PreprintPDF Available

Global worming: massive invasion of North America by earthworms revealed

Authors:
  • Oligochaetology Laboratory, Kitchener, Ontario, Canada
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Human activities cause major ecological changes by reshuffling the spatial distribution of species. The extent to which this process affects belowground biota is a critical issue because soil organisms play a key role in ecosystem functioning and maintenance. However, the magnitude of the reshuffling of soil species remains unknown so far because of the lack of a historic baseline. Here, we begin to fill this gap with the largest spatiotemporal database of native and alien earthworms in North America. Our results reveal that the entire continent is being invaded by non-native earthworms through a variety of pathways. We show that these aliens bring novel ecological functions in most regions and thus represent a major threat to native ecosystems. Our findings demonstrate that earthworms, and most likely other soil organisms, represent a major but overlooked pool of invasive species with strong ecological impact. They need to be better integrated in control and mitigation strategies.
Content may be subject to copyright.
1
Title: Global worming: massive invasion of North America by earthworms
1
revealed
2
3
Authors: Jérôme Mathieu1*, John Warren Reynolds2,3, Carlos Fragoso4, Elizabeth Hadly5,6,7
4
Affiliations:
5
1Sorbonne Université, CNRS, UPEC, INRAE, IRD, Institut d’Ecologie et des Sciences de
6
l’Environnement de Paris; 4 place Jussieu, F-75005, Paris, France.
7
2Oligochaetology Laboratory; 1250 Weber Street East, Kitchener, ON Canada N2A 4E1.
8
3New Brunswick Museum; Saint John, NB Canada E2K 1E5.
9
4Red de Biodiversidad y Sistemática, Instituto de Ecología A.C.; Xalapa, Ver. México
10
5Department of Biology, Stanford University; Stanford, CA, United States.
11
6Jasper Ridge Biological Preserve, Stanford University; Stanford, CA, United States.
12
7Stanford Woods Institute for the Environment, Stanford University; Stanford, CA, United
13
States.
14
15
*Corresponding author. Email: jerome.mathieu@sorbonne-universite.fr
16
17
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
2
Abstract:
18
Human activities cause major ecological changes by reshuffling the spatial distribution of
19
species. The extent to which this process affects belowground biota is a critical issue
20
because soil organisms play a key role in ecosystem functioning and maintenance.
21
However, the magnitude of the reshuffling of soil species remains unknown so far because
22
of the lack of a historic baseline. Here, we begin to fill this gap with the largest
23
spatiotemporal database of native and alien earthworms in North America. Our results
24
reveal that the entire continent is being invaded by non-native earthworms through a
25
variety of pathways. We show that these aliens bring novel ecological functions in most
26
regions and thus represent a major threat to native ecosystems. Our findings demonstrate
27
that earthworms, and most likely other soil organisms, represent a major but overlooked
28
pool of invasive species with strong ecological impact. They need to be better integrated in
29
control and mitigation strategies.
30
31
Main Text:
32
The transport of species beyond their biogeographic barriers plays a major role in global
33
homogenization of biodiversity1, biodiversity loss, and has countless consequences on
34
ecosystems (1). Invasive insects alone cost a minimum of US$70.0 billion per year globally2.
35
Biological invasions are a major driver of global change, synergized by and influencing
36
biodiversity loss, climate change and land use change3. Identifying aliens that can turn invasive
37
and anticipating their spread is thus critical to build environmental management strategies and
38
policies.
39
40
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
3
The spatial distribution and dynamics of alien species is now well documented at the global scale
41
in a large number above-ground taxa4. This knowledge provides a background to develop
42
management strategies of aliens at national and international scales. In contrast, the spatial
43
distribution of alien species belowground is still largely unknown5. A limited number of studies
44
documented the expansion of a few specific soil organisms, and to date, none have quantified the
45
colonization dynamics of multiple soil species on a continental scale. Ignoring these organisms
46
will result in missing critical mechanisms of ecological changes for soil organisms have a strong
47
impact on overall ecosystem function. Many soil organisms are qualified as ecosystem
48
engineers6 precisely because they play a central role in many processes that cascade to
49
aboveground communities and to the atmosphere7. Consequently, any differences in life
50
histories, network linkages (i.e., trophic or symbiotic), and metabolisms of alien soil species will
51
have a strong impact on native ecosystems.
52
53
Among alien soil taxa, earthworms are of particular concern because of their dramatic impact on
54
native ecosystems. Earthworms, in general, are viewed favourably because they are perceived to
55
increase crop productivity and maintain soil fertility8,9. They are iconic of good soil
56
management. As a consequence, colonization by alien earthworms is often viewed positively,
57
and sometimes even encouraged1013. However, not all earthworms are the same. In tropical
58
areas, alien earthworm species such as Pontoscolex corethrurus can result in compacted soil and
59
make it unsuitable for plant growth in less than a year14. In the northern broadleaf forests of USA
60
and Canada, particularly where few earthworms survived Pleistocene glaciation, colonization by
61
alien earthworms affects all components of the native ecosystems15. Native soils are transformed
62
by the mixing of the horizons and the rapid disappearance of the litter layer16. Alien earthworm
63
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
4
bioturbation results in a simplification of the understory vegetation and increases mortality of
64
important economic species such as sugar maple (Acer sacharrum)1719. Alien earthworms also
65
impact native soil food webs and the organisms that depend on them, such as springtails, ground
66
beetles and the red-backed salamander20,21. Ultimately, the colonization by alien earthworms can
67
facilitate the spread of invasive plants22.
68
69
Even though the strong effects of alien earthworms are well known, their spatial distribution and
70
their colonization dynamics over large scales has not been well characterized. Therefore, we
71
have limited baselines to assess where alien earthworms may represent a threat, or where they
72
have already impacted ecosystems. This is surprising because it is known since the beginning of
73
the 20th century that several species of earthworm have colonized different parts of the world by
74
inadvertent human transport23. Indeed, it was predicted in 1900 that alien earthworms should
75
have conquered the whole of North America by the end of the XXth century24. But, to date, no
76
data were available to test this prediction. North America is particularly sensitive to biological
77
invasions25,26. Many North American ecosystems evolved in the absence of earthworms at least
78
since the Last Glacial Maximum, until their recent introduction by humans 27. In these
79
ecosystems, earthworms arrived into an empty belowground niche with limited competitors. In
80
addition, climate change is facilitating their spread from South America to Mexico and into the
81
northern parts of the continent, where the permafrost is melting, potentially introducing a
82
positive feedback loop with climate change2830.
83
84
Here, we reconstruct the spatiotemporal spread and putative introduction pathways of exotic
85
earthworm species at the scale of a continent over a century. To achieve this, we built the
86
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
5
database EWINA, which is the most comprehensive database of native and exotic earthworm
87
species occurrence ever built. This database collates more than 68k records dating from 1850 to
88
2021, across 2510 geographical units in North America (Extended Data Fig. 1). We completed
89
this database using EWINA_IPATHS, a second database which focuses on the introduction
90
pathways of alien earthworm species into the USA between 1945 and 1975. Combined, these
91
two databases provide a baseline with which to ascertain earthworm invasion dynamics in North
92
America, which can be used to investigate alternative scenarios and policies for management of
93
alien earthworms and other soil taxa here and globally.
94
95
Results
96
A massive invasion of North American soils by earthworms
97
To examine the extent of the geographic distribution of alien earthworm species, we mapped the
98
predicted Relative Alien Species Richness (RASR) of earthworm across North America (Fig.
99
1A), based on present environmental condidtions. RASR is the ratio of the number of alien
100
species over the total number of species in a given geographical unit (typically a county herein,
101
see Extended Data Fig. 1). We modeled RASR as a function of 12 environmental drivers
102
(Extended Data Table 1) using data between 2000 and 2021, with a machine learning approach.
103
Predicted RSAR of earthworm in actual conditions is overall extremely high with a median of
104
73% across North America (Fig. 1B), meaning that alien species dominate in a major proportion
105
of the continent. Only 3% of studied geographical units are devoid of alien earthworms, while
106
28% are devoid of any native ones. The map reveals an opposition between the northern part of
107
the continent, which has a RASR typically higher than 50%, and the south and west of the
108
continent, where RASR is typically below 50%.
109
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
6
110
Overall, exotics represent 23% of the 308 earthworm species recorded in North America (Table
111
1). RASR at the country level is twice as high in Canada than in the US and in Mexico. There are
112
many more alien species than native ones in Canada. Over North America, exotic earthworms
113
have on average a larger geographical range than the natives (Extended Data Fig. 2), typical of
114
other invasive species. A startling twelve of the thirteen topmost widespread earthworm species
115
are alien. Aporrectodea trapezoides and Lumbricus rubellus are the most ubiquitous species (60
116
and 43% of geographical units respectively). Alien species are also proportionally more
117
parthenogenetic than native ones (Extended Data Fig. 3).
118
119
Climate is the best predictor of aliens' distribution
120
Surprisingly, climate, but not human activity, was the strongest predictor of RASR (Extended
121
Data Fig. 4). Mean Annual Temperature (bio1) clearly had the highest effect on RASR, followed
122
by carbon at soil surface and annual precipitation (bio12). The other environmental features,
123
including human activity, had a very limited impact on RASR. Mean annual temperature and
124
annual precipitation were negatively related to RASR (Extended Data Fig. 5). Soil carbon at soil
125
surface was positively related to RASR. The effect of the other covariates was very limited,
126
except the human impact index, which was positively related to RASR in a few environmental
127
conditions (Extended Data Fig. 5).
128
129
Earthworm alien species richness is correlated to above ground taxa
130
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
7
To test the specificity of the distribution of alien earthworms versus aboveground exotics, we
131
correlated the alien earthworm species richness to the one of non-native Plants, Spiders,
132
Mammals, and Birds across the geographical units defined by the Biodiversity Information
133
Standards31 (TDWG). We used the data from Dawson et al.4 to compute aboveground alien
134
species richness (Extended Data Fig. 6). Earthworm alien species richness was positively
135
correlated to all taxa, particularly with plants and spiders (Fig. 2). This suggests that alien
136
earthworm species richness in North America is not independent to that of aboveground taxa and
137
indeed, they may be linked, at least at the resolution of the TDWG units.
138
139
A slow but certain colonization of North America from Europe and Asia
140
Anticipating the spread of alien species requires a general understanding of the temporal
141
dynamics of the colonization process. To tackle this issue, we evaluated three critical aspects of
142
the dynamics.
143
First, we looked at how alien species number have developed in time in North America since
144
1860 (Fig. 3A). The temporal accumulation of alien species is constant and steep after 1900, with
145
a peak of first species records in 1950. The average rate of colonization since 1950 is one new
146
alien species every three years.
147
148
Second, we assessed the strength of the propagule pressure - the flux of individuals and
149
earthworm species - arriving from other places of the world. To do so, we compiled reports of
150
events of earthworm interceptions at the US borders over 30 years. Although these data are not
151
standardized and only represent a fraction of the total introductions, they show that introductions
152
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
8
were happening consistently between 1945 and 1970, with a peak in 1949 (Fig. 3B). This means
153
that apart from 1950, the propagule pressure was fairly constant over this period of time.
154
155
Third, we analyzed the nature of the introduction pathways and their geographical features. Alien
156
earthworm species principally originate from Europe and Asia (Fig. 4A). Most species were
157
introduced by airplane (Fig. 4B). Islands such as Hawaii and Caribbean islands seem to have
158
played an important role as both sinks and sources of alien species. Alien earthworms were
159
mostly intercepted in coastal regions of the US, in airports and harbors, which is consistent with
160
the observed spatiotemporal colonization pattern (Fig. 5).
161
162
A colonization from the coast to the interior
163
To determine the colonization rate of aliens within the continent, we reconstructed the spatial
164
temporal colonization of alien earthworm species across North America since the beginning of
165
the 20th century (Fig. 5). Even though data are scare before 1960, it appears that colonization
166
started from the both Atlantic and Pacific coasts - where major ports were located - and gradually
167
reached the interior of the continent. The temporal dynamics of the individual alien species
168
spread reveals a continuous and pulsed increase in the geographical range of species (Extended
169
Data Fig. 7).
170
171
Alien earthworm species bring ecological novelty
172
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
9
To assess the potential impact of alien earthworm species on native ecosystems, we evaluated
173
their ecological redundancy with native earthworm species. High redundancy would suggest
174
little potential impact, while low redundancy would suggest ecological novelty and high
175
potential impact. We mapped the functional enrichment brought by the arrival of alien
176
earthworm species across the TDWG units to identify the zones under highest threat. Native and
177
alien earthworm species share little functional redundancy (Fig. 6A). Native species mostly feed
178
in/on soil, while aliens primarily feed in/on litter. This suggests that in the majority of locations,
179
the arrival of alien earthworm species at least increases litter decomposition. In locations where
180
no native earthworms were present, such as the northern Great Plains, aliens should also increase
181
soil bioturbation. Overall, alien earthworms bring higher rates of biomass turnover except in the
182
parts of the eastern USA and a few locations in Mexico and the west coast.
183
184
Discussion
185
A massive colonization of North America by alien earthworm species
186
The majority of the North American soils contain alien earthworms. Only 3% of the
187
geographical units that we investigated were free of exotic earthworm species. At the county
188
level, RASR was dramatically high, with a median above 70%, showing that in most of the
189
continent, alien earthworm species are in fact more diverse than the native species. Temporal
190
dynamics suggest that there may be a limit to the introduction of new alien earthworm species, at
191
least in the present environmental conditions. However, the spatiotemporal maps suggest that
192
alien species are still becoming invasive within the continent. Overall, our study demonstrates
193
that multiple species of earthworms, and by inference, other soil macro and microinvertebrates,
194
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
10
have been massively translocated between and within continents, and should be considered in
195
biological invasion policies where they are not already. For example the trade of alien earthworm
196
species is allowed without restrictions between most states of the US.
197
198
It is difficult to assess whether the trend we found is specific to North America because of the
199
lack of other similar databases on alien earthworms in other parts of the world. The few other
200
available data on earthworm alien richness suggest that they may also be substantial in other
201
regions. For example, the Northern Russian Plain is believed to be inhabited mainly by European
202
alien species32. A few countries such as China, Korea, and Myanmar also seem to host a high
203
level of alien species, between 20 and 30%33. At the other end of the gradient, Africa is believed
204
to have very few alien earthworm species34, though no quantitative data are available to confirm
205
these statements.
206
207
Surprisingly, earthworms' RASR increases from the tropics to the North Pole, whereas it
208
decreases towards the pole in other aboveground terrestrial invertebrates35. In contrast, alien
209
species richness of earthworms shows a parallel gradient with some aboveground invasives,
210
especially plants and spiders. Correlation with plants and spiders is likely due to a common
211
history of introduction, as earthworms and spiders are often imported accidentally as
212
contaminants of plants, either in the soil around the roots, in the leaves, or among fruits36.
213
Earthworms are easily at the top of all invasive animal taxa in North American in terms of alien
214
versus native species richness ratio, with 25% of North American earthworm species being
215
exotic. In the United States alone, only 6% of mammal species are exotic, 2% of insects and
216
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
11
arachnids, and 8% of fish species37. The lack of awareness of how much the soil biota has been
217
transformed is alarming. The high proportion of alien species in the earthworm group may be
218
explained by the limited pool of North American native species: 240 species, likely due to the
219
absence of native earthworms in much of the continent for reasons that are still debated27,38. The
220
presence of the Ice Sheet during Last Glacial Maximum (LGM) in North America has often been
221
put forward as an explanation of this peculiar pattern. Another nonexclusive - explanation is
222
that the diversity of native earthworm species is underestimated because not all native species
223
have been described, particularly in southern and central parts of the continent.
224
Accidental introductions are typical of invertebrates in general39. We show that this dynamic in
225
earthworms was sustained over 30 years and that the species originated mostly from Europe and
226
Asia. Interestingly, the peak of new aliens in 1950 was also reported in insects and mollusks 40. It
227
can be linked to the increased trade that followed the General Agreement on Tariffs and Trade
228
(TAGG) of 1947, which reduced trade barriers such as tariffs and quotas. Overall, our results
229
suggest a strong link with the trading routes and policies. It must be noticed that our data do not
230
cover legal importations, in particular between Canada and the US. Indeed, over 500 million
231
alien earthworm now are exported yearly from Canada to other countries, especially to the USA,
232
to be used as fishing or composting commodity41. This massive flux constitutes a major and legal
233
pathway of alien earthworm species transportation. It appears in contradiction with the efforts
234
developed by the U.S. Department of Agriculture to prevent introductions of alien earthworms in
235
the US.
236
Invasive or alien earthworms?
237
Due to their large distribution and their specific ecological role, alien earthworms represent a
238
serious threat to native ecosystems in North America. Whether these organisms will turn into
239
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
12
invasive depends on their impact on native ecosystems. The impact of alien earthworms in
240
broadleaf forest has been well documented, showing that these ecosystems are under particular
241
threat. However the impact of alien earthworms on other native ecosystems such as natural
242
grasslands, conifer forests, or chaparral, characteristic of unique species assemblages in North
243
America, or in our croplands, is poorly known (but see4245). It may be more critical to deliberate
244
on the status of alien earthworm in these systems. We show that when alien earthworms colonize
245
a new area, most of the time they fill an empty functional niche (Fig. 6), because natives are
246
absent or tend to be soil feeders, while aliens tend to feed on litter. Aliens that establish thus have
247
a strong potential impact on native ecosystems.
248
249
It is thus tempting to classify non-native earthworm species as invaders. However, many alien
250
earthworm species in North America provide important ecosystem services, particularly to
251
agriculture7. For instance, their presence dramatically increases crop productivity7, which
252
motived their deliberated introductions in many places of the world10,11. In some parts of Canada,
253
alien earthworm picking and trading represent an important economy41, demonstrating that
254
although alien earthworm species may be considered as a pest by one sector, they may be a
255
valuable ecosystem provider to another46,47. Earthworms highlight that the one major group of
256
organisms may be perceived as beneficial in some regions, typically agricultural ones, but
257
negative in others, such as more natural ones, a case similarly derived for the European
258
honeybee48. We know little about how either native or non-native earthworms have impacted
259
agricultural lands or less-impacted ecosystems.
260
261
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
13
Earthworm's spread in the future
262
Expansion of earthworms in North America will depend both on the flux of introductions into
263
the continent and on the capacity of aliens to spread within the continent upon arrival. The
264
temporal trend in the accumulated number of alien earthworm species shows a plateau since
265
2001. This plateau suggests that the number of earthworm species in North American soils may
266
have stabilized. But this doesn't mean that the corridors for new species or for in situ increase in
267
abundances have vanished. For example the flux of aliens between Canada and the US is still
268
massive due to the fishing bait market41, and the further demise of native earthworms may induce
269
new limits to introductions and invasions.
270
The spatial spread of alien species within the continent directly depends on the dispersal capacity
271
of earthworms. Dispersal of earthworms is mostly passive and achieved by the transport of
272
individuals and human activities49. In North America, known transports include uses for fishing,
273
vermicomposting, and agriculture. It has been predicted that in a business as usual scenario, 49%
274
of suitable habitats in northern Alberta will host alien earthworm species within 50 years, and
275
more than 92% in Ottawa within a century50,51. However, there are large uncertainties around
276
these estimations because the population dynamics of these species, and their competitive
277
interactions with natives, are poorly known. Our results suggest that aliens will increase in
278
numbers and distribution rapidly because many are parthenogenetic (Extended Data Fig. 3).
279
However, we need a better understanding of life histories of both native and alien species to be
280
able to anticipate their potential development at local-to-continental scales.
281
Our results also suggest that climate change will play an important role in the future spread of
282
alien earthworms, because temperature and precipitation are among the most important
283
predictors of earthworms' RASR. Climate has also been identified as the main driver of future
284
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
14
biological invasions52 in other invertebrate taxa such as ants53 and gastropods54. Though, whereas
285
invasion hotspots of ants and several other taxa are predicted to occur in warm areas53,55, alien
286
earthworms are more prevalent in colder, wetter areas. Hence, alien earthworm species may
287
expand northwards more rapidly to track cool, humid soils, which are more favorable to the
288
exotic species. Canada and northern parts of the US thus appear as potential future hotspots for
289
alien earthworm species in changing climates56. In southern regions, climate change will less
290
likely facilitate the spread of alien earthworms from Asia and Europe. However, warming of
291
temperate areas may open colonization opportunities for tropical alien earthworm species, as
292
already observed in a few temperate grasslands57.
293
Because it is virtually impossible to remove established populations of alien earthworms5862, the
294
best management options we have is to focus on prevention and early detection63. Prevention can
295
take several forms such as education of gardeners, fishermen and farmers. Such effort has
296
already begun in the regions of Great Lakes ("Great Lakes Worm Watch" initiative) and could be
297
extended via citizen science, through for example the #WorldWormWeek. Another approach is
298
to encourage scientific research and use of native species in fishing and vermicomposting of
299
agriculture. For instance, alien composting species such as Eisenia foetida can be replaced by
300
native ones such as Bimastos tumidus 64. Several species of the native genus Diplocardia are
301
currently collected commercially for bait from natural populations in Kansas, Missouri, and
302
Florida64. Overall, because earthworms are dispersal limited but easily transported by human
303
activities, the future spread of alien earthworm, and probably other soil organisms, strongly
304
depend on our capacity to develop policies to manage fluxes of alien soil organisms at national
305
and international levels62.
306
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
15
Main references
307
1. Ehrenfeld, J. G. Ecosystem Consequences of Biological Invasions. Annu. Rev. Ecol. Evol. Syst. 41, 59
308
80 (2010).
309
2. Bradshaw, C. J. A. et al. Massive yet grossly underestimated global costs of invasive insects. Nat.
310
Commun. 7, 12986 (2016).
311
3. Murphy, G. E. P. & Romanuk, T. N. A meta‐analysis of declines in local species richness from human
312
disturbances. Ecol. Evol. 4, 91103 (2014).
313
4. Dawson, W. et al. Global hotspots and correlates of alien species richness across taxonomic groups.
314
Nat. Ecol. Evol. 1, 0186 (2017).
315
5. Ehrenfeld, J. G. & Scott, N. Invasive species and the soil: Effects on organisms and ecosystem
316
processes. Ecol. Appl. 11, 12591260 (2001).
317
6. Cameron, E. K., Vilà, M. & Cabeza, M. Global meta-analysis of the impacts of terrestrial invertebrate
318
invaders on species, communities and ecosystems: Ecological impacts of terrestrial invertebrate
319
invaders. Glob. Ecol. Biogeogr. 25, 596606 (2016).
320
7. Blouin, M. et al. A review of earthworm impact on soil function and ecosystem services. Eur. J. Soil
321
Sci. 64, 161182 (2013).
322
8. Pauli, N., Abbott, L. K., Negrete-Yankelevich, S. & Andrés, P. Farmers’ knowledge and use of soil
323
fauna in agriculture: a worldwide review. Ecol. Soc. 21, (2016).
324
9. van Groenigen, J. W. et al. Earthworms increase plant production: a meta-analysis. Sci. Rep. 4, 6365
325
(2014).
326
10. Stockdill, S. M. J. Effects of introduced earthworms on the productivity of New Zealand pastures.
327
Pedobiologia 24, 2935 (1982).
328
11. Ghilarov, M. S. & Perel, T. S. Transfer of earthworms (Lumbricidae,Oligochaeta) for soil amelioration
329
in the USRR. Pedobiologia 27, 107113 (1984).
330
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
16
12. Seidl, D. E. & Klepeis, P. Human Dimensions of Earthworm Invasion in the Adirondack State Park.
331
Hum. Ecol. 39, 641655 (2011).
332
13. Blakemore, R. J. Agronomic potential of earthworms in brigalow soils of south-east Queensland. Soil
333
Biol. Biochem. 29, 603608 (1997).
334
14. Chauvel, A. et al. Pasture damage by an Amazonian earthworm. Nature 398, 3233 (1999).
335
15. Bohlen, P. J. et al. Ecosystem Consequences of Exotic Earthworm Invasion of North Temperate
336
Forests. Ecosystems 7, 112 (2004).
337
16. Bohlen, P. J., Pelletier, D. M., Groffman, P. M., Fahey, T. J. & Fisk, M. C. Influence of Earthworm
338
Invasion on Redistribution and Retention of Soil Carbon and Nitrogen in Northern Temperate
339
Forests. Ecosystems 7, 1327 (2004).
340
17. Bal, T. L., Storer, A. J. & Jurgensen, M. F. Evidence of damage from exotic invasive earthworm
341
activity was highly correlated to sugar maple dieback in the Upper Great Lakes region. Biol.
342
Invasions 20, 151164 (2018).
343
18. Frelich, L. E. et al. Earthworm invasion into previously earthworm-free temperate and boreal
344
forests. Biol. Invasions 8, 12351245 (2006).
345
19. Craven, D. et al. The unseen invaders: introduced earthworms as drivers of change in plant
346
communities in North American forests (a meta-analysis). Glob. Change Biol. 23, 10651074 (2017).
347
20. Jochum, M. et al. Earthworm invasion causes declines across soil fauna size classes and biodiversity
348
facets in northern North American forests. Oikos 130, 766780 (2021).
349
21. Ransom, T. Earthworms, as ecosystem engineers, influence multiple aspects of a salamander’s
350
ecology. Oecologia 165, 74554 (2011).
351
22. Nuzzo, V. A., Maerz, J. C. & Blossey, B. Earthworm Invasion as the Driving Force Behind Plant
352
Invasion and Community Change in Northeastern North American Forests. Conserv. Biol. 23, 966
353
974 (2009).
354
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
17
23. Hendrix, P. F. et al. Pandora’s Box Contained Bait: The Global Problem of Introduced Earthworms.
355
Annu. Rev. Ecol. Evol. Syst. 39, 593613 (2008).
356
24. Eisen, G. Researches in American Oligochaeta, with Especial Reference to those of the Pacific Coast
357
and Adjacent Islands. Proc. Calif. Acad. Sci. 2, 85268 (1900).
358
25. Early, R. et al. Global threats from invasive alien species in the twenty-first century and national
359
response capacities. Nat. Commun. 7, 12485 (2016).
360
26. Czech, B. & Krausman, P. R. Distribution and Causation of Species Endangerment in the United
361
States. Science 277, 11161117 (1997).
362
27. Reynolds, J. W. The distribution of the earthworms (Oligochaeta) of Indiana: a case for the post
363
quaternary introduction theory for megadrile migration in North America. Megadrilogica 5, 1332
364
(1994).
365
28. Mckenzie, J. et al. Invited Perspective: What Lies Beneath a Changing Arctic? (2020). doi:10.5194/tc-
366
2020-132.
367
29. Lubbers, I. M. et al. Greenhouse-gas emissions from soils increased by earthworms. Nat. Clim.
368
Change 3, 187194 (2013).
369
30. Conant, R. T., Paustian, K. & Elliott, E. T. Grassland Management and Conversion into Grassland:
370
Effects on Soil Carbon. Ecol. Appl. 11, 343355 (2001).
371
31. TDWG. Vocabulary Maintenance Specification Task Group. 2017. Standards Documentation
372
Standard. Biodiversity Information Standards (TDWG) http://www.tdwg.org/standards/147. (2017).
373
32. Tiunov, A. V., Hale, C. M., Holdsworth, A. R. & Vsevolodova-Perel, T. S. Invasion patterns of
374
Lumbricidae into previously earthworm-free areas of northeastern Europe and the western Great
375
Lakes region of North America. Biol. Invasions 8, 12231234 (2006).
376
33. Blakemore, R. Cosmopolitan EarthwormsA Global and Historical Perspective. in Annelids in
377
Modern Biology 257283 (2009). doi:10.1002/9780470455203.ch14.
378
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
18
34. Fragoso, C. et al. Earthworm Communities of Tropical Agroecosystems: Origin, Structure and
379
Influence of Management Practices. in Earthworm Management in Tropical Agroecosystems (eds.
380
Lavelle, P., Brussard, L. & Hendrix, P.) 2755 (CAB International, 1999).
381
35. Lilleskov, E. A., Mattson, W. J. & Storer, A. J. Divergent biogeography of native and introduced soil
382
macroinvertebrates in North America north of Mexico. Divers. Distrib. 14, 893904 (2008).
383
36. Nentwig, W. Introduction, establishment rate, pathways and impact of spiders alien to Europe. Biol.
384
Invasions 17, 27572778 (2015).
385
37. OTA. Harmful Non-Indigenous Species in the United States. (Office of Technology Assessment, 1993).
386
38. Hansen, P. L., Holmstrup, M., Bayley, M. & Simonsen, V. Low genetic variation for Dendrobaena
387
octaedra from Greenland compared to populations from Europe and North America: Refuge or
388
selection? Pedobiologia 50, 225234 (2006).
389
39. Hulme, P. E. et al. Grasping at the routes of biological invasions: a framework for integrating
390
pathways into policy. J. Appl. Ecol. 45, 403414 (2008).
391
40. Seebens, H. et al. No saturation in the accumulation of alien species worldwide. Nat. Commun. 8,
392
14435 (2017).
393
41. Steckley, J. Cash cropping worms: How the Lumbricus terrestris bait worm market operates in
394
Ontario, Canada. Geoderma 363, 114128 (2020).
395
42. Callaham, M. A., Blair, J. M. & Hendricks, W. A. Different behavioural patterns of the earthworms
396
Octolasion tyrtaeum and Diplocardia spp. in tallgrass prairie soils: potential influences on plant
397
growth. Biol. Fertil. Soils 34, 4956 (2001).
398
43. Regnier, E. et al. Impact of an exotic earthworm on seed dispersal of an indigenous US weed. J. Appl.
399
Ecol. 45, 16211629 (2008).
400
44. Fragoso, C. et al. Agricultural intensification, soil biodiversity and agroecosystem function in the
401
tropics: the role of earthworms. Appl. Soil Ecol. 6, 1735 (1997).
402
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
19
45. James, S. W. Soil, nitrogen, phosphorus, and organic matter processing by earthworms in tallgrass
403
prairie. Ecology 72, 21012109 (1991).
404
46. McNeely, J. A. The Great Reshuffling: Human Dimensions of Invasive Alien Species. (IUCN, 2001).
405
47. Estévez, R. A., Anderson, C. B., Pizarro, J. C. & Burgman, M. A. Clarifying values, risk perceptions, and
406
attitudes to resolve or avoid social conflicts in invasive species management: Confronting Invasive
407
Species Conflicts. Conserv. Biol. 29, 1930 (2015).
408
48. Moritz, R. F. A., Härtel, S. & Neumann, P. Global invasions of the western honeybee (Apis mellifera)
409
and the consequences for biodiversity. Écoscience 12, 289301 (2005).
410
49. Dupont, L., Torres-Leguizamon, M., Rene-Corail, P. & Mathieu, J. Landscape features impact
411
connectivity between soil populations: a comparative study of gene flow in earthworms. Mol. Ecol.
412
26, 31283140 (2017).
413
50. Cameron, E. K. & Bayne, E. M. Road age and its importance in earthworm invasion of northern
414
boreal forests. J. Appl. Ecol. 46, 2836 (2009).
415
51. Gundale, M. J., Jolly, W. M. & Deluca, T. H. Susceptibility of a Northern Hardwood Forest to Exotic
416
Earthworm Invasion. Conserv. Biol. 19, 10751083 (2005).
417
52. Bellard, C. et al. Will climate change promote future invasions? Glob. Change Biol. 19, 37403748
418
(2013).
419
53. Bertelsmeier, C., Luque, G. M., Hoffmann, B. D. & Courchamp, F. Worldwide ant invasions under
420
climate change. Biodivers. Conserv. 24, 117128 (2015).
421
54. Capinha, C., Essl, F., Seebens, H., Moser, D. & Pereira, H. M. The dispersal of alien species redefines
422
biogeography in the Anthropocene. Science 348, 12481251 (2015).
423
55. Bellard, C., Leroy, B., Thuiller, W., Rysman, J. ‐F. & Courchamp, F. Major drivers of invasion risks
424
throughout the world. Ecosphere 7, (2016).
425
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
20
56. Lawler, J. J., Ruesch, A. S., Olden, J. D. & McRae, B. H. Projected climate-driven faunal movement
426
routes. Ecol. Lett. 16, 10141022 (2013).
427
57. Ortiz-Gamino, D., Pérez-Rodríguez, P. & Ortiz-Ceballos, A. I. Invasion of the tropical earthworm
428
Pontoscolex corethrurus(Rhinodrilidae, Oligochaeta) in temperate grasslands. PeerJ 20 (2016).
429
58. Callaham, M. A. & Blair, J. M. Influence of differing land, management on the invasion of North
430
American tall grass prairie by European earthworms. Pedobiologia 43, 507512 (1999).
431
59. Ikeda, H., Callaham, M. A., O’Brien, J. J., Hornsby, B. S. & Wenk, E. S. Can the invasive earthworm,
432
Amynthas agrestis, be controlled with prescribed fire? Soil Biol. Biochem. 82, 2127 (2015).
433
60. Madritch, M. D. & Lindroth, R. L. Removal of invasive shrubs reduces exotic earthworm populations.
434
Biol. Invasions 11, 663671 (2009).
435
61. Callaham, M. A. et al. Policy and management responses to earthworm invasions in North America.
436
Biol. Invasions 8, 13171329 (2006).
437
62. Hendrix, P. F. & Bohlen, P. J. Exotic Earthworm Invasions in North America: Ecological and Policy
438
Implications. Bioscience 52, 801811 (2002).
439
63. Simberloff, D. et al. Impacts of biological invasions: what’s what and the way forward. Trends Ecol.
440
Evol. 28, 5866 (2013).
441
64. James, S. W. Systematics, biogeography and ecology of earthworms from eastern, central, southern
442
and southwestern USA. in Earthworm Ecology and Biogeography in North America 2951 (CRC,
443
1995).
444
445
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
21
Tables
446
447
448
Region
Total Number of
species
Native Species
Alien Species
% of Alien
Species
Canada
34
8
26
77
United States
193
135
58
30
Mexico
156
106
50
32
North America
308
238
70
23
449
Table 1. Number of native and alien earthworm species in North America.
450
451
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
22
Figures
452
453
454
455
456
457
458
459
460
461
462
463
Figure 1. Relative alien earthworm species richness in North America
464
(A) Map of predicted Relative Alien Species Richness of Earthworm in North America (RASR).
465
The colors indicate the value of predicted RASR in each geographical units. RASR is calculated
466
as the proportion of species within a geographical unit that are alien. (B) Statistical distribution
467
of the predicted values of RASR across the geographical units.
468
469
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
23
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
Figure 2. Correlation between earthworms and aboveground groups alien species richness.
485
Each dot represent the number of Alien Earthworm Species Richness and (A) Plants, (B)
486
Spiders, (C) Mammals, (D) Birds across TDWG level-4 regions in North America. Data of these
487
groups come from Dawson et al.4. The correlations were tested with the Spearman correlation
488
test.
489
490
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
24
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
Figure 3. Historical dynamics of North America's colonization by alien earthworm species.
510
(A) Temporal Evolution of the number of alien earthworm Species recorded in North America
511
(Mexico, United States, and Canada). The number of established species is detailed by the
512
continent of origin of the species. The respective total number of alien species are indicated
513
separately on the right. (B) Temporal trend of interception rate of earthworm specimens at the
514
US borders between 1945 and 1975. The continuous line on top indicates the number of
515
interceptions per year. The continuous line on the right indicates the average monthly number of
516
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
25
interceptions. The color of the tiles indicates the monthly rate of earthworms intercepted at the
517
US Borders.
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
Figure 4. Introduction pathways of alien earthworm species in the United States between
534
1945 and 1975. (A) Each line indicates an intercepted introduction pathway of earthworms at the
535
US border. The green dots indicate the origin of departure, while the red ones indicate the
536
location of interception in the United States. When the exact location of departure was not
537
known, we located it in the center of the country of departure. The color of the lines indicates the
538
transportation means, following (B): The number of species intercepted through the different
539
introduction pathways. Please note that Puerto Rico and Guam islands were part of the United
540
States over this period.
541
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
26
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
Figure 5 Spatio-temporal dynamics of earthworm alien species richness across North
560
America. These maps relates the accumulated number of known alien earthworm species in
561
TDWG4 units across North America at different dates. The colors are proportional to the number
562
of exotic species. Zero means that only native species have been reported at the considered date.
563
Gray colors represents areas with no data.
564
565
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
27
566
567
568
569
570
571
572
573
574
575
576
577
578
Figure 6 Functional niche of earthworm species in North America and functional
579
enrichment brought by aliens (A) Functional niche of Native and Exotic species. The bars
580
represent the number of species that feed or inhabit a specific food or habitat. A species can
581
occur in several categories. (B) Ecological function brought by alien earthworm species.
582
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
28
Methods
583
We built two separate databases: EWINA_RICH, which gathers native and alien earthworm
584
species richness across North America, and EWINA_IPATHS, which collates data of earthworm
585
interceptions at the US Borders.
586
EWINA_RICH: Native and alien Species richness
587
To estimate the magnitude of introductions of Alien earthworm across North America, we
588
calculated the Relative Alien Species Richness (RASR) of each of the North American
589
geographical units, as defined below. RASR was computed as the ratio of the number of alien
590
species versus the total number of earthworm species (natives and aliens). This metric has
591
numerous advantages: it is independent of region size and scale of analysis, it is comparable
592
across regions and ecosystems and between groups of taxa65. We used only data from 2000 to
593
2021 for the modeling and mapping of RASR.
594
To build the database, we compiled data of both native and alien earthworm species occurrence
595
across 2510 geographic units in North America (Mexico, United States, and Canada, Greenland),
596
covering 73% of the land surface (Extended Data Fig. 1). The data gathers the species name, the
597
locality and date of observation, and when available, the abundance, the geographical
598
coordinates and habitat features. The database contains individual records of species, but also
599
community data, where all species were sampled together at the same place. We used the finest
600
possible geographical resolution to define the geographic units, which in the US and Canada is
601
the level 3 of the GADM (https://www.gadm.org). It corresponds to counties in the United States
602
and to Parishes in Canada. The earthworm data were collated from 459 sources dating from 1891
603
to 2021. Data were searched in several manners. First, we compiled all data published or edited
604
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
29
by the authors. J.W Reynolds (JWR) published over 500 articles and 25 books about earthworm's
605
spatial distribution in North America. He is the chief editor of the Journal Megadrilogica since
606
1972. This journal specializes on the ecology of earthworm in North America. JWR regularly
607
published accounts of earthworm distribution in North America. As a senior author and editor, he
608
has a deep knowledge of the literature published regarding earthworms in North American.
609
Carlos Fragoso worked over 35 years on the distribution of earthworms of Mexico and published
610
over 75 articles and 46 book chapters, with a specific emphasis on native and exotic species. He
611
published the great majority of available data regarding Mexico. In a second step, we reviewed
612
systematically all the 391 articles published in the Journal Megadrilogica. We identified 230
613
papers from this journal with earthworm data and extracted the data from the text and from the
614
maps. After this step, we compiled all references cited in these papers, get all possible original
615
sources, and entered their data. To complete this process, we compiled the list of species present
616
in North America and did a specific search for each native species. Once this step was achieved,
617
we established the list of authors who published data on native earthworm in North America and
618
did a systematic search of all the papers published by these authors. Finally, we did classical
619
research on web of science and Google Scholar, with two approaches. We combined the key
620
word "earthworm" or "oligochaete" and a key word for location. We used several location key
621
words such as country names and each state or province names of North America. After this step,
622
we used the list of all native species to do a search on web of science and in GBIF. In order to
623
minimize identification issues in GBIF, we removed all GBIF data originating from human
624
observations and kept only the trusted ones from museums and molecular biology portals. Once
625
all data were compiled, we homogenized the taxonomy with the Csuzdi's database66 and the
626
GBIF Backbone Taxonomy. We aggregated subspecies at the species level. We obtained 68 938
627
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
30
records, from which we eliminated the duplicates based on their locality and date of observation
628
(month and year). At the end, the database holds 66 512 unique species occurrence. With this list
629
of species occurrences, we computed the list of species by geographical units at two resolutions:
630
at county level and at TDWG-level 4 units. TDWG4 units are standardized geographical entities
631
used to compare biodiversity data across taxa31. Counties boundaries were obtained from GADM
632
level 2 (parishes) in Canada, from the tiger 2014 database of the US Bureau of Census for the
633
US, and from GDAM level 1 for Mexico. The TDWG-4 boundaries were obtained from a paper4
634
on invasive species from which we extracted data (see below). The different spatial layers were
635
then merged with snapping activated and cleaned by geo-scripting to fix geometry issues such as
636
duplicated points, auto intersections or sliver polygons. Last, we checked and fixed topology by
637
neighborhood analysis and simplified the boundaries of the polygons, while preserving topology,
638
with mapshaper.
639
640
EWINA_IPATHS: Introductions pathways' database
641
The second database, "EWINA_IPATHS" (Earthworm Introduction Pathways) documents the
642
introduction pathways of earthworm in the United States. This database centralizes data of
643
earthworm interception events at the US borders between 1945 and 1975. These data came from
644
the U.S. Bureau of Plant Quarantine, U.S. Department of Agriculture. They were principally
645
reported by G.E. Gates in a list of articles published between 1956 and 1982. Each record in the
646
EWINA_IPATHS database relates an interception event of introduced earthworms. Interception
647
events are described by the name of the intercepted species, its abundance, the date and locality
648
of interception, the geographical point of origin, the transportation mode (boat, plane, car), and
649
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
31
the substrate in which the earthworms were found (e.g. soil, leaves, fish bait). EWINA_IPATHS
650
contains 1016 events of earthworm interceptions.
651
Definition of alien and native species
652
The geographical origin of earthworm species is relatively well established at the country level in
653
North America. Historically, due to the absence of earthworm fossils, the origin of species was
654
estimated from the distribution of group of species and genus. Indeed, it is well known that
655
certain genus such as Lavellodrilus are restricted to certain regions of North America and are
656
considered native for this reason. More recent work based on molecular biology revisited these
657
first imputations. They confirmed that in general alien species belong to distinct taxonomic and
658
phylogenetic groups67. For our analyses, we considered the native/alien status of each species
659
both at the continent and at the country scales. At the continent scale, we considered species as
660
native when they were native in any of the three countries of North America. This classification
661
was used for the global analyses of the North American continent, such as the cumulative
662
number of alien species (Fig. 3) or the analysis of species attributes such as functional type (Fig.
663
6A) , reproduction (Fig.S3) and spread dynamics (Extended Data Fig. 7). A few north American
664
species, for example from the genus Ramiellona and Arctiostrotus, were native from only one or
665
two North American countries and could be considered as alien in other parts of North America.
666
This was the case for species that were rare outside a clearly delimited area and were only
667
present in disturbed areas outside of these regions. To take into account this fact, we also defined
668
the native/alien status of each species at the country level. This system was used for the analyses
669
at the TDWG4 levels or county levels (Fig 1, 2, 5, 6B, Extended Data Fig. 6).
670
671
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
32
Environmental drivers of RASR
672
We tested 12 environmental features as potential determinants of RASR level (see Extended
673
Data Table 1 for a full description of all examined variables).
674
First, we selected variables that were identified as global drivers of invasions in several other
675
aboveground groups. These general macroecological variables included climate68 : Mean Annual
676
Temperature (bio1) and Annual Precipitation (bio12), human activity (Global Human Influence
677
Index69), the proportion of grassland and cropland in landscape70, and two habitat features:
678
elevation and elevation heterogeneity (GTOPO3071).
679
Second, we added environmental features that are - according to previous studies - specific to
680
earthworm invasions 50,72. These variables were related to soil properties: the soil carbon content
681
on surface and at 30 cm depth, the soil pH in the ten first centimeters73, as well as the (log)
682
density of roads50 (length of roads by units of surface). We included the effect of the area of
683
geographical unit in order to estimate any bias associated to the size of the geographical units,
684
but this variable had no effect (Extended Data Fig. 4 and 5). Finally, we checked multi-
685
collinearity among predictors using qr-matrix decomposition and pairwise plots. No covariates
686
were correlated.
687
688
Model of RASR
689
We modeled and predicted RASR as a function of the 12 environmental covariates with a
690
random forest model. We used only observations posterior to 2000 and removed geographical
691
units with less than five species, to remove under sampled units. The model was bagged in order
692
to assess the test error without the need to perform additional cross validation to estimate the
693
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
33
predictive accuracy74. Bagging (Bootstrap aggregation) consists in drawing many random
694
samples from the training dataset (bootstraps), fit a separate model on each bootstrapped dataset,
695
and average the predictions. On average each model use around two-third of the observations.
696
The remaining one third observations, not used to fit the model, are referred to as the out of bag
697
(OOB) observations. We can predict RASR and estimate the prediction accuracy of each
698
observation using all the models in which the observation was OOB. The prediction is obtained
699
by averaging these predicted responses. With this approach, the out of the bag error is a valid
700
estimate of the test error for the bagged model74.
701
The performance of the model was checked by plotting the predicted RASR versus the observed
702
RASR (Supplementary Extended Data Fig. 8) and by inspecting the residuals and prediction
703
uncertainty. The model attained an OOB r2 of 0.6 with a OOB prediction error (MSE) of 0.015
704
(Extended Data Fig. 8). We checked for the presence of spatial correlation in the residuals in
705
several ways. To quantify the spatial proximity between geographical units, we used a
706
topological approach that considers geographical units as neighbors when they share a
707
contiguous boundary. This allowed us determining which areal units should be considered as
708
neighbors, taking into account natural barriers and the non-convex shape of the continent. We
709
computed the graph of neighborhood based on topology and used it to test autocorrelation with
710
global and local Moran tests75. Both tests were not significant, indicating the absence of spatial
711
patterns in the residuals of the model (Extended Data Fig. 9). Finally we mapped the prediction
712
uncertainty (Extended Data Fig. 10) and found that it was generally below 20%. We used the
713
jackknife-after-bootstrap76 for bagging to estimate the standard errors based on out-of-bag
714
(OOB) predictions with the package ranger77.
715
716
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
34
Variable importance
717
To hierarchize the effect of the potential drivers of RASR, we used a permutation variable
718
importance approach with the ranger package77. This approach considers a variable important if
719
it has a positive effect on the prediction performance. To evaluate this, a tree is grown in the first
720
step, and the prediction accuracy in the OOB observations is calculated. In the second step, any
721
association between the variable of interest X_i and the outcome is broken by permuting the
722
values of all individuals for X_i, and the prediction accuracy is computed again. The difference
723
between the two accuracy values is the permutation importance for X_i from a single tree. The
724
average of all tree importance values in a random forest then gives the random forest
725
permutation importance of this variable. The procedure is repeated for all variables of interest.
726
727
Influence of variables on RASR
728
We explored the influence of the 12 environmental variables on RASR in two steps.
729
First, we analyzed the influence of the variables separately with centered Individual Conditional
730
Expectation Plots78 (c-ICE Plots). This type of plot shows the effect of a given predictor X_i on
731
the variable of interest Y (RASR here) for an average situation, meaning for all other covariates
732
at their average value, like in Partial Dependent Plots, and also for each individual environmental
733
condition meaning all observed combinations of the other covariates. To achieve this, it
734
computes the expected values of Y across the range of the covariate X_i, while keeping all other
735
covariates constant. C-ICE plots thus allows us to see the average effect of the covariates, and
736
also the heterogeneity of this effect in individual combinations of covariates across the
737
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
35
predictors' space. In these plots, the expected values are centered to their value at the lowest
738
value of the covariate X_i in order to facilitate the comparison among individual conditions.
739
In a second step, we explored potential interactions between predictors by a double approach.
740
We started by drawing all pairwise 2-D Partial dependent plots, restricting the co-variables to lie
741
within with convex hull of their possible values, and searched for potential interactions. Then we
742
used the joint-Variable Importance Measure (JVIMP) test79 to assess the significant interactions.
743
In this approach, two co-variates are paired and their paired VIMP is calculated. The VIMP for
744
each separate variable is also calculated. The sum of these two values is referred to as 'Additive'
745
importance. A large positive or negative difference between 'Paired' and 'Additive' indicates an
746
association worth pursuing if the univariate VIMP for each of the paired-variables is reasonably
747
large. The find.interaction () function of the R package randomForestSRC80 was used with the
748
option vimp. We found no evidence of significant interaction.
749
750
Comparison with the invasion patterns in other groups
751
We compared our results on earthworm to data of invasion patterns on several aboveground
752
organisms. These data came from a global database of invasive species4. We restricted our
753
analysis on taxa with the best geographical coverage across North America: mammals, birds,
754
plants, and spiders. This database reports the number of invasive species in the TGWD-44
755
geographical units previously mentioned. We computed the number of alien earthworm species
756
in the TDWG-level 4 units from the complete EWINA database, excluding Greenland,
757
(Extended Data Fig. 6) and compared it to the above ground organisms with a Spearman
758
correlation.
759
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
36
Data availability
760
Data are available online on the Zenodo datawarehouse. EWINA_RICH, the database of
761
earthworm native and alien species richness in North America at the two spatial resolutions,
762
together with the aboveground alien species richness in TDWG4 units, is available at
763
https://doi.org/10.5281/zenodo.6421014. EWINA_IPATHS, the database of intercepted
764
earthworm introduction pathways from 1945 to 1975 is available at
765
https://doi.org/10.5281/zenodo.6408609. EWINA_SP, the tabular data of Ecological profile of
766
earthworm species found in North America, including geographical origin, alien status,
767
functional role and reproduction mode is available at https://doi.org/10.5281/zenodo.6462831.
768
EWINA_1st_RECORDS, the list of first year of observation of each earthworm species in North
769
America can be found at https://zenodo.org/record/6759725. The sources of the environmental
770
data are given in Extended Data Table 1. The list of data sources is provided in Supplementary
771
Information.
772
773
Code availability
774
R scripts used for the analyses are available at
775
https://github.com/JeromeMathieuEcology/GlobalWorming
776
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
37
Methods references
777
65. Catford, J. A., Vesk, P. A., Richardson, D. M. & Pyšek, P. Quantifying levels of biological invasion:
778
towards the objective classification of invaded and invasible ecosystems. Glob. Change Biol. 18, 44
779
62 (2012).
780
66. Csuzdi, C. Earthworm species, a searchable database. Opusc. Zool. Bp. 43, 9799 (2012).
781
67. Csuzdi, C. et al. Molecular phylogeny and systematics of native North American lumbricid
782
earthworms (Clitellata: Megadrili). Plos One 12, e0181504 (2017).
783
68. Fick, S. E. & Hijmans, R. J. WorldClim 2: new 1-km spatial resolution climate surfaces for global land
784
areas. Int. J. Climatol. 37, 43024315 (2017).
785
69. Wildlife Conservation Society - WCS & Center for International Earth Science Information Network -
786
CIESIN - Columbia University. Last of the Wild Project, Version 2, 2005 (LWP-2): Global Human
787
Influence Index (HII) Dataset (Geographic). (2005).
788
70. Klein Goldewijk, K., Beusen, A., Doelman, J. & Stehfest, E. Anthropogenic land use estimates for the
789
Holocene HYDE 3.2. Earth Syst. Sci. Data 9, 927953 (2017).
790
71. Earth Resources Observation And Science (EROS) Center. Global 30 Arc-Second Elevation
791
(GTOPO30). (2017) doi:10.5066/F7DF6PQS.
792
72. Holdsworth, A. R., Frelich, L. E. & Reich, P. B. Regional extent of an ecosystem engineer: Earthworm
793
invasion in northern hardwood forests. Ecol. Appl. 17, 16661677 (2007).
794
73. Shartell, L. M., Lilleskov, E. A. & Storer, A. J. Predicting exotic earthworm distribution in the northern
795
Great Lakes region. Biol. Invasions 15, 16651675 (2013).
796
74. James, G., Witten, D., Hastie, T. & Tibshirani, R. An Introduction to Statistical Learning: with
797
Applications in R. (Springer-Verlag, 2013). doi:10.1007/978-1-4614-7138-7.
798
75. Bivand, R. S., Pebesma, E. & Gómez-Rubio, V. Applied Spatial Data Analysis with R. (Springer New
799
York, 2013). doi:10.1007/978-1-4614-7618-4.
800
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
38
76. Wager, S., Hastie, T. & Efron, B. Confidence Intervals for Random Forests: The Jackknife and the
801
Infinitesimal Jackknife. J. Mach. Learn. Res. 15, 16251651 (2014).
802
77. Wright, M. N. & Ziegler, A. ranger: A Fast Implementation of Random Forests for High Dimensional
803
Data in C++ and R. J. Stat. Softw. 77, (2017).
804
78. Goldstein, A., Kapelner, A., Bleich, J. & Pitkin, E. Peeking Inside the Black Box: Visualizing Statistical
805
Learning with Plots of Individual Conditional Expectation. ArXiv13096392 Stat (2014).
806
79. Ishwaran, H. Variable importance in binary regression trees and forests. Electron. J. Stat. 1, 519537
807
(2007).
808
80. Ishwaran, H. & Kogalur, U. B. Fast Unified Random Forests for Survival, Regression, and
809
Classification. (2021).
810
811
812
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
39
Acknowledgments:
813
We thank the people who helped with digitizing information on the distribution of native and
814
alien earthworm species, particularly E. Pedarros, S. Dandrifosse, C. Mathieu, J. Nabias, and T.
815
Allain. We highly appreciate the comments and suggestions of Isabelle Gounand, Gérard Lacroix
816
and Olivier Moine, and the help of W. Dawson and R. Early regarding the use of their databases,
817
as well as the help of Marvin Wright regarding the use of the package ranger. Finally, we thank
818
the Animal Plant Health Inspection Service from the United States Department of Agriculture
819
and the Invasive Alien Species & Domestic Programs Section from the Canadian Food
820
Inspection Agency for their help regarding the import-export regulation laws. Funding was
821
provided by the France-Stanford Center for Interdisciplinary Studies.
822
823
Author contributions
824
J.M. and E.H. conceived the original idea and secured the principal funding. J.M. curated the data
825
and ran the analyses. J.M. J.W.R. and C.F. searched for data, J.M. wrote the first draft and all
826
authors reviewed and edited the manuscript.
827
828
Competing interests
829
The authors declare no competing interests.
830
831
Additional information
832
Supplementary Information is available for this paper.
833
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
40
Correspondence and requests for materials should be addressed to Jérôme Mathieu
834
Reprints and permissions information is available at www.nature.com/reprints.
835
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
41
Extended data
836
837
838
839
840
841
842
843
844
845
846
Extended Data Fig. 1 Geographical coverage of the EWINA database
847
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
42
848
849
850
851
852
853
854
855
856
857
858
Extended Data Fig. 2 Geographical range of Native and Alien species across North
859
America
860
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
43
861
862
863
864
865
866
867
868
869
870
871
Extended Data Fig. 3 Reproduction type of native and alien earthworm species in North
872
America
873
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
44
874
875
876
877
878
879
880
881
882
883
884
885
Extended Data Fig. 4 Variable importance in the random forest model of earthworms'
886
RASR
887
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
45
888
889
890
891
892
893
894
895
896
897
898
899
Extended Data Fig. 5 C-ICE plots: influence of the covariates on predicted RASR.
900
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
46
901
902
903
904
905
906
907
908
909
910
Extended Data Fig. 6 Earthworm species richness in the TDWG-level 4 geographical units
911
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
47
912
913
914
915
916
917
918
919
920
921
922
Extended Data Fig. 7 Temporal dynamics of the known geographical range of native and
923
alien earthworm species across North America
924
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
48
925
926
927
928
929
930
931
932
933
934
935
936
Extended Data Fig. 8 Model performance
937
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
49
938
939
940
941
942
943
944
945
946
947
Extended Data Fig. 9 Test of the autocorrelation in the residuals
948
949
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
50
950
951
952
953
954
955
956
957
958
959
960
Extended Data Fig. 10 Map of predictions' uncertainty
961
962
963
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
51
Extended Data Figure legends
964
Extended Data Fig. 1 Geographical coverage of the EWINA database
965
Data that are available in the EWINA database are indicated in dark grey, while geographical
966
units with no data are indicated in light grey. A/ Complete coverage. B/ Coverage for the period
967
2000-2021. Only the data from 2000-2021 were used for modelling RASR.
968
969
Extended Data Fig. 2 Geographical range of Native and Alien species across North
970
America
971
Each horizontal bar represents the geographical range of a species. The range is expressed as the
972
proportion of geographic units where the species occur. The color of the bars indicates the origin
973
of the species. The 15 most widespread species are list in the box, by decreasing order of
974
geographical range. * B. rubidus is a synonym of Dendrodrilus rubidus, m : species considered as
975
alien in Mexico, c species considered as alien in Canada, cm species considered as alien in Canada
976
and Mexico.
977
978
Extended Data Fig. 3 Reproduction type of native and alien earthworm species in North
979
America
980
A : Amphimictic : reproduction sexual and biparenthal, AP : generaly amphimictic with
981
parthenogenesis in some morphs, P : parthenogenetic, reproduction uniparental, PF :
982
parthenogenesis facultative 2 = 42.8, p = 1.15 10-8).
983
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
52
984
Extended Data Fig. 4 Variable importance in the random forest model of earthworms'
985
RASR.
986
Variable importance represents the magnitude of accuracy loss of the predictions across all trees
987
when randomizing the considered variable. The procedure is repeated for all variables of interest.
988
Variable importance was computed with the ranger package77.
989
990
Extended Data Fig. 5 C-ICE plots: influence of the covariates on predicted RASR.
991
The thick yellow line indicates the effect of the covariate for an average environmental condition
992
(All other covariates being at their average value). The blue lines indicate the effect of the
993
covariate for the individual condititions across the training dataset. The heterogeneity of the
994
individual blue lines indicates the variability of the predictor across individual conditions.
995
996
Extended Data Fig. 6 Earthworm species richness in the TDWG-level 4 geographical units
997
Maps of the species richness of earthworm across the TDWG-level 4 regions. The TDWG
998
regions are standardized geographical units defined to compare species richness across taxa at
999
large scale 31. The maps detail the (A) Alien, (B) Native, and (C) Total species richness of
1000
earthworm. These data were used for the comparison with above ground taxa (Fig. 2).
1001
1002
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
53
Extended Data Fig. 7 Temporal dynamics of the known geographical range of native and
1003
alien earthworm species across North America. Each line represents the geographical range of
1004
a species over time. Geographical range is expressed as the percentage of geographical units
1005
(county level, Extended Data Fig. 1) that are occupied by a species.
1006
1007
Extended Data Fig. 8 Model performance
1008
This plot shows the predicted value of each observation, using a model in which the observation
1009
was not used to train the model (OOB test data), against the observed RASR. The closer the data
1010
are from the line, the best the model performs. We can see that the model tends to slightly
1011
underestimate RASR in actual high values RASR and slightly overestimate RASR values in
1012
actual low values of RASR. OOB r2 is 0.6.
1013
1014
Extended Data Fig. 9 Test of the autocorrelation in the residuals
1015
These plots are aimed at testing the existence of spatial correlation in the residuals of the random
1016
forest. (A) The global Moran's I test the existence of global correlation by a permutation
1017
procedure. The distribution of the random values are indicated by the bell shaped curve. The
1018
observed value of the global Moran, indicated by the vertical bar, falls within the null
1019
distribution, indicating an absence of autocorrelation. (B) The correlation is tested between
1020
geographical units separated by an increasing number of neighbors (lags), indicated on the X
1021
axis. This allows to test the existence of spatial correlation at precise lags of neighboring. The
1022
horizontal line indicates a null spatial correlation. The whiskers indicate the confidence interval
1023
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
54
of the Moran indices for each lag. There is significant correlation when the whiskers do not
1024
intercept the horizontal line. Here, the trend is fairly flat with the majority of Moran's value being
1025
near 0, indicating the absence of spatial autocorrelation of the residuals.
1026
1027
Extended Data Fig. 10 Map of predictions' uncertainty
1028
This map shows the uncertainty of the predictions made by the model. This is useful to interpret
1029
the predicted values mapped in Fig.1. We see that the uncertainty is overall relatively low.
1030
1031
1032
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
55
Extended Data Table 1 Detail of the environmental covariates used to model earthworm RASR.
1033
Source
Weblink
Native
Spatial
Resolution
Geographical features
Area
GADM 3.6
https://www.gadm.org/data.html
-
Elevation
gtopo30 (USGS)
https://doi.org/10.5066/F7DF6PQS
30
Seconds
Elevation
heterogeneity
Climate
bio1 - Annual
Mean
Temperature
Wordclim
https://www.worldclim.org/
10
minutes
bio12 -
Annual
Precipitation
Human Activity
Human
Impact Index
SEDAC
https://sedac.ciesin.columbia.edu/data/set/wildareas-
v2-human-influence-index-geographic/data-download
30
Seconds
% of Area in
grassland
HYDE 3.2.1
https://easy.dans.knaw.nl/ui/datasets/id/easy-
dataset:74467
5 Minutes
% of Area in
croplant
Roads Density
Natural Earth
https://www.naturalearthdata.com/downloads/10m-
physical-vectors/
1:10m
Soil Properties
Soil pH at 10
cm depth
SoilGrids
https://soilgrids.org/
1 km
C0 - Soil
Carbon
Content on
Soil Surface
C30 - Soil
Carbon
Content at 30
cm depth
1034
.CC-BY-NC-ND 4.0 International licenseavailable under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprintthis version posted June 29, 2022. ; https://doi.org/10.1101/2022.06.27.497722doi: bioRxiv preprint
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The first earthworm species named was Lumbricus terrestris Linnaeus, 1758. Since then, there were some 6000 earthworm (Oligochaeta: Megadrili) species names described, from which ca. 3000-3500 are valid. In order to help the orientation in such a huge amount of data a web-based database was created. Each record contains the basic data of the species names described; i.e. family, genus, specific epithet, author, year, reference to the original description and optionally the valid combination of the species name and deposition of type specimens. The database is searchable by every field mentioned and the resulted list can be arranged alphabetically.
Article
Full-text available
This paper presents an update and extension of HYDE, the History Database of the Global Environment (HYDE version 3.2). HYDE is an internally consistent combination of historical population estimates and allocation algorithms with time-dependent weighting maps for land use. Categories include cropland, with new distinctions for irrigated and rain-fed crops (other than rice) and irrigated and rain-fed rice. Grazing lands are also provided, divided into more intensively used pasture and less intensively used rangeland, and further specified with respect to conversion of natural vegetation to facilitate global change modellers. Population is represented by maps of total, urban, rural population, population density and built-up area. The period covered is 10 000 before Common Era (BCE) to 2015 Common Era (CE). All data can be downloaded from https://doi.org/10.17026/dans-25g-gez3. We estimate that global population increased from 4.4 million people (we also estimate a lower range
Article
Full-text available
The family Lumbricidae is arguably the most well-known and well-studied earthworm group due to its dominance in the European earthworm fauna and its invasion in temperate regions worldwide. However, its North American members, especially the genus Bimastos Moore, 1893, are poorly understood. We revised the systematics of the genus Bimastos and tested the hypothesis of the monophyly of North American lumbricids using morphological characters and eight molecular markers. Phylogenetic analyses based on our extensive sampling of Bimastos and inclusion of Dendrodrilus and Allolobophoridella indicated a well-supported clade containing Bimastos and Eisenoides Gates, 1969, and provided the first evidence supporting that North American lumbricids are monophyletic. Assuming the available divergence time estimations and dating of land bridges are correct, it would suggest that the ancestor of this clade arrived North America through Beringia or the De Geer route during Late Cretaceous, and since then the clade has diverged from its Eurasian sister group, Eisenia. The peregrine genera Dendrodrilus and Allolobophoridella are nested within the Bimastos clade; we propose to treat them as junior synonyms of the genus Bimastos, and, contradictory to the commonly held belief of being European, they are indeed part of the indigenous North American earthworm fauna. Morphological characters, such as red-violet pigmentation, proclinate U-shaped nephridial bladders and calciferous diverticula in segment 10 further support this placement. The East Mediterranean–Levantine Spermophorodrilus Bouché, 1975 and Healyella Omodeo & Rota, 1989 are nested within the Dendrobaena sensu lato clade; therefore their close relationship with the North American Bimastos is refuted. Species fit the revised diagnosis of Bimastos are reviewed and keyed, and a new species, Bimastos schwerti sp. nov., is described.
Article
Full-text available
We created a new dataset of spatially interpolated monthly climate data for global land areas at a very high spatial resolution (approximately 1 km 2). We included monthly temperature (minimum, maximum and average), precipitation, solar radiation, vapour pressure and wind speed, aggregated across a target temporal range of 1970–2000, using data from between 9000 and 60 000 weather stations. Weather station data were interpolated using thin-plate splines with covariates including elevation, distance to the coast and three satellite-derived covariates: maximum and minimum land surface temperature as well as cloud cover, obtained with the MODIS satellite platform. Interpolation was done for 23 regions of varying size depending on station density. Satellite data improved prediction accuracy for temperature variables 5–15% (0.07–0.17 ∘ C), particularly for areas with a low station density, although prediction error remained high in such regions for all climate variables. Contributions of satellite covariates were mostly negligible for the other variables, although their importance varied by region. In contrast to the common approach to use a single model formulation for the entire world, we constructed the final product by selecting the best performing model for each region and variable. Global cross-validation correlations were ≥ 0.99 for temperature and humidity, 0.86 for precipitation and 0.76 for wind speed. The fact that most of our climate surface estimates were only marginally improved by use of satellite covariates highlights the importance having a dense, high-quality network of climate station data.
Article
Full-text available
Identifying influences of earthworm invasion and distribution in the northern Great Lakes is an important step in predicting the potential extent and impact of earthworms across the region. The occurrence of earthworm signs, indicating presence in general, and middens, indicating presence of Lumbricus terrestris exclusively, in the Huron Mountains located in the Upper Peninsula of Michigan were modeled using generalized linear models and stepwise regression to identify important environmental variables. Models were then applied to earthworm occurrence data from Seney National Wildlife Refuge, also located in the Upper Peninsula of Michigan to validate results. Occurrence of earthworm signs was associated with high soil pH, high basal area of earthworm preferred overstory species, and north facing aspects. Middens of L. terrestris were associated with high soil pH, high basal area of preferred species, and close proximity to roads. The resulting model for L. terrestris was incorporated into a geographic information system (GIS) to map the expected distribution, both current and potential, across the study area. Results indicate that L. terrestris has not yet fully saturated its potential habitat, as it is currently found close to roads and has yet to establish in most interior forests sampled. Comparing field measured data to GIS layers revealed limitations in the precision of publicly available spatial data layers that should be addressed in future attempts to predict the extent of earthworm invasion across the larger Great Lakes region. However, within the Huron Mountains, it is predicted that the distribution of L. terrestris will cover, at minimum, 41 % of the area.
Article
Biological invasions are a global phenomenon that threatens biodiversity and few, if any, ecosystems are free from alien species. The outcome of human-mediated introductions are affected by the invasiveness of species and invasibility of ecosystems, but research has primarily focused on defining, characterizing and identifying invasive species; ecosystem invasibility has received much less attention. A prerequisite for characterising invasibility is the ability to compare levels of invasion across ecosystems. In this paper, we aim to identify the best way to quantify the level of invasion by non-native animals and plants by reviewing the advantages and disadvantages of different metrics. We explore how interpretation and choice of these measures can depend on the objective of a study or management intervention. Based on our review, we recommend two invasion indices and illustrate their use by applying them to two case studies. Relative alien species richness and relative alien species abundance indicate the contribution that alien species make to a community. They are easy to measure, can be applied to various taxa, are independent of scale, are comparable across regions and ecosystems, and historical data are available. The relationship between relative alien richness and abundance can indicate the presence of dominant alien species and the trajectory of invasion over time, and can highlight ecosystems and sites that are heavily invaded or especially susceptible to invasion. Splitting species into functional groups and examining invasion patterns of transformer species may be particularly instructive for gauging effects of alien invasion on ecosystem structure and function. Establishing standard, transparent ways to define and quantify invasion level will facilitate meaningful comparisons among studies, ecosystem types and regions. It is essential for progress in ecology and will help guide ecosystem restoration and management.
Article
The invasion of exotic earthworms into northern temperate and boreal forests previously devoid of earthworms is an important driver of ecosystem change. Earthworm invasion can cause significant changes in soil structure and communities, nutrient cycles, and the diversity and abundance of herbaceous plants. However, the regional extent and patterns of this invasion are poorly known. We conducted a regional survey in the Chippewa and Chequamegon National Forests, in Minnesota and Wisconsin, U.S.A., respectively, to measure the extent and patterns of earthworm invasion and their relationship to potential earthworm introduction sites. We sampled earthworms, soils, and vegetation in 20 mature, sugar maple-dominated forest stands in each national forest and analyzed the relationship between the presence of five earthworm taxonomic groups, habitat variables, and distance to the nearest potential introduction site. Earthworm invasion was extensive but incomplete in the two national forests. Four of the six earthworm taxonomic groups occurred in 55-95% of transects; however 20% of all transects were invaded by only one taxonomic group that has relatively minor ecological effects. Earthworm taxonomic groups exhibited a similar sequence of invasion found in other studies: Dendrobaena > Aporrectodea = Lumbricus juveniles > L. rubellus > L. terrestris. Distance to the nearest road was the best predictor of earthworm invasion in Wisconsin while distance to the nearest cabin was the best predictor in Minnesota. These data allow us to make preliminary assessments of landscape patterns of earthworm invasion. As an example, we estimate that 82% of upland mesic hardwood stands in the Wisconsin region are likely invaded by most taxonomic groups while only 3% are unlikely to be invaded at present. Distance to roads and cabins provides a coarse-scale predictor of earthworm invasion to focus stand-level assessments that will help forest managers better understand current and potential forest conditions and identify uninvaded areas that could serve as important refugia for plant species threatened by earthworm invasion.
  • Holocene
  • Hyde
Holocene -HYDE 3.2. Earth Syst. Sci. Data 9, 927-953 (2017).