Effects of biogeographical origin on the �owering phenology of exotic plant communities

Although numerous attempts have been made to clarify traits associated with the invasiveness of exotic species, the effect of the biogeographical origin of exotic species on trait variability in invaded regions in a community level remains poorly understood. In this study, we examined how biogeographical origin affects the �owering phenology of exotic herbaceous plant communities. We collected �owering phenology data for herbaceous plants in Japan based on three types of independent data source: records of 537 exotic species from a picture book; records of 319 exotic and native species from �eld surveys in 1994–2015 conducted using a citizen science approach; and records of 186 exotic and native species from a �eld survey conducted over a 1-year period. We used generalized linear models to examine the effects of three attributes (family, lifeform, and biogeographical origin) on the �owering phenology of exotic plant communities. For comparisons between exotic and native communities, we included habitat factors in the analyses. The results indicated that biogeographical origin has a signi�cant effect on the �owering phenology of exotic plant communities. In Japan, exotic plants of European origin �ower signi�cantly earlier than those of North American origin and native plants. The �ndings of this study highlight the need to consider biogeographical origin when assessing the traits of exotic species.


Introduction
Biological invasion is increasingly being acknowledged as a major threat to the conservation of biodiversity worldwide (Hejda, Pyšek, & Jarošík, 2009 & Richardson, 2010).Since the seminal study of Baker in the 1960s (Baker, 1965), there has been a long history of research that has attempted to identify traits associated with the invasiveness of exotic plants (Daehler, 2003;Pyšek & Richardson, 2008;Richardson & Pyšek, 2006; Van Kleunen, Weber, & Fischer, 2010).
Although numerous studies have focused on traits associated with invasiveness, the possibility that exotic species would have origin-dependent traits in invaded regions has yet to be su ciently examined.
If exotic species do not encounter biotic or environmental ltering through its invasion processes that modify the traits, the species would share similar traits among the communities with the same biogeographical origin even in their invaded region.Flowering phenology, the timing of seasonal owering, is an important functional trait for exotic plants associated with invasiveness (Pyšek & Richardson, 2008).It not only affects the tness of plants but also modi es native pollination patterns and the stability of insect and plant communities ( Aizen, 2018).If species do not encounter biotic or environmental ltering through its invasion processes that modify the owering phenology, exotic plants may have different origin-dependent owering phenologies in invaded communities.In Spain, some exotic plant species did not shift their owering phenology determined by the climate condition in their native range after the invasion (Godoy, Castro-Díez, Valladares, & Costa-Tenorio, 2009).However, there have been no studies that have sought to assess a link between the owering phenology of exotic plant communities and their biogeographical origins (i) at a community level, (ii) considering other factors such as family and lifeform of plants, which are determinants of owering phenology (Davies et al., 2013;Kochmer & Handel, 1986;Ramírez, 2002;Willis et al., 2008;Wright & Calderon, 1995), (iii) through both database and actual eld surveys.
In this study, we examined the effect of biogeographical origin on the owering phenology patterns of exotic plant communities and the relationships between native and exotic plant communities.As a regional model, we focused on herbaceous plant communities in Japan, a country that has accumulated a large number of non-native naturalized herbaceous plants originating from a number of different continents (Muranaka, 2008;Pyšek et al., 2017;Van Kleunen et al., 2015).In this regard, we sought to address the following questions.(a) Does the owering phenology of exotic plants in invaded region differ according to biogeographical origin considering other factors such as family and lifeform?(b) If so, how do relationships between the owering phenologies of native and exotic plant communities in Japan differ according to the biogeographical origins?(c) Does the owering phenology of exotic plants in invade region differ from that in their native range?
We collected data on owering phenology from three independent data sources and analyzed separately: (i) data in a published picture book (Shimizu, Hirota, & Morita, 2001) covering almost all the naturalized plant species in Japan, (ii) citizen's exotic and native plants' owering records collected over a period of 22 years (Maruyama et al., 2021), and (iii) records of monthly eld surveys of exotic and native plants conducted throughout the course of a single year.We veri ed the consistent effect of biogeographical origin on owering phenology using multiple datasets collected by different methods.

Data collection
The Japanese Archipelago study site is located off the east coast of the Eurasian continent in a region characterized by a largely temperate climate.
Historically, for almost 200 years between 1639 and 1854, Japan imposed strict limitations on international trade.However, since the reopening of trade links in the mid-19th century, there has been an increasingly rapid in ux of exotic plants in Japan (Muranaka, 2008), and the present-day Japanese ora comprises a large number of exotic plants from diverse botanical continents (Van Kleunen et al., 2015).
To collect data relating to the owering phenological data of herbaceous exotic plants in Japan, we used three independent data sources and analyzed separately: (i) a picture book covering almost all naturalized plants in Japan (Shimizu et al., 2001); (ii) a site-speci c survey conducted based on a citizen science approach (Maruyama et al., 2021); and (iii) eld surveys conducted over multiple sites during a single year.The latter two included records of co-occurrence native plants.

The picture book
We compiled a list of exotic plant species that have become naturalized in Japan based on reference to the publication Plant Invader 600 (Shimizu et al., 2001), which covers almost all the naturalized plant species that have colonized regions throughout Japan.With the exception of those of species that have only colonized Okinawa Prefecture, which consists of isolated islands lying in a subtropical zone, we obtained the following datasets for herbaceous plant species: species name, family, lifeform, biogeographical origin, and owering phenology.With respect to biogeographical origin, we compiled the information based on the continental scheme proposed by the Taxonomic Databases Working Group (TDWG) (Brummitt, 2001) and classi ed origin with respect to seven areas: Europe, North America, South America, temperate Asia, tropical Asia, Africa, and Australasia.For the purposes of analysis, we excluded species from areas that did not fall into any of these seven continental regions.The owering season of each species was generally described in terms of the four seasons spring, summer, fall, and winter.For data analysis, we represented these seasons as months (spring: March to May, summer: June to August, fall: September to November, winter: December to February) according to the de nition provided by the Ministry of the Environment in Japan.For those species for which the owering season is indicated as "from spring to summer," we designated this as March to August.Plant Invader 600 covers general information on owering season throughout Japan, and thus the stated owering seasons tends to be rough, rather than speci c.For species lacking lifeform data, we obtained the relevant information by referencing alternative publications (Hoshino, Masaki, & Nishimoto, 2011;Numata & Yoshizawa, 1975;Ohashi et al., 2016;Osada, 1993).

Citizen science surveys
To obtain detailed site-speci c owering data covering exotic and native herbaceous plant species, we used eld survey data using a citizen science approach (Maruyama et al., 2021), which previously has contributed as an ecological research tool in collecting phenological data (Dickinson, Zuckerberg, & Bonter, 2010;Silvertown, 2009).Over a 22-year period between 1994 and 2015, monthly surveys were conducted by trained volunteers in a forested area and agricultural elds of the Institute for Sustainable Agro-ecosystem Services (ISAS) and Tanashi Forest owned by The University of Tokyo, Tokyo, Japan, which lie in a warm temperate zone.During each survey, the participants recorded the names of all herbaceous plant species that were owering at the time.The complete dataset and detailed method of the survey are provided by Maruyama et al. (2021).Of the complete dataset comprising 401 species, we used data obtained for 329 species (67 exotics and 262 natives), which the volunteers observed on more than two occasions at monitored sites over the survey period.We de ned the owering season of each species in terms months based on recording owering in at least 5% of the total number of surveys conducted during all months over the 22-year survey period.For data relating to biogeographical origin, family, and lifeform, we referred to relevant publications (Hoshino et al., 2011;Numata & Yoshizawa, 1975;Ohashi et al., 2016;Osada, 1993).

Field surveys
To obtain detailed site-speci c owering phenology data for exotic and native herbaceous plant species, we conducted a speci cally designed eld surveys.At 2-week intervals from June 2020 to August 2021, belt transect (2 m wide; 30 m long) surveys were conducted at eight sites distributed in Saitama and Tokyo prefectures (warm temperate zone).Survey sites were selected in agricultural elds, riparian areas, and hilly semi-natural areas to obtain data on an extensive range of different species owering in different environments, some of which were located at the same ISAS and Tanashi forest sites covered by the citizen science surveys.During the surveys, we walked along the designated transect routes and recorded abundances of the fully blooming in orescences of each species to obtain quantitative owering records.Neither rare nor endangered species were observed during these surveys.For the purposes of data analysis, we calculated mean abundances for each month at each site and summed the abundance records for species at all sites.The owering season for each species was de ned as those months in which owering abundance was at least 5% of the total abundance recorded for all months.For data relating to biogeographical origin, family, and lifeform, we referred to relevant publications (Hoshino et al., 2011;Numata & Yoshizawa, 1975;Ohashi et al., 2016;Osada, 1993).

Flowering phenology in native range
We compared the owering phenology of native plant communities in in Europe and North America and the corresponding communities of exotic plants originating from North America and Europe, in an invaded site, Japan.For comparative purposes, we obtained data on exotic plants of North American and European origin in Japan based on reference to Plant Invader 600 (Shimizu et al., 2001).We collected owering phenological data of native herbaceous plants community in arable elds in Europe and North America.To obtain data on families and owering phenologies of common North American native weeds, we compiled lists of common native weeds based on reference to the dataset assembled by Van Wychen (2020) and veri ed their biogeographical origin based on reference to The PLANTS Database (USDA, NRCS., 2021).We then obtained family and owering phenology data for species from an online version of the Flora of North America North of Mexico (Flora of North America Editorial Committee, eds., 1993).With respect to the family and owering phenology of common native weeds in Europe, we compiled lists of common native weeds found growing on conventional cereal farms in Spain and Germany based on reference to Armengot et al. (2012) and veri ed their biogeographical origin and family by reference to Euro + Med (2006-) (Euro + Med PlantBase).We combined the species list compiled for Spain (Mediterranean climate) and Germany (oceanic, temperate climate) to obtain data for the two climates, which are predominant in Europe, and thereafter obtained data on the owering phenologies of species based on the references respectively (Castroviejo et al., 1986(Castroviejo et al., -2005; Klotz, Kühn & Durka, eds., 2002).

Determinants of exotic plant owering phenology
We investigated the effect of the three attributes family, lifeform, and biogeographical origin on the owering phenology of exotic plants, for which we performed independent statistical analyses for each dataset.We used generalized linear models (GLMs) with normal error distribution and identity link function.Likelihood ratio tests were conducted to calculate P values.As explanatory variables, we used common attributes for all datasets: family, lifeform (annual, biennial, or perennial), and biogeographical origin (natives or exotics from Europe, North America, South America, temperate Asia, tropical Asia, Africa, and Australasia), and all possible two-way interactions.As response variables, we de ned the owering season of each species in terms of the start, peak, end, and duration of owering.Given that few of the monitored species ower through the winter months from December to January, we did not treat owering season as circular variable.De nitions of the start, end, and duration of owering time were the same for all datasets.For example, if the owering season of a species is from March to May, the start value is 3, end value is 5 and duration value is 3.However, we used different de nitions for the peak value depending on the datasets.For the picture book data, we de ned the peak as the mean month of the owering season (Kochmer & Handel, 1986).For example, if the owering season of a species extends from March to May, the peak value was designated 4. For the citizen science data, we de ned the peak as the month with the largest number of ower observations over the 22-year survey period.In cases where there were multiple months with most ower observations, we used the mean of the months.For the eld survey, we de ned the peak owering as the month in which the largest number of in orescences were recorded at the eight study sites.Having conducted GLMs analyses, if the effect of biogeographical origin was signi cant, we performed multiple comparison analysis with Bonferroni correction.In this regard, it should be note that we examined the effect of biogeographical origin in each family only for picture book dataset, given that the interaction between family and biogeographical origin was signi cant.All statistical analyses were performed using R software version 4.0.0 (R Core Team, 2020).

Comparison of the owering phenologies of native and exotic plant communities
To examine the relationships between the owering phenology of native plant community and that of exotic plants of different biogeographical origin, we analyzed data obtained from the citizen science surveys (261 natives vs. 58 exotics) and the 1-year eld survey (126 natives vs. 60 exotics), separately.
For both native and exotic plants, we used the same de nitions of owering season.In all analyses, we used GLMs with normal error distribution and the identity link function, and likelihood ratio tests were used to calculate P values.As response variables, we used the start, peak, end, and duration of the owering season for each species independently.As explanatory variables, in addition to the family, lifeform, and biogeographical origin, we included habitat factors and all possible two-way interactions.
For biogeographical origin, native species was set as Japan.We extracted information relating to the habitat of native plants from a published source (Ohashi et. al., 2015), because habitat condition affects the timing of owering (Thies & Kalko 2004).Habitat was classi ed as disturbed, woodland, lakes and rivers, grassland, and coastal.To determine the in uence of the biogeographical origin of exotic communities on relationships between the owering phenology of exotic and native plant communities, having initially conducted GLM analysis, we performed pairwise multiple comparisons using Dunnett's test in the R package ` multcomp´ (Hothorn et al., 2008).

Comparison of the owering phenologies of exotic plant communities between the invaded and native range
We performed GLM analyses with normal error distribution and identity link function.Likelihood ratio tests were conducted to calculate P values.As response variables, we used the mean values of owering season.As explanatory variables, we used family, species status (exotic plants in Japan or native plants in Europe/North America) and the interaction between family and kinds of datasets to compare the peak of owering seasons between (i) the communities of exotic plants in Japan originating from Europe and native plant communities in Europe, (ii) the communities of exotic plants in Japan originating from North America and native plant communities in North America, and (iii) native plant communities in Europe and North America.Using Plant Invader 600 (Shimizu et al., 2001), we obtained datasets for the majority of exotic herbaceous plants that have colonized Japan (Fig. 1).Among the data obtained for 929 species, we used the datasets of 537 species in 66 families.In the analysis of peak owering season, we found that biogeographical origin, family, life form, and the interaction between biogeographical origin and family had signi cant effects (Table 1).Given our ndings of a signi cant interaction between biogeographical origin and family, we examined the effect of biogeographical origin, life form, and all possible two-way interactions on the peak owering season of species in the families Asteraceae (158 species) and Poaceae (129 species), which were the two families with the highest numbers of recorded species and accounted for more than 30% of all species.The results indicated that only biogeographical origin affected the peak of the owering season for both families (Table 2).Among the Asteraceae species, we found that the peak owering season of European species was earlier than those of North American, South American, temperate Asian, and African origins (Fig. 2a, Table 2).Similarly, for Poaceae, the peak owering season of European species was observed to be earlier than that of North American, South American, tropical Asian, and African species (Fig. 2b, Table 2).Analysis of start of owering season indicated that biogeographical origin, family and the interaction between these two variables had signi cant effects (see Appendix Table S1 in the Supporting information).Similar to our analysis of peak owering, we examined the effect of biogeographical origin on the start of the owering season among species in the families Asteraceae and Poaceae, and consistently found that only biogeographical origin has a signi cant in uence on the start of owering of species in both families (Supporting information Table S2).Among Asteraceae species, European species were found to start owering earlier than those of North American, South American, and temperate Asian origin (Supporting information Table S2), whereas for species of Poaceae, European species started owering earlier than those originating from North and South America (Supporting information Table S2).With respect to the end the owering season, biogeographical origin and family were found to have signi cant effects (Supporting information Table S3), with the end of owering season of European species being earlier than that of species of North American, South American, tropical Asian, and African origin (Supporting information Table S3).Furthermore, the owering seasons of temperate Asian and Australasian species tended to end earlier than that of species originating from tropical Asia (Supporting information Table S3).Our analysis of owering season duration indicated that biogeographical origin and family had signi cant effects (Supporting information Table S4), however, on the basis of multiple comparison test, we were unable to detect any signi cant difference among exotic plants differing in biogeographical origin with respect to owering season duration.A summary of the results obtained based on our analyses the Plant Invader 600 (Shimizu et al., 2001) data is shown in Supporting information Table S5.have tropical Asia or Australasian origins.For analysis of the peak of owering season, the biogeographical origin was found to be signi cant (Table 1), and the owering of European species was found to peak earlier than that of North and South American species (Fig. 3a, Table 1).Similarly, biogeographical origin was also found to have a signi cant effect on the start of the owering season, with the start of owering among European species being earlier than that of North American species (Supporting information Table S1).Contrastingly, we detected no signi cant effects among the assessed variables with respect to the end and duration of the owering season (Supporting information Tables S3 and S4, respectively).

Field surveys
During the 1-year eld survey, we collected 3112 records for 71 exotic plants species, among which, datasets were obtained for 60 exotic herbaceous species in 25 families, none of which are of African and Australasian origin.All species are considered agricultural weeds in Japan.We found that biogeographical origin and family had signi cant effects on peak owering (Table 1), although the results multiple comparison test revealed no signi cant differences in the peak owering seasons of different families (data not shown).With respect to biogeographical origin, the peak owering season of European species was observed to be earlier than that of North American species (Fig. 4a, Table 1).
Biogeographical origin and family were also found to have signi cant effects on the start of the owering season (Supporting information Table S1), although similarly to peak owering, we detected no signi cant differences among the different families based on the ndings of multiple comparison tests (data not shown).European species did, however, start to ower earlier than those of North American origin.With respect to the end of the owering season, signi cant effects were detected for the interaction between biogeographical origin and family (Supporting information Table S3), whereas we identi ed no variables having any signi cant effect on the duration of owering season (Supporting information Table S4).
3.2 Comparison of the owering phenologies of native and exotic plant communities

Citizen science surveys
On the basis of the observations reported by citizen scientists, we obtained complete datasets for 261 native herbaceous species in 66 families and 58 exotic herbaceous plant species in 25 families.Analysis of the peak owering season identi ed biogeographical origin, family, and habitat as variables having signi cant effects (Table 3).The owering season of European species was found peak earlier than that of native species (Fig. 3a and 3b, Table 3), and the owering of species growing in disturbed areas peaked earlier than that of species found in woodland.With respect to the start of the owering season, we detected signi cant effects of biogeographical origin, family, and the interaction between life form and habitat, with exotic species of European origin being found to start owering at an earlier date than native species (Supporting information Table S6).Life form and the interaction between biogeographical origin and family were found to have signi cant effects on the end of the owering season, (Supporting information Table S7), with the end of perennial species owering being earlier than that of annual species.Moreover, the duration of the owering season was shown to be signi cantly affected by the interaction between biogeographic origin and family (Supporting information Table S8).A summary of the results obtained based on our analysis of the citizen science data is presented in Supporting information Table S9.Observations made in our 1-year eld survey enabled us to compile complete datasets for 126 native herbaceous species in 39 families and 60 exotic herbaceous species in 25 families.Among the assessed variables, we found that biogeographical origin, family, life form, and habitat all had signi cant effects on peak owering (Table 3), with the owering of European species peaking earlier than that of native species (Fig. 4a and 4b, Table 3), and that of species colonizing disturbed areas being earlier than that of species growing in woodland areas.Furthermore, the owering of perennial species was observed to peak earlier than that of annual species.Biogeographical origin, life form, and the interaction between family and habitat were found to have signi cant effects on the start of the owering season, with European exotic species tending to start owering earlier than native species, and perennial species starting to ower earlier than annual species (Supporting information Table S6).Regarding the end of owering season, only life form had a signi cant effect, with the end of perennial species owering being earlier than that of annual species (Supporting information Table S7).In addition, the duration of owering season was found to be signi cantly affected by the interaction between biogeographical origin and family (Supporting information Table S8).A summary of the results obtained based on our analysis of the eld survey data is presented in Supporting information Table S10.

Comparison of the owering phenologies of exotic plant communities between the invaded and native range
We compared the owering phenology of native plant communities in arable elds in Europe (42 species) and North America (42 species) and that of exotic communities in Japan, originating from Europe (261 species) and North America (156 species), respectively.We detected no signi cant difference between the owering season of native plant communities in Europe and North America and the communities of exotic species of European and North American origin in Japan, respectively (Table 4, Fig. 5).These results indicate that the owering phenology of exotic plant communities did not differ signi cantly from that of native plant communities.In addition, comparisons of the peak owering season of European and North American native plants revealed that the peak owering season of native communities in Europe did not differ signi cantly from that of native communities in North America (Table 4, Fig. 5).Collectively, the ndings of this study indicate that biogeographical origin was a signi cant source of variation in owering phenology of exotic plant communities in Japan and that owering phenology varied among biogeographical origins (Table 1).These effects were particularly evident with respect to the start and peak of the owering season.Notably, analyses of all three datasets indicated that exotic species of European origin, ower signi cantly earlier (from spring to early summer) than those originated from North America and plants native to Japan (late summer to autumn) (Figs. 2, 3a, and 4a; Table 1), though in the datasets of the picture book, which included many families, this pattern was observed among species within the same family (Fig. 2; Table 2).These ndings thus provide convincing evidence to indicate that biogeographical origin does play an important role in determining the patterns of owering at a community level, considering other factors such as family and lifeform of plants, which are determinants of owering phenology (Davies et al., 2013;Kochmer & Handel, 1986;Ramírez, 2002;Willis et al., 2008;Wright & Calderon, 1995).Pollination type was also assumed to be unrelated to biogeographical origin, because biogeographical origin affected owering phenology of Poaceae species, which are wind-pollinated (Fig. 2b, Table 2).
To assess the ecological factors underlying the effects of different biogeographical origins on owering phenology, we applied the ecological lter model used in community ecology (Funk et al., 2008).In this model, the ecological lter functions to determine nal species composition in a community with a certain subset of traits.In this model, three main assembly processes are considered to be involved in the success of invaders outside their native range: dispersal, biotic and environmental ltering ( (Elzinga et al., 2007).Also, owering phenology will be anticipated to function as s dispersal ltering in invasion process in a certain condition (Knapp & Kühn, 2012).Thus, at the community level, these factors are believed to function as lters for invasive plants with a certain owering phenology.
On the basis of this approach, we propose three non-exclusive hypotheses to explain differences in the owering phenology of exotic plants of different biogeographical origin (Fig. 6a-c).Firstly, it can be posited that the owering phenology of exotic plant communities in an invaded region may directly re ect owering in a native region (Fig. 6a).Flowering phenology of plants has an endogenous genetic determinant associated with climate conditions in their native distribution range to which they are adapted (Dahlgren, Zeipel, & Ehrlén, 2007;Godoy, Richardson, et al., 2009;Rathcke & Lacey, 1985), and this hypothesis assumes that no speci c biotic or environmental ltering acts on owering phenology during the invasion process.If correct, it would be assumed that in Japan, the owering phenology of herbaceous exotic plants community of European and North American origin would not differ from that in their original distribution ranges in Europe and North America, respectively.Moreover, it would be anticipated that herbaceous plants of European origin will ower earlier than those originating from North America.
The second hypothesis proposes that during the invasive process, owering phenology may differ among plants of different biogeographical origin as a consequence of ecological ltering (Goodwin, McAllister, & Fahrig, 1999;Kolar & Lodge, 2001;Theoharides & Dukes, 2007), which has differential effects on plants derived from different biogeographical regions (Fig. 6b).This hypothesis assumes that one or more sources of dispersal or biotic/environmental ltering act on owering phenology during the invasion process.For instance, transport, the rst step of invasion process (Theoharides & Dukes, 2007), may act as a human-mediated dispersal, lter determining owering phenology.For example, as a consequence of agricultural practices, agricultural weeds sometimes have life histories similar to those of the crop plants amongst which they grow (Barrett, 1983).It is thus probable that the timing of the seed dispersal in these weeds will coincide with that of the crop plants, and consequently, the seeds of these weeds contaminating harvested grain may be unintentionally transported into new regions via international trade (Benvenuti, 2007;Shimono & Konuma, 2008).Given this scenario, it might be anticipated that in Japan, the plants of contaminant agricultural weeds of North American origin will ower in autumn, owing to the similar timing of spring wheat harvesting.In contrast, the plants of contaminant agricultural weeds of European origin may ower in spring in Japan, given that this main period during which winter wheat or forage crops such as alfalfa and perennial ryegrass are harvested in Europe.If this hypothesis is correct, it would be predicted that in Japan, the owering phenology of exotic plants originating from Europe and/or North America would differ from that of the corresponding native plant communities in Europe and/or North America, respectively.In this regard, a number of studies have presented lists of weed seeds that have been identi ed as contaminants of imported grain commodities (Mekky et Kurokawa, 2001;Shimono & Konuma, 2008).Accordingly, by assessing these data, it should be possible to con rm whether inter-regional transport serves as a human-mediated lter determining the owering phenology of exotic plant communities.
The third hypothesis postulates that owering phenology may differ among exotic plant communities of different biogeographical origin, owing to changes relating to evolutionary processes or phenotypic plasticity (Fig. 6c).In this regard, the ndings of some studies have indicated that having invaded a new regions, exotic plants can undergo rapid evolutionary changes associated with owering phenology (Lustenhouwer, Wilschut, Williams, van der Putten, & Levine, 2018; Novy, Flory, & Hartman, 2013).
However, this hypothesis is generally considered to be implausible, given that it is di cult to identify any particular ecological processes that would exclusively drive changes in the owering phenology of exotic plants derived from any one speci c origin.

Hypotheses testing
To assess the potential applicability of rst and second of the aforementioned hypotheses, we compared the owering phenologies of exotic plant communities with European and North American origin between the invaded and native range.We accordingly established that the owering phenology of these plants as exotic communities in invaded regions did not differ signi cantly from that in their respective native ranges (Table 4, Fig. 5).These results would conceivably provide evidence in support of the rst hypothesis, which states that the owering phenology of exotic plant communities in Japan directly re ects that in the native ranges of these plants, and that there is no particular ecological ltering in uencing owering phenology during the invasion process.
One of the factors affecting owering phenology is the climate where plants have evolved.For example, in the Mediterranean climate, plants ower mainly in spring to avoid the arid summer and cold winter (Nahal, 1981; Orshan, 1989; De Lillis et al., 1992).In temperate region, where winter cold is harsh, plants ower from spring to late summer (Rathcke & Lacey, 1985).Some studies insisted that some plants conserved their owering phenology in their native range after the invasion (Godoy, Castro-Díez, Valladares, & Costa-Tenorio, 2009; Godoy, Richardson, et al., 2009).In this study, the effect of climate in the native range would be the case for some exotic species, but some would not because datasets on biogeographical origin of exotic plants in Japan were classi ed by the botanical continents proposed by the Taxonomic Databases Working Group (Brummitt, 2001)  Consequently, it is plausible that invasive species with different biogeographical origins could have differing effects on invaded region, thereby posing multiple threats to native ecosystems.Given the diversity of plant-pollinator interactions (Carvalheiro et al., 2014;Morales & Traveset, 2009;Muñoz & Cavieres, 2008), further research will be necessary to establish how differences in the owering phenologies of exotic plant of different biogeographical origin affect native plant-pollinator systems.
Although a number of studies have compared the owering phenology of exotic and native plants, their ndings have been inconsistent (Pyšek & Richardson, 2008).Some studies reported the earlier (Belmonte ) owering of exotic plant community than native one, whereas others have found no signi cant differences in the owering phenologies of exotic and native plant communities (Thompson, Hodgson & Rich, 1995;Williamson & Fitter, 1996).Such disparate observations could well be attributed to differences in the biogeographical origin of exotic plants.
Our ndings highlight the importance of obtaining information relating to biogeographical origin when assessing or predicting the potential threats posed by the introduction of exotic species.For example, in cases where an invasive species of European origin has the potential to colonize regions in Japan, we might predict that it will ower in spring and thereby compete with native ora with similar owering phenologies for available pollinators, which could accordingly have particularly detrimental effects on endangered native species owering in spring.Given that the in uence of exotic plants on native plantpollinator networks is determined to a large extent by disparate owering phenologies (Bartomeus et al.,

2008;
McKinney & Goodell, 2011), management of exotic plants based on a knowledge of their biogeographical origin may be necessary for developing appropriate conservation strategies designed to protect native plant communities.Furthermore, it is reasonable to assume that traits other than owering phenology will also differ among exotic species depending on biogeographical origin, and consequently, further research is required to clarify the effects of biogeographical origin on the different traits of exotic species in invaded regions.

Declarations Figures
Page 28/ Seasonal pattern of the number of owering species from each dataset during each month.

Results 3 . 1
Factors in uencing the owering phenology of exotic plants 3.1.1The picture book

Table 1
Generalized linear models (GLMs) of representatives of owering season of exotic plants for three datasets.Bold type indicates a statistically signi cant difference at the 0.05 level in GLM analysis and any level with Bonferroni correction in multiple comparison among biogeographical origins.

Table 2
Generalized linear models (GLMs) of the mean of the owering season of the two top families based on data obtained from the publication Plant Invader 600.Bold type indicates a statistically signi cant difference at the 0.05 level in GLM analysis and any level with Bonferroni correction in multiple comparison among biogeographical origins.
(Maruyama et al., 2021)obtained for 401 species(Maruyama et al., 2021)based on citizen science surveys, there were dataset for 58 exotic herbaceous plant species in 25 families, none of which, however,

Table 3
Generalized linear models (GLMs) of representatives of owering season of exotic and native plants for two datasets.Bold type indicates a statistically signi cant difference at the 0.05 level in GLM analysis and any level with Bonferroni correction in multiple comparison among biogeographical origins.

Table 4
Generalized linear models (GLMs) of the mean of the owering season of native plants in North America, Europe, and exotic plants in Japan.Bold type indicates a statistically signi cant difference at the 0.05 level in GLM analysis and any level with Bonferroni correction in multiple comparison among biogeographical origins.
al., 2010; Pheloung, Swarbrick, & Roberts, 1999; Wilson et al., 2016), including cases in Japan (Asai, Kurokawa, Shimizu, & Enomoto, 2007; Asai, Kurokawa, Shimizu, & Enomoto, 2009; , which consisted of diverse of climate in a region.Both Europe and North America continents are classi ed into some climate zones; Mediterranean, temperate, continental, etc according to Köppen climate classi cation.Therefore, it is plausible that exotic plants community could have different owering phenology depending on the oristic region, at the continental level in invaded regions.Improved data of global exchange of plants at the continent level, which can nowadays be readily obtained (Van Kleunen et al., 2015), is a key to understanding owering phenologies of exotic plant communities.4.3.Relationships between exotic and native plant communities Based on our observations in the present study, we predict that the effects of exotic plants on invaded native ecosystem would tend to be dependent on the biogeographical origins of these exotic species.Exotic plant species community of European origin showed asynchronous owering with native one, whereas those from continents other than Europe showed synchronously owering.the asynchronous owering phenology could have detrimental effects on native plant communities through reduced pollinator visitation (McKinney & Goodell, 2011) or by providing a pollen source when native plants do not ower (Koyama et al., 2018; Salisbury et al., 2015).Conversely, if exotic plants ower synchronously with native species, these alien plants may either compete with the native ora for available pollinators