019

461
00:20:55,470 --> 00:20:58,460
you can see that the tree
of heaven is even bigger,

462
00:20:58,460 --> 00:21:01,190
taking up a lot more space
and there's not as much

463
00:21:01,190 --> 00:21:04,330
Japanese stilt grass in the understory.

464
00:21:04,330 --> 00:21:06,100
And so all of this is to show

465
00:21:06,100 --> 00:21:07,440
that we have pretty high confidence

466
00:21:07,440 --> 00:21:09,130
that the trends that
we're seeing in the data

467
00:21:09,130 --> 00:21:12,563
do in fact reflect real
trends on the ground.

468
00:21:14,640 --> 00:21:17,993
Okay, so just to summarize
what we found here.

469
00:21:17,993 --> 00:21:21,060
We found that invasive
species, they're widespread

470
00:21:21,060 --> 00:21:23,610
and consistently increasing in abundance

471
00:21:23,610 --> 00:21:25,323
across our eastern parks.

472
00:21:26,300 --> 00:21:29,490
35 out of 39 parks had invasives

473
00:21:29,490 --> 00:21:31,393
in at least half of their plots.

474
00:21:32,290 --> 00:21:34,970
30 parks had at least
one significant increase

475
00:21:34,970 --> 00:21:36,533
in invasive abundance.

476
00:21:37,460 --> 00:21:39,660
Only two parks showed overall declines.

477
00:21:39,660 --> 00:21:41,950
That was Prince William
Forest Park in Virginia

478
00:21:41,950 --> 00:21:43,280
and Marsh-Billings-Rockefeller

479
00:21:43,280 --> 00:21:46,060
National Historical Park in Vermont.

480
00:21:46,060 --> 00:21:50,280
I don't know a whole lot
about the management history

481
00:21:50,280 --> 00:21:51,530
of Prince William Forest Park,

482
00:21:51,530 --> 00:21:53,820
but I do know that in
Marsh-Billings, they have been

483
00:21:53,820 --> 00:21:58,240
managing invasive species
intensively for over 10 years.

484
00:21:58,240 --> 00:22:00,630
And so I think that the
decline we're seeing,

485
00:22:00,630 --> 00:22:01,870
in Marsh-Billings at least,

486
00:22:01,870 --> 00:22:06,517
is because of their management
efforts to remove invasives.

487
00:22:08,030 --> 00:22:09,680
There were also two parks that kept

488
00:22:09,680 --> 00:22:12,260
relatively low invasive levels over time.

489
00:22:12,260 --> 00:22:14,780
And that was Acadia National Park in Maine

490
00:22:14,780 --> 00:22:18,313
and Saint-Gaudens National
Historical Park in New Hampshire.

491
00:22:20,980 --> 00:22:23,600
The most aggressive invaders
were Japanese stilt grass

492
00:22:23,600 --> 00:22:25,580
and shrubs and woody vines.

493
00:22:25,580 --> 00:22:29,720
The likely drivers were
overabundant deer, forest pests,

494
00:22:29,720 --> 00:22:31,800
and other canopy disturbances,

495
00:22:31,800 --> 00:22:35,140
exotic earthworms, fragmentation,
and land use history.

496
00:22:35,140 --> 00:22:36,630
And I highlight these three

497
00:22:36,630 --> 00:22:38,540
because they seem to be
contributing the most

498
00:22:38,540 --> 00:22:40,620
to the invasive problems these days,

499
00:22:40,620 --> 00:22:42,430
and are something that managers

500
00:22:42,430 --> 00:22:43,827
may be able to do something about.

501
00:22:43,827 --> 00:22:46,203
And so I'm gonna elaborate
more on that next.

502
00:22:48,140 --> 00:22:51,100
Deer are a major problem
in many of our parks,

503
00:22:51,100 --> 00:22:53,560
not in Acadia, but pretty
much everywhere else.

504
00:22:53,560 --> 00:22:55,340
And we're finding that the impacts

505
00:22:55,340 --> 00:22:57,080
from chronic deer overabundance

506
00:22:57,080 --> 00:22:59,510
continue to increase over time.

507
00:22:59,510 --> 00:23:02,670
And deer eat native seedlings and herbs.

508
00:23:02,670 --> 00:23:04,690
They facilitate invasive species

509
00:23:04,690 --> 00:23:06,870
through seed dispersal
and soil compaction.

510
00:23:06,870 --> 00:23:09,280
And then also obviously
eating the native species

511
00:23:09,280 --> 00:23:11,620
and making room for the invasives.

512
00:23:11,620 --> 00:23:14,640
And most parks do not allow deer hunting

513
00:23:14,640 --> 00:23:16,840
as part of their enabling legislation.

514
00:23:16,840 --> 00:23:18,500
And so they can't just open the park

515
00:23:18,500 --> 00:23:20,780
to hunting to reduce the deer herd.

516
00:23:20,780 --> 00:23:22,000
And so in most cases,

517
00:23:22,000 --> 00:23:24,480
the parks that are actively managing deer

518
00:23:24,480 --> 00:23:27,820
are hiring APHIS sharpshooters
to reduce the herd.

519
00:23:27,820 --> 00:23:29,730
And it's really expensive.

520
00:23:29,730 --> 00:23:31,920
And pretty much as soon as a park decides

521
00:23:31,920 --> 00:23:34,870
to go down this path, they
basically have to do it forever

522
00:23:34,870 --> 00:23:38,040
to keep deer numbers in check.

523
00:23:38,040 --> 00:23:40,380
And even getting to the
point where parks are allowed

524
00:23:40,380 --> 00:23:44,090
to manage deer is a huge
undertaking, requiring many years

525
00:23:44,090 --> 00:23:46,160
to develop environmental
impact statements.

526
00:23:46,160 --> 00:23:47,970
And each park, at this point,

527
00:23:47,970 --> 00:23:50,130
has to go through the same process.

528
00:23:50,130 --> 00:23:52,160
And parks are often sued by hunters

529
00:23:52,160 --> 00:23:54,707
who want to actually
be in the parks hunting

530
00:23:54,707 --> 00:23:56,410
and animal rights activists

531
00:23:56,410 --> 00:24:00,400
that don't want the park
to be killing animals.

532
00:24:00,400 --> 00:24:03,617
And so that can really
slow things down as well.

533
00:24:03,617 --> 00:24:05,820
As more parks are going down this path,

534
00:24:05,820 --> 00:24:09,140
our hope is that the approval
process can happen faster

535
00:24:09,140 --> 00:24:11,060
and management becomes easier.

536
00:24:11,060 --> 00:24:13,313
But we've still got a long way to go.

537
00:24:14,470 --> 00:24:15,850
So looking at the parks

538
00:24:15,850 --> 00:24:19,720
that have the most impacts from deer,

539
00:24:19,720 --> 00:24:24,060
we see this area from
Saratoga up in New York,

540
00:24:24,060 --> 00:24:27,400
down through Colonial
in Southern Virginia.

541
00:24:27,400 --> 00:24:28,260
This is where we see

542
00:24:28,260 --> 00:24:31,340
the highest deer impacts in our parks.

543
00:24:31,340 --> 00:24:35,110
And looking at the
regeneration stocking index,

544
00:24:35,110 --> 00:24:37,750
which was developed by the
Northern Research Station,

545
00:24:37,750 --> 00:24:40,810
with the Forest Service,
we also find that the parks

546
00:24:40,810 --> 00:24:43,210
that have the least percent of plots

547
00:24:43,210 --> 00:24:44,930
that are stocked with regeneration

548
00:24:44,930 --> 00:24:47,840
are also in the area with
high deer browse impacts.

549
00:24:47,840 --> 00:24:51,178
And in fact, there are
five parks in this circle

550
00:24:51,178 --> 00:24:53,170
that don't have a single plot

551
00:24:53,170 --> 00:24:55,063
that has sufficient regeneration.

552
00:24:57,750 --> 00:25:01,900
And even the parks that do have

553
00:25:01,900 --> 00:25:03,560
an abundant regeneration layer,

554
00:25:03,560 --> 00:25:05,970
like Gettysburg, pictured here,

555
00:25:05,970 --> 00:25:07,970
a closer look at the composition

556
00:25:07,970 --> 00:25:10,350
of the regeneration
layer is also concerning.

557
00:25:10,350 --> 00:25:13,680
So this map is of our
forest plots in Gettysburg.

558
00:25:13,680 --> 00:25:15,357
And each pie is a forest plot.

559
00:25:15,357 --> 00:25:17,162
And the size of the plots is relative

560
00:25:17,162 --> 00:25:20,840
to the total seedling
density for that plot.

561
00:25:20,840 --> 00:25:22,470
And the colors represent the species

562
00:25:22,470 --> 00:25:24,360
making up the seedling layer.

563
00:25:24,360 --> 00:25:27,080
And as you can see here,
most of the regeneration

564
00:25:27,080 --> 00:25:30,770
in Gettysburg is dominated
by ash, which is brown,

565
00:25:30,770 --> 00:25:33,550
or low canopy species, which are in pink.

566
00:25:33,550 --> 00:25:36,390
And here, low canopy species are species

567
00:25:36,390 --> 00:25:38,450
like flowering dogwoods, sassafras,

568
00:25:38,450 --> 00:25:40,920
and ironwood, or red bud,

569
00:25:40,920 --> 00:25:43,190
which they're species
that have a tree-like form

570
00:25:43,190 --> 00:25:44,560
but they don't get very tall

571
00:25:44,560 --> 00:25:47,190
and they don't really
reach into the canopy.

572
00:25:47,190 --> 00:25:50,200
And both are not optimal
species for maintaining

573
00:25:50,200 --> 00:25:52,423
a closed canopy forest.

574
00:25:55,420 --> 00:25:58,110
Forest pests, especially
emerald ash borer now,

575
00:25:58,110 --> 00:26:00,460
and gaps formed by intense storms,

576
00:26:00,460 --> 00:26:02,330
are also contributing to the invasive

577
00:26:02,330 --> 00:26:04,560
and regeneration problems in our parks.

578
00:26:04,560 --> 00:26:07,770
So this is a photo of the
canopy thinning, in Morristown,

579
00:26:07,770 --> 00:26:10,090
as ash dieback from emerald ash borer.

580
00:26:10,090 --> 00:26:12,710
But this could easily be from
other parks in the region

581
00:26:12,710 --> 00:26:14,440
where emerald ash borer has hit.

582
00:26:14,440 --> 00:26:17,150
And the understory response
to the canopy thinning

583
00:26:17,150 --> 00:26:20,100
and the gaps, unfortunately,
in these areas,

584
00:26:20,100 --> 00:26:22,890
is dominated by increases
in invasive species,

585
00:26:22,890 --> 00:26:24,697
particularly invasive shrubs,

586
00:26:24,697 --> 00:26:27,493
with little response in tree regeneration.

587
00:26:29,100 --> 00:26:31,850
When deer and invasive
problems go on for long enough,

588
00:26:31,850 --> 00:26:33,900
the end result, which
we're seeing more of,

589
00:26:33,900 --> 00:26:36,391
is the formation of these
invasive shrub thickets.

590
00:26:36,391 --> 00:26:38,870
This is a photo from the gap in Morristown

591
00:26:38,870 --> 00:26:42,080
that was formed from
Hurricane Sandy in 2012.

592
00:26:42,080 --> 00:26:44,000
We're also seeing gaps form like this

593
00:26:44,000 --> 00:26:46,380
in ash stands in some of our parks.

594
00:26:46,380 --> 00:26:47,370
And believe it or not,

595
00:26:47,370 --> 00:26:49,790
there are actually four
people in this photo.

596
00:26:49,790 --> 00:26:53,380
This was taken in 2017 when
we last sampled the plot.

597
00:26:53,380 --> 00:26:56,410
And the only regeneration
we found in this plot

598
00:26:56,410 --> 00:26:57,820
was staghorn sumac,

599
00:26:57,820 --> 00:26:59,970
which is the flagged sapling on the left.

600
00:26:59,970 --> 00:27:01,860
No canopy forming tree species

601
00:27:01,860 --> 00:27:03,360
were regenerating in this stand.

602
00:27:03,360 --> 00:27:07,180
And it's really hard to
imagine this being anything but

603
00:27:07,180 --> 00:27:10,100
exotic shrub thicket, long-term.

604
00:27:10,100 --> 00:27:13,040
And many of the forests
in our eastern parks

605
00:27:13,040 --> 00:27:16,340
are just one major disturbance
away from becoming this.

606
00:27:16,340 --> 00:27:18,960
And this is obviously
bad for forest health

607
00:27:18,960 --> 00:27:20,510
and for all the biodiversity

608
00:27:20,510 --> 00:27:22,720
that depends on forest habitats.

609
00:27:22,720 --> 00:27:24,540
But it's also bad for human health.

610
00:27:24,540 --> 00:27:26,530
So studies have found that these thickets

611
00:27:26,530 --> 00:27:28,340
can host more than double the density

612
00:27:28,340 --> 00:27:30,770
of black-legged ticks
than uninvaded forests,

613
00:27:30,770 --> 00:27:33,583
and the ticks are more
likely to carry Lyme disease.

614
00:27:34,760 --> 00:27:38,830
So you've seen the issues that
are facing many of our parks.

615
00:27:38,830 --> 00:27:40,530
One question that we had as we were

616
00:27:40,530 --> 00:27:42,510
starting to document these patterns,

617
00:27:42,510 --> 00:27:44,230
was whether it's just in our parks,

618
00:27:44,230 --> 00:27:48,260
or if this is part of a more
widespread, regional issue?

619
00:27:48,260 --> 00:27:50,593
And so that's what I'm
gonna talk about next.

620
00:27:54,200 --> 00:27:57,510
So to look at regional
regeneration patterns

621
00:27:57,510 --> 00:28:02,150
we used FIA data from
17 eastern U.S. states

622
00:28:02,150 --> 00:28:06,090
to quantify regeneration
debt in eastern forests.

623
00:28:06,090 --> 00:28:07,740
We coined the term regeneration debt

624
00:28:07,740 --> 00:28:09,340
to collectively describe issues

625
00:28:09,340 --> 00:28:11,440
of either not enough regeneration

626
00:28:11,440 --> 00:28:14,000
or mismatch in species composition

627
00:28:14,000 --> 00:28:17,300
between the regeneration
layer and the canopy species.

628
00:28:17,300 --> 00:28:19,640
So a regeneration debt
occurs if you either

629
00:28:19,640 --> 00:28:22,560
don't have enough regeneration
or if the composition

630
00:28:22,560 --> 00:28:25,300
of the regeneration
doesn't match the canopy.

631
00:28:25,300 --> 00:28:27,680
And regeneration debt is
not always a bad thing.

632
00:28:27,680 --> 00:28:29,750
So if you think of a forest succeeding

633
00:28:29,750 --> 00:28:32,150
from aspen and birch to spruce fir,

634
00:28:32,150 --> 00:28:34,680
the regeneration debt here
merely predicts a change

635
00:28:34,680 --> 00:28:37,690
in the future composition of
the canopy and is not bad.

636
00:28:37,690 --> 00:28:40,770
However, regeneration debt
can indicate a problem

637
00:28:40,770 --> 00:28:43,450
where anthropogenic stressors
are driving the patterns

638
00:28:43,450 --> 00:28:45,930
and you have suboptimal
species regenerating,

639
00:28:45,930 --> 00:28:48,350
which I'll talk about in a minute.

640
00:28:48,350 --> 00:28:51,450
And in addition to
quantifying regeneration debt,

641
00:28:51,450 --> 00:28:54,240
we related these findings
to anthropogenic stressors

642
00:28:54,240 --> 00:28:57,010
and metrics of climate
change using model selection

643
00:28:57,010 --> 00:28:58,853
to determine the likely drivers.

644
00:29:00,890 --> 00:29:04,250
Okay, so these are some of
the results from our analysis.

645
00:29:04,250 --> 00:29:07,330
We're looking at seedling
density across the region.

646
00:29:07,330 --> 00:29:09,930
And again, this was using FIA data.

647
00:29:09,930 --> 00:29:12,160
And so you'll notice that in Maine

648
00:29:12,160 --> 00:29:13,210
and Northern New Hampshire,

649
00:29:13,210 --> 00:29:15,340
seedling densities are really high.

650
00:29:15,340 --> 00:29:17,610
But if you look in the coastal region,

651
00:29:17,610 --> 00:29:20,968
from Massachusetts down
to DC and Virginia,

652
00:29:20,968 --> 00:29:24,113
there's some areas with
really low seedling density.

653
00:29:25,190 --> 00:29:27,480
So zooming in closer to this area

654
00:29:27,480 --> 00:29:29,800
for Massachusetts down to Virginia,

655
00:29:29,800 --> 00:29:32,810
there are areas, the darkest gray hatches,

656
00:29:32,810 --> 00:29:36,130
where there's less than one
seedling per 10 square meters.

657
00:29:36,130 --> 00:29:37,670
And then the lighter gray hatches

658
00:29:37,670 --> 00:29:40,890
are less than one seedling
for four square meters.

659
00:29:40,890 --> 00:29:42,760
That's really, really low.

660
00:29:42,760 --> 00:29:45,460
And so next, I'm gonna walk
you through the metrics

661
00:29:45,460 --> 00:29:47,440
that best predicted these patterns

662
00:29:47,440 --> 00:29:48,793
based on model selection.

663
00:29:50,000 --> 00:29:52,940
So average invasive cover in these plots,

664
00:29:52,940 --> 00:29:55,520
invasive plant cover, was
an important predictor

665
00:29:55,520 --> 00:29:56,950
for low seedling densities.

666
00:29:56,950 --> 00:29:59,380
You can see strong overlap
between the hatches

667
00:29:59,380 --> 00:30:01,100
where there's low seedling density

668
00:30:01,100 --> 00:30:03,000
and where there's high invasive cover.

669
00:30:04,210 --> 00:30:06,360
Looking at deer browse impacts.

670
00:30:06,360 --> 00:30:08,030
Unfortunately, this is only collected

671
00:30:08,030 --> 00:30:09,480
in the Northern Research Station,

672
00:30:09,480 --> 00:30:11,400
so I don't have data for Virginia.

673
00:30:11,400 --> 00:30:13,760
But where we have low seedling densities,

674
00:30:13,760 --> 00:30:17,303
we also have high deer
browse impacts, in general.

675
00:30:18,660 --> 00:30:20,323
And then finally looking at

676
00:30:20,323 --> 00:30:23,120
degree of human modified land cover,

677
00:30:23,120 --> 00:30:25,910
at the 300 meter square scale.

678
00:30:25,910 --> 00:30:27,580
That was also an important predictor.

679
00:30:27,580 --> 00:30:30,110
So where there's high modified land cover,

680
00:30:30,110 --> 00:30:32,420
so basically non-forest land cover,

681
00:30:32,420 --> 00:30:35,143
we have low seedling densities.

682
00:30:38,240 --> 00:30:40,790
Next we looked at composition between

683
00:30:40,790 --> 00:30:43,840
the regeneration and canopy
using similarity metrics,

684
00:30:43,840 --> 00:30:46,090
which are plotted on the map on the right.

685
00:30:46,090 --> 00:30:48,320
So similarity is a measure of how similar

686
00:30:48,320 --> 00:30:50,680
species composition is between two groups,

687
00:30:50,680 --> 00:30:53,360
in this case, the seedling
versus canopy strata.

688
00:30:53,360 --> 00:30:54,910
So if you have all the same species

689
00:30:54,910 --> 00:30:56,550
in the seedling and canopy layer,

690
00:30:56,550 --> 00:30:58,510
then your similarity will be one.

691
00:30:58,510 --> 00:31:00,140
If you have no species in common,

692
00:31:00,140 --> 00:31:02,010
then similarity will be zero.

693
00:31:02,010 --> 00:31:03,990
And in this case, we
used Horn's similarity,

694
00:31:03,990 --> 00:31:07,260
which also incorporates species
abundance into the equation.

695
00:31:07,260 --> 00:31:09,960
And the main takeaway
here is that the same area

696
00:31:09,960 --> 00:31:12,090
where we saw very low seedling abundance,

697
00:31:12,090 --> 00:31:13,430
the seedlings that are present

698
00:31:13,430 --> 00:31:16,340
are not the same as the
species in the canopy.

699
00:31:16,340 --> 00:31:20,220
And based on these patterns,
we use model selection again

700
00:31:20,220 --> 00:31:22,380
to determine which species and stressors

701
00:31:22,380 --> 00:31:24,060
were most likely driving these patterns,

702
00:31:24,060 --> 00:31:25,460
which we'll talk about next.

703
00:31:27,200 --> 00:31:28,940
To determine which species or groups

704
00:31:28,940 --> 00:31:30,670
are driving the low similarity,

705
00:31:30,670 --> 00:31:33,780
we mapped relative abundance
based on stem densities.

706
00:31:33,780 --> 00:31:36,530
And so looking at oak
first, which is on the top,

707
00:31:36,530 --> 00:31:38,390
we see that oak dominates the canopy

708
00:31:38,390 --> 00:31:41,220
across much of the region,
that's the map on the right,

709
00:31:41,220 --> 00:31:43,450
but is not as abundant
throughout that area

710
00:31:43,450 --> 00:31:45,970
in the seedling layer,
which is on the left.

711
00:31:45,970 --> 00:31:48,970
In contrast, beech is not
very abundant in the canopy,

712
00:31:48,970 --> 00:31:52,133
so that's the bottom right
map, but it is pretty abundant

713
00:31:52,133 --> 00:31:55,203
throughout the region
in the seedling layer.

714
00:31:57,760 --> 00:32:00,240
Invasive tree species
and low canopy species

715
00:32:00,240 --> 00:32:02,790
were also important
predictors in our model.

716
00:32:02,790 --> 00:32:05,970
And I think most importantly,
looking at low canopy species,

717
00:32:05,970 --> 00:32:08,956
on the left, there's very
high relative abundance

718
00:32:08,956 --> 00:32:13,620
of low canopy seedlings present
throughout the study area,

719
00:32:13,620 --> 00:32:16,890
but pretty low abundance
in the tree layer.

720
00:32:16,890 --> 00:32:19,910
And so again, low canopy
species are species

721
00:32:19,910 --> 00:32:22,393
like musclewood, ironwood,
sassafras, pawpaw,

722
00:32:23,230 --> 00:32:25,770
that have a tree-like growth form,

723
00:32:25,770 --> 00:32:26,770
but they're not gonna be able

724
00:32:26,770 --> 00:32:29,990
to form a tall closed canopy forest.

725
00:32:29,990 --> 00:32:32,140
And some research has also shown

726
00:32:32,140 --> 00:32:35,023
that many of those species
are browse resilient.

727
00:32:37,430 --> 00:32:40,010
So we finally, we combined the results

728
00:32:40,010 --> 00:32:42,060
from the seedling and sapling densities

729
00:32:42,060 --> 00:32:45,258
and the similarity between
seedling and canopy

730
00:32:45,258 --> 00:32:47,460
and sapling and canopy.

731
00:32:47,460 --> 00:32:51,200
And so here, this was
to look at the severity

732
00:32:51,200 --> 00:32:52,890
of the regeneration debt.

733
00:32:52,890 --> 00:32:55,930
And we can find in
purple is the most severe

734
00:32:55,930 --> 00:32:57,690
and then dark blue is high.

735
00:32:57,690 --> 00:33:00,550
And that means that
there was very low score

736
00:33:00,550 --> 00:33:02,633
in three of the four metrics.

737
00:33:05,640 --> 00:33:08,040
And overlaying our parks in the area

738
00:33:08,040 --> 00:33:10,710
where we find the most
severe regeneration debt

739
00:33:10,710 --> 00:33:13,950
with the FIA data that
also matches the parks

740
00:33:13,950 --> 00:33:16,060
where we have the highest deer impacts

741
00:33:16,060 --> 00:33:18,073
and the most regeneration problems.

742
00:33:21,220 --> 00:33:23,010
Okay, so just to summarize the results

743
00:33:23,010 --> 00:33:24,900
of the regeneration debt study.

744
00:33:24,900 --> 00:33:28,140
We found severe regeneration
debt in mid-Atlantic forests,

745
00:33:28,140 --> 00:33:31,090
particularly in oak, hickory forests.

746
00:33:31,090 --> 00:33:33,540
Anthropogenic stressors predicted the debt

747
00:33:33,540 --> 00:33:36,070
better than climate change metrics.

748
00:33:36,070 --> 00:33:38,310
And the most affected species were

749
00:33:38,310 --> 00:33:40,493
oak, hickory, and southern pine species.

750
00:33:43,060 --> 00:33:45,600
Okay, so now I'm going to shift

751
00:33:45,600 --> 00:33:48,720
to talking about future
concerns in eastern parks,

752
00:33:48,720 --> 00:33:51,623
and this is primarily
focused on climate change.

753
00:33:53,820 --> 00:33:55,920
I'm sure many of you are familiar

754
00:33:55,920 --> 00:33:58,290
with the Climate Change
Tree Atlas predictions,

755
00:33:58,290 --> 00:34:00,440
which predict changes in suitable habitat

756
00:34:00,440 --> 00:34:01,860
by the turn of the century,

757
00:34:01,860 --> 00:34:04,240
using species distribution models.

758
00:34:04,240 --> 00:34:06,050
And species distribution models.

759
00:34:06,050 --> 00:34:07,600
they essentially take environmental

760
00:34:07,600 --> 00:34:09,370
and climate variables that best predict

761
00:34:09,370 --> 00:34:11,420
the current distribution of a species,

762
00:34:11,420 --> 00:34:13,500
and then predict the
future suitable habitat

763
00:34:13,500 --> 00:34:16,320
for that species using
general circulation models

764
00:34:16,320 --> 00:34:18,237
of future climate scenarios.

765
00:34:18,237 --> 00:34:20,500
They're really helpful for predicting

766
00:34:20,500 --> 00:34:23,050
which species are likely
to be most vulnerable

767
00:34:23,050 --> 00:34:25,530
or robust to climate change.

768
00:34:25,530 --> 00:34:27,840
And as you can see in the maps here,

769
00:34:27,840 --> 00:34:29,540
the Climate Change Tree Atlas predicts

770
00:34:29,540 --> 00:34:31,890
a pretty major expansion
of suitable habitat

771
00:34:31,890 --> 00:34:34,400
for oak, hickory into
the Northeastern U.S.

772
00:34:34,400 --> 00:34:36,070
by the turn of the century.

773
00:34:36,070 --> 00:34:38,100
And even though everyone involved

774
00:34:38,100 --> 00:34:39,730
in the Climate Change Tree Atlas

775
00:34:39,730 --> 00:34:42,720
is super clear about how
to interpret these maps,

776
00:34:42,720 --> 00:34:45,570
which are merely predictions
of future suitable habitat,

777
00:34:45,570 --> 00:34:47,940
but not future forest composition,

778
00:34:47,940 --> 00:34:49,860
I've heard on many occasions,

779
00:34:49,860 --> 00:34:51,690
a number of researchers or managers

780
00:34:51,690 --> 00:34:54,030
misinterpret these maps as predictions

781
00:34:54,030 --> 00:34:57,710
of what the forest will be
in the next 50 to 100 years.

782
00:34:57,710 --> 00:35:00,720
And interpreting the predictions this way,

783
00:35:00,720 --> 00:35:04,080
even though it's wrong, isn't
that alarming to managers.

784
00:35:04,080 --> 00:35:05,880
They're generally okay with the idea

785
00:35:05,880 --> 00:35:08,730
of northern hardwoods
converting to oak, hickory.

786
00:35:08,730 --> 00:35:11,126
And the research I'm gonna talk about next

787
00:35:11,126 --> 00:35:14,010
was really largely in response to this.

788
00:35:14,010 --> 00:35:16,507
I wanted to see how likely tree species

789
00:35:16,507 --> 00:35:19,883
will be able to migrate into
their newly suitable habitat.

790
00:35:21,720 --> 00:35:24,340
Thinking about tree migration is important

791
00:35:24,340 --> 00:35:27,060
because temperate tree
species largely responded

792
00:35:27,060 --> 00:35:29,530
to past climate change through migration,

793
00:35:29,530 --> 00:35:32,220
rather than adapting to changes in place.

794
00:35:32,220 --> 00:35:35,970
And tree migration depends
on seed dispersal ability,

795
00:35:35,970 --> 00:35:38,560
life history traits, like age to maturity,

796
00:35:38,560 --> 00:35:41,710
habitat connectivity and
landscape configuration,

797
00:35:41,710 --> 00:35:45,050
between where a species is
now and where it needs to be,

798
00:35:45,050 --> 00:35:47,210
and then also regeneration.

799
00:35:47,210 --> 00:35:49,660
So we already talked
about regeneration debt

800
00:35:49,660 --> 00:35:51,895
and regional patterns of regeneration.

801
00:35:51,895 --> 00:35:55,760
Now I'm gonna talk about a
dispersal simulation we did

802
00:35:55,760 --> 00:35:57,653
to look into these three factors.

803
00:35:59,950 --> 00:36:03,100
For this study, we simulated
100 years of dispersal

804
00:36:03,100 --> 00:36:06,350
for 15 common eastern tree
species that are predicted

805
00:36:06,350 --> 00:36:08,970
to gain suitable habitat in the northeast.

806
00:36:08,970 --> 00:36:12,150
And our objectives for
the study were to assess

807
00:36:12,150 --> 00:36:15,090
species-specific dispersal ability.

808
00:36:15,090 --> 00:36:17,410
Determine influence of non-forest land use

809
00:36:17,410 --> 00:36:19,150
on dispersal ability.

810
00:36:19,150 --> 00:36:22,070
Identify if there are any
major dispersal barriers.

811
00:36:22,070 --> 00:36:24,420
And then also we compared dispersal rates

812
00:36:24,420 --> 00:36:25,910
with rates of habitat shift

813
00:36:25,910 --> 00:36:28,143
based on the Climate Change Tree Atlas.

814
00:36:30,240 --> 00:36:32,620
Our simulations were
informed by what we know

815
00:36:32,620 --> 00:36:35,290
about species-specific dispersal rates,

816
00:36:35,290 --> 00:36:38,570
including published dispersal
kernels for each species,

817
00:36:38,570 --> 00:36:41,180
life history traits, like age to maturity,

818
00:36:41,180 --> 00:36:43,430
and we used a 90 meter grid size.

819
00:36:43,430 --> 00:36:46,160
For our current range, we used
the Climate Change Tree Atlas

820
00:36:46,160 --> 00:36:48,020
current range for each species,

821
00:36:48,020 --> 00:36:49,850
so that's in green on the map.

822
00:36:49,850 --> 00:36:51,860
And for the range to migrate into,

823
00:36:51,860 --> 00:36:54,110
we use future suitable habitat predicted

824
00:36:54,110 --> 00:36:56,790
by the Climate Change Tree
Atlas, using the average

825
00:36:56,790 --> 00:37:00,740
of 3 GCMs and high emissions
scenario, and that's in yellow.

826
00:37:00,740 --> 00:37:04,010
And the map on the left
simulates dispersal

827
00:37:04,010 --> 00:37:07,760
with no dispersal barriers,
so that's our null dispersal.

828
00:37:07,760 --> 00:37:09,880
The map on the right simulates dispersal

829
00:37:09,880 --> 00:37:11,630
with a dispersal barrier,

830
00:37:11,630 --> 00:37:14,910
represented as non-forest
land use, that's in gray.

831
00:37:14,910 --> 00:37:16,850
And this is the start of the simulation

832
00:37:16,850 --> 00:37:17,850
for southern red oak.

833
00:37:17,850 --> 00:37:20,900
And next I'm gonna show you
how the simulations play out

834
00:37:20,900 --> 00:37:23,143
over 25 year time intervals.

835
00:37:25,700 --> 00:37:28,680
So 25 years after our simulation starts,

836
00:37:28,680 --> 00:37:32,350
we see that there's a pretty
continuous band of dispersal

837
00:37:32,350 --> 00:37:34,250
on the left, in the null model.

838
00:37:34,250 --> 00:37:35,650
Not so much on the right,

839
00:37:35,650 --> 00:37:39,773
where non-forest land use
is effectively a barrier.

840
00:37:41,040 --> 00:37:42,713
Here's 50 years later.

841
00:37:44,340 --> 00:37:46,023
75 years later.

842
00:37:48,520 --> 00:37:51,310
So at the end of the 100 year simulation,

843
00:37:51,310 --> 00:37:54,330
in the null dispersal results on the left,

844
00:37:54,330 --> 00:37:56,450
we found southern red oak only got

845
00:37:56,450 --> 00:37:59,710
about 50 kilometers
from its current range.

846
00:37:59,710 --> 00:38:03,050
Considering that most oak
species take 20 to 25 years

847
00:38:03,050 --> 00:38:05,740
to reach maturity under ideal conditions,

848
00:38:05,740 --> 00:38:08,180
this shouldn't really be a surprise.

849
00:38:08,180 --> 00:38:10,870
There's only a maximum of
four new dispersal events

850
00:38:10,870 --> 00:38:13,120
for newly colonized cells.

851
00:38:13,120 --> 00:38:16,050
I should also note that our
null dispersal simulations

852
00:38:16,050 --> 00:38:18,740
average roughly one to
10 kilometers per decade

853
00:38:18,740 --> 00:38:21,640
for their dispersal rate, which is similar

854
00:38:21,640 --> 00:38:23,763
to the rates of post-glacial migration

855
00:38:23,763 --> 00:38:26,160
determined by the pollen record.

856
00:38:26,160 --> 00:38:28,440
So we feel pretty confident
that our simulations

857
00:38:28,440 --> 00:38:30,820
are a decent representation of dispersal.

858
00:38:30,820 --> 00:38:34,210
That's just not very far on
the scale of our lifetimes.

859
00:38:34,210 --> 00:38:36,590
And now looking at the
simulation on the right

860
00:38:36,590 --> 00:38:38,640
that incorporated non-forest land use,

861
00:38:38,640 --> 00:38:42,730
we see that dispersal basically failed.

862
00:38:42,730 --> 00:38:45,630
There was no effective
dispersal between this area

863
00:38:45,630 --> 00:38:49,610
from Northern Virginia up
through central New Jersey.

864
00:38:49,610 --> 00:38:52,093
And this area from Northern Virginia

865
00:38:52,093 --> 00:38:54,000
up through central New Jersey

866
00:38:54,000 --> 00:38:56,010
was a consistent dispersal barrier

867
00:38:56,010 --> 00:39:00,173
for species whose ranges were
south of or near this region.

868
00:39:02,470 --> 00:39:06,050
So just to kind of summarize
what we found in this study,

869
00:39:06,050 --> 00:39:09,400
species dispersal rates
were considerably slower

870
00:39:09,400 --> 00:39:11,150
than their species habitat shifts,

871
00:39:11,150 --> 00:39:13,820
as predicted by the
Climate Change Tree Atlas,

872
00:39:13,820 --> 00:39:15,510
even in our null models,

873
00:39:15,510 --> 00:39:17,790
with 50 kilometers long
distance dispersals.

874
00:39:17,790 --> 00:39:20,750
So kind of maximum possible dispersal,

875
00:39:20,750 --> 00:39:23,490
it was still slower than habitat shifts.

876
00:39:23,490 --> 00:39:26,270
Oak species had the
slowest dispersal rates

877
00:39:26,270 --> 00:39:29,550
and also the fastest
species habitat shifts.

878
00:39:29,550 --> 00:39:31,520
No species filled the majority

879
00:39:31,520 --> 00:39:35,190
of its predicted newly
suitable habitat by 2100.

880
00:39:35,190 --> 00:39:38,450
And we also identified these
two major dispersal barriers

881
00:39:38,450 --> 00:39:42,320
from Washington DC area up
through central New Jersey.

882
00:39:42,320 --> 00:39:45,343
And then there was an east-west
barrier in central New York.

883
00:39:47,200 --> 00:39:49,900
So putting this all together,

884
00:39:49,900 --> 00:39:53,150
looking at the results from
the dispersal simulation

885
00:39:53,150 --> 00:39:57,160
and also our regeneration
debt study, we find that

886
00:39:57,160 --> 00:40:00,970
where the most extreme
dispersal barrier occurred

887
00:40:00,970 --> 00:40:03,740
overlaps almost exactly
with where we found

888
00:40:03,740 --> 00:40:06,083
the most severe regeneration debt.

889
00:40:07,830 --> 00:40:10,550
And then, of course,
overlaying where our parks are

890
00:40:10,550 --> 00:40:13,810
in this map, we see that
a number of our parks

891
00:40:13,810 --> 00:40:18,380
are in the area that is both
a severe regeneration debt

892
00:40:18,380 --> 00:40:22,240
and a migration barrier,
but also many of our parks

893
00:40:22,240 --> 00:40:24,960
are north of this, and so
they're likely to be affected

894
00:40:24,960 --> 00:40:27,643
because of this migration
barrier in the way.

895
00:40:30,610 --> 00:40:34,190
Tying this all together,
we found invasive species

896
00:40:34,190 --> 00:40:37,970
and overabundant deer are
impacting forest regeneration

897
00:40:37,970 --> 00:40:40,800
in parks and throughout the region.

898
00:40:40,800 --> 00:40:43,310
Many species predicted
to gain suitable habitat

899
00:40:43,310 --> 00:40:45,190
in the northeast have regeneration debts

900
00:40:45,190 --> 00:40:46,950
in their current range.

901
00:40:46,950 --> 00:40:48,780
And there's strong spatial overlap

902
00:40:48,780 --> 00:40:51,000
of regeneration debt
and dispersal barriers

903
00:40:51,000 --> 00:40:53,660
that will likely impact tree migration.

904
00:40:53,660 --> 00:40:55,280
And it's possible that these issues

905
00:40:55,280 --> 00:40:57,810
are already impacting tree migration,

906
00:40:57,810 --> 00:41:00,070
as there's some evidence
of northern hardwood

907
00:41:00,070 --> 00:41:02,690
and spruce fir species moving north.

908
00:41:02,690 --> 00:41:06,040
Some evidence of oak
hickory species moving west

909
00:41:06,040 --> 00:41:08,810
because of increased
moisture availability.

910
00:41:08,810 --> 00:41:10,860
However, there's almost no evidence

911
00:41:10,860 --> 00:41:13,280
of oak, hickory species moving north.

912
00:41:13,280 --> 00:41:15,760
And it may just be lack
of climate response,

913
00:41:15,760 --> 00:41:17,540
or it may be that they can't migrate

914
00:41:17,540 --> 00:41:20,630
through these dispersal
and regeneration barriers.

915
00:41:20,630 --> 00:41:22,990
And even though most of
the issues I documented

916
00:41:22,990 --> 00:41:26,030
are south of the region that
many at this meeting work in,

917
00:41:26,030 --> 00:41:28,810
this affects us too, because
the oak, hickory species

918
00:41:28,810 --> 00:41:31,570
that are predicted to gain
suitable habitat in the northeast

919
00:41:31,570 --> 00:41:34,293
are probably not gonna
make it on their own.

920
00:41:36,270 --> 00:41:39,090
Okay, so what should we
do about these issues?

921
00:41:39,090 --> 00:41:40,750
Management recommendations really are

922
00:41:40,750 --> 00:41:42,700
that we need to start doing something

923
00:41:42,700 --> 00:41:44,640
about these non-climate stressors

924
00:41:44,640 --> 00:41:46,810
that are impacting tree regeneration,

925
00:41:46,810 --> 00:41:49,620
particularly deer and invasive plants.

926
00:41:49,620 --> 00:41:52,570
And I really think we need to
do this at a regional level.

927
00:41:52,570 --> 00:41:54,790
This obviously poses
an enormous challenge,

928
00:41:54,790 --> 00:41:56,840
considering how many
states, jurisdictions,

929
00:41:56,840 --> 00:41:58,640
and landowners are involved.

930
00:41:58,640 --> 00:42:00,870
But if we don't start doing something big,

931
00:42:00,870 --> 00:42:02,500
more and more of the eastern forest

932
00:42:02,500 --> 00:42:05,850
is gonna look like the barberry
thickets in Morristown.

933
00:42:05,850 --> 00:42:08,390
We probably also need to
consider range expansion

934
00:42:08,390 --> 00:42:11,380
and assisted migration
beyond the dispersal

935
00:42:11,380 --> 00:42:13,680
and regeneration barriers
that I talked about,

936
00:42:13,680 --> 00:42:15,730
so that tree species are able to migrate

937
00:42:15,730 --> 00:42:17,253
with their suitable habitat.

938
00:42:19,300 --> 00:42:22,160
Okay, with the last
minute or so I have left

939
00:42:22,160 --> 00:42:23,900
I'm gonna completely shift gears

940
00:42:23,900 --> 00:42:25,600
by ending on some advice for folks

941
00:42:25,600 --> 00:42:27,800
who are starting up long-term
monitoring projects.

942
00:42:27,800 --> 00:42:29,760
And this is advice I would give myself

943
00:42:29,760 --> 00:42:30,870
if I could go back in time

944
00:42:30,870 --> 00:42:32,880
to when we were developing our protocols.

945
00:42:32,880 --> 00:42:35,130
And so the biggest piece
of advice I would give

946
00:42:35,130 --> 00:42:38,580
is to be consistent and borrow
from existing protocols.

947
00:42:38,580 --> 00:42:40,020
So once you start using a method,

948
00:42:40,020 --> 00:42:42,100
don't change it unless you
have really good reason to,

949
00:42:42,100 --> 00:42:44,090
so you don't lose data continuity.

950
00:42:44,090 --> 00:42:46,010
And if you're monitoring forests,

951
00:42:46,010 --> 00:42:49,190
do whatever the Forest
Service FIA program is doing,

952
00:42:49,190 --> 00:42:52,010
unless you have really good reason not to.

953
00:42:52,010 --> 00:42:54,420
Also avoid complex sample designs.

954
00:42:54,420 --> 00:42:57,000
So there were several networks,
particularly some out west,

955
00:42:57,000 --> 00:42:59,010
that use really complex panel designs.

956
00:42:59,010 --> 00:43:00,920
And they had completely legitimate reasons

957
00:43:00,920 --> 00:43:02,530
for doing so at the time.

958
00:43:02,530 --> 00:43:04,530
But losing a year of monitoring to COVID

959
00:43:04,530 --> 00:43:06,950
has created way more
headaches for their analyses

960
00:43:06,950 --> 00:43:10,283
than it has for those of us
with simpler panel designs.

961
00:43:11,620 --> 00:43:15,482
Quality control and quality
assurance is very important.

962
00:43:15,482 --> 00:43:18,980
So we re-sample 5% of our plots every year

963
00:43:18,980 --> 00:43:20,320
with a different crew.

964
00:43:20,320 --> 00:43:22,230
And this has helped us catch issues

965
00:43:22,230 --> 00:43:23,780
from the field crew early on,

966
00:43:23,780 --> 00:43:25,310
but it's also helped us improve

967
00:43:25,310 --> 00:43:27,830
or drop methods that
weren't very repeatable.

968
00:43:27,830 --> 00:43:29,630
It also gives you confidence in your data

969
00:43:29,630 --> 00:43:32,343
if you have very good
precision across crews.

970
00:43:34,670 --> 00:43:36,340
It's also helpful to try collecting

971
00:43:36,340 --> 00:43:38,610
your data digitally wherever possible.

972
00:43:38,610 --> 00:43:40,640
So for most of our monitoring protocols,

973
00:43:40,640 --> 00:43:42,100
we record our data directly

974
00:43:42,100 --> 00:43:44,440
into a database or app in the field.

975
00:43:44,440 --> 00:43:47,210
And data quality has greatly
improved as a result,

976
00:43:47,210 --> 00:43:48,990
because we have built in checks

977
00:43:48,990 --> 00:43:50,960
that check against previous surveys.

978
00:43:50,960 --> 00:43:52,670
They check for missing data.

979
00:43:52,670 --> 00:43:55,060
And they also require standardized values.

980
00:43:55,060 --> 00:43:58,020
And as a result, within a day or two

981
00:43:58,020 --> 00:43:59,520
of our crews leaving a park,

982
00:43:59,520 --> 00:44:01,370
we're able to actually certify the data

983
00:44:01,370 --> 00:44:03,700
and start analyzing
it, instead of spending

984
00:44:03,700 --> 00:44:06,610
the next month entering and checking data.

985
00:44:06,610 --> 00:44:09,960
And finally look at your
data as it's coming in.

986
00:44:09,960 --> 00:44:13,430
So resolving errors is often
possible if you find it

987
00:44:13,430 --> 00:44:15,690
within a day or two of
the data being collected.

988
00:44:15,690 --> 00:44:17,180
This is also a lot easier to do

989
00:44:17,180 --> 00:44:19,570
if you're collecting your data digitally.

990
00:44:19,570 --> 00:44:22,050
Months or years later,
there's really no way to know

991
00:44:22,050 --> 00:44:25,320
whether a weird data point you
found was real or an error.

992
00:44:25,320 --> 00:44:27,883
So checking your data
frequently is really helpful.

993
00:44:30,630 --> 00:44:32,242
Okay, so that's all I have.

994
00:44:32,242 --> 00:44:35,730
If there's time, I'm happy
to answer any questions.

995
00:44:35,730 --> 00:44:37,950
And for those of you looking
for summer field work,

996
00:44:37,950 --> 00:44:40,740
we're always looking for
crew members for the summer.

997
00:44:40,740 --> 00:44:43,760
And our job announcements
usually go out in mid-February.

998
00:44:43,760 --> 00:44:46,860
And if you are interested,
feel free to contact me.

999
00:44:46,860 --> 00:44:49,040
My email is included on this slide.

1000
00:44:49,040 --> 00:44:49,873
Thank you.