British species that are present in Australia have different traits from British species that are not present in Australia

Introduced species spreading to natural ecosystems is a leading cause of environmental change and a key feature of the Anthropocene. While there have been many studies of the traits of introduced and invasive species, less is known about the traits that affect a species' chances of reaching and establishing in new areas. We asked whether British species that are present in Australia have different traits to British species that are not present in Australia.


| INTRODUC TI ON
Plants are introduced into new environments by humans both deliberately and accidentally, to such an extent that introduced species are a defining feature of the Anthropocene (Lewis & Maslin, 2015).
Once in new environments, introduced species may alter ecosystem structure and function, change disturbance regimes and threaten biodiversity (McGeoch et al., 2010).This makes introduced species a major component of global environmental change (Vitousek et al., 1996).Preventing biological invasions before they occur has been recognised as a more effective approach than attempting to control a species after it has been introduced (Leung et al., 2002).
This has led to a large body of literature aimed at predicting which species will become invasive based on biological traits.Prior studies have compared the traits of co-occurring native and introduced species (Ordonez et al., 2010;Pyšek & Richardson, 2007) or of invasive and non-invasive introduced species (Gallagher et al., 2011;Ollivier et al., 2020;van Kleunen et al., 2010).These studies generally focus on the role of traits once a species has arrived in a new range.
However, plant traits could also contribute to a species' chances of being introduced to a new environment in the first place.In this study, we present a broad empirical test of the idea that a plant species' chance of being introduced to and establishing in a new country is related to its traits.
Many of Australia's introduced species come from Great Britain, introduced in the years since Europeans colonised Australia.The flow of people, things and technologies brought flora both deliberately and inadvertently (Beinart & Middleton, 2004;Brockway, 1979).
Rapid worldwide travel, facilitated by globalisation, has led to a significant increase in the rate of species introductions (Capinha et al., 2015;Seebens et al., 2017).Association with humans is a strong indicator of invasion success for birds, fish and mammals in Europe, the United States and Canada (Jeschke & Strayer, 2006) and recent studies have linked plant introduction success with human association (Dehnen-Schmutz et al., 2007;Egawa et al., 2019;Maurel et al., 2016).A greater understanding of how human activities influence introduced plant species' traits is likely to advance our understanding of the mechanisms of invasion.In this study, we compare plant species native to Great Britain that are present in Australia with plant species native to Great Britain that are not present in Australia.
We ask whether these species pools (Figure 1) differ in a suite of traits predicted to influence the likelihood of introduction by humans and/or the chances of species successfully establishing a population in a new range, including flower colour and size, plant height, seedbank longevity, seed mass and geographic extent in Great Britain.
Our first aim is to determine if British species present in Australia have a longer period of seedbank longevity than British species that are not present in Australia.The soil seedbank acts as a reservoir of individuals able to buffer environmental change, and as such is an important trait for understanding the population dynamics and environmental management of introduced species (Gioria et al., 2012).For species to reach and successfully establish in a new range, they must first survive the journey.Before highspeed air travel and modern quarantine procedures, when species' introductions occurred under colonial frameworks (Beinart & Middleton, 2004), the ability for a species' seeds to remain viable for extended periods may have conferred a benefit, especially in unintentional introductions (e.g.species transported via ship ballast or cattle; Hulme et al., 2008;Seebens et al., 2013).Once the seeds reach a new geographical range, the ability to remain viable for longer periods may also enable the species to 'wait' until the right environmental conditions are found before germination, further increasing their chance of successful colonisation of a new range (Gioria et al., 2012).The evidence linking soil seedbank characteristics and plant invasiveness is mixed, with introduced/invasive species having larger and/or more persistent soil seed banks than native/less invasive species in six out of 10 published comparisons (Pyšek & Richardson, 2007).
showing the pool of species present in Great Britain and the pool of species present in Australia, highlighting some of the biotic, abiotic and stochastic filters that species must pass through to become present in a new region.
Our second aim is to determine if British species that are present in Australia have smaller seeds compared to British species that are not present in Australia.In the search for trait differences between introduced and native plant species, seed mass is an oftenresearched trait.Yet the data are mixed, with some studies finding a negative relationship between seed mass and invasion success (Hamilton et al., 2005;Rejmánek, 1996) and other studies finding no relationship (Lloret et al., 2005;Thompson et al., 1995).Seed size may influence the dispersal of seeds past their natural ranges, with smaller seeds more likely to be inadvertently introduced.The effect of seed size on a species' chances of establishing once seeds have reached Australia is less clear.Smaller seeded species have higher rates of seedling mortality, but this appears to be counterbalanced by a shorter time to maturity (Moles, 2018).
Our third hypothesis is that British species present in Australia are taller than British species that are not present in Australia.Evidence suggests that research interest is biased towards conspicuous, colourful and broadly distributed plants (Adamo et al., 2021), and we wondered whether there may have been similar introduction biases towards taller, more conspicuous plant species.Taller plants are also more conspicuous, and their inflorescences are more accessible to humans (Adamo et al., 2021), which might increase the chances of their being transported to a new range.There is also some evidence that height might positively affect species once they have arrived in a new range.For example, success in urban habitats is positively linked to plant height (Thompson & McCarthy, 2008).Previous studies have compared the height of introduced species with native species (Ordonez et al., 2010) and naturalised species with invasive species (Mathakutha et al., 2019;Phillips et al., 2010).These studies have yielded mixed results, finding evidence for introduced species being shorter than native species (Ordonez et al., 2010), taller than native species (van Kleunen et al., 2007;Williamson & Fitter, 1996) and not significantly different from native species (Thompson et al., 1995).
Next, we test the hypothesis that British species that are present in Australia have a larger geographical extent in Britain than British species that are not present in Australia.As humans are major mediators of plant introductions into new ranges (Runghen et al., 2021), more widespread plants may have more frequent encounters with humans, increasing propagule pressure, potentially leading to greater introduction success (Pigot et al., 2018;Wichmann et al., 2009).
Species with a larger geographical extent in Britain might also have greater niche breadths (Sheth et al., 2020), which might increase their chances of finding suitable habitat in which they can successfully establish a population once they have reached Australia.
Finally, we aim to determine whether flower colour and size are different between British species that are present in Australia and British species that are not.Ornamental horticulture is a major reason for species to be imported into new ranges around the world, with domestic and botanic gardens growing 75% and 93% of the world's naturalised non-native flora, respectively (van Kleunen et al., 2018).In Australia, New Zealand and the United States, 50%-75% of invasive plant species were originally imported and sold as garden ornamentals (Csurhes et al., 2006;Dodd et al., 2015;Downey & Glanznig, 2006).Human preference for brightly coloured exotic species could play a role in which species are introduced to new environments.For example, advertising through modern communication networks reinforces strong messages about what is beautiful and trendy, potentially making human-influenced introduction biases more pronounced (Staples, 2001).This is reflected in the over-representation of bright and colourful introduced crayfish in German waterways following the mid-2000s 'crayfish hype' when interest in ornamental crayfish for home aquaria spiked in popularity (Chucholl & Chucholl, 2017).In plants, this effect is less studied but corroborative evidence has been found.For example, many species from the family Acanthaceae introduced to tropical Indo-Pacific islands for their showy and colourful flowers as garden ornamentals have become invasive (Meyer & Lavergne, 2004).Another example is how media imagery of Hawaii over the past 50 years has acted as a hidden factor in the dramatic increase in introduced species, as the tourism industry has worked to create the island paradise that visitors expect (Brock & Daehler, 2021;Staples, 2001).Similarly, Proteaceae species with showy flowers are over-represented in introduced species pools (Moodley et al., 2013), which is consistent with the idea that attractive flowers lead to species being preferentially selected for introduction.In addition to increasing the chances of a species being deliberately transported to a new range, having large, showy flowers might increase the chances that species being cultivated in gardens, from where they might go on to become naturalised through horticultural escapes (Reichard & White, 2001).
Our study tests whether British species that are present in Australia have larger flowers than British species not present in Australia, and whether the proportion of species with different flower colours in each species pool is different.
The presence of British species in Australia could be affected by introduction bias and/or by differences in the likelihood of species successfully establishing a population (Figure 1).Introduction bias describes the process where particular species are more likely to be moved to a new region.Previous studies on animals have provided empirical evidence for introduction biases in amphibians (Tingley et al., 2010) and birds (Blackburn & Duncan, 2001).Similarly, evidence for introduction bias has been found in cultivated plants introduced to Southern Africa and northwestern USA (Minton & Mack, 2010;Omer et al., 2021), as well as at the taxon-scale in Trifolium and the family Proteaceae (Gravuer et al., 2008;Moodley et al., 2013).The establishment of a population in a new range depends on several processes, including stochasticity, filtering resulting from biotic and abiotic pressures, and demography (HilleRisLambers et al., 2012; Figure 1).While we cannot disentangle these processes, our work adds to the understanding of the factors affecting the early stages of species' establishment in new ranges.

| ME THODS
We began with The Vascular Plant Red Data List for Great Britain (Cheffings & Farrell, 2005), which contains 1755 taxa in total and includes all native and archaeophyte vascular plant species of Great Britain.We included all taxa that have been analysed according to IUCN criteria, including taxa listed as threatened and taxa listed as 'Least Concern'.We excluded taxa in the Waiting List, taxa whose native status is yet to be determined, taxa with inadequate distributional or taxonomic data, and taxa on the Parking List.We included archaeophytes (species that may not be native to an area but were introduced/naturalised before 1500 AD [Cheffings & Farrell, 2005]), but excluded neophytes (species that were introduced/naturalised after 1500 AD).Neophytes were identified by comparing data published in the New Atlas of the British and Irish Flora (Preston et al., 2002), with the Atlas of the British Flora (Perring & Walters, 1962).To the list of 1755 taxa, we added 19 plant name synonyms used in Australia, resulting in a list of 1774 taxa.We chose to exclude infraspecific taxa to maximise compatibility between both lists, since our comparative Australian species list detailed below was not resolved to the same taxonomic resolution.This reduced this dataset to 1565 species.
To create the list of British species present in Australia, we accessed the Australasian Virtual Herbarium (AVH) (available at https://avh.chah.org.au/) and conducted a batch taxon search using the list of British vascular plants.The resulting dataset includes records from all major herbaria spanning invasive, naturalised, casual and cultivated specimens.To avoid confounding our analyses, we identified species that are native to both Australia and Britain (using native status from the AusTraits database [Falster et al., 2021; Table S1]), and excluded these species from our analyses.Some species were listed in the Australasian Virtual Herbarium both with and without the qualifier 'sens.lat.'.In these cases, data for the narrower definition was used.We then removed the subset of British species present in Australia from the total list of British species.This resulted in a list of 1170 species of British species which are not present in Australia.After removing infraspecific taxa, cosmopolitan species considered native to both Great Britain and Australia, and harmonising the species list with the Australian Plant Census, the list of British species present in Australia included 395 species.
Note that while we can identify British species that are present in Australia, we do not have sufficient historic information to know whether these species were introduced to Australia from Britain, or from another part of their range.
Data for seed mass, seed longevity, plant height, range size and flower colour (units in Table 1) were downloaded from the TRY plant trait database on 14 September 2021 (Kattge et al., 2020).Where there were repeated observations for a single species in the database, the arithmetic mean was taken to create a single value, except for plant height, where the maximum value was taken.Each analysis included different subsets of the total species list, based on data availability (see Table 1).
Flower size was the only trait in our study not included in the TRY database.To obtain data for each species on this trait we searched each species on the World Flora Online database (available at: http://www.worldflora online.org/).This international online Flora provided a brief description of each species through a managed collection of published floristic data (Borsch et al., 2020).As we were interested in the 'showy' part of the floral display, for composite flowers, such as those in the daisy family (Asteraceae), the measurement taken was of the diameter of the capitula.For grasses and sedges (Poaceae and Cyperaceae), we used the measurement for the spikelet or glume.For umbellate flowers, such as those in the celery family (Apiaceae), the inflorescence measurement was taken as the entire umbel or cyme length.For species with a racemose inflorescence, such as some in the orchid family (Orchidaceae), we used the measurement of the length of the raceme.Where measurements were given for both male and female flowers, the larger one was taken.

| Statistical analyses
All analyses were conducted using R version 3.6.3and R studio version 1.1.463.All data and code that support the findings of this paper are openly available online (https://osf.io/8guw5/?view_only=c2935 955df dd484 aaf48 5f052 a00c895).We analysed flower colour (a categorical variable) with Pearson's chi-squared test for homogeneity.
The assumptions of independence and sample size were met.

TA B L E 1
Final number of species for each trait, average underlying sample size for each species-level average, and units for each plant trait.Note that flower size was compiled separately from World Flora Online.To analyse traits with continuous data (plant height, seed dry mass, geographic extent in the British Isles, seed bank longevity and flower size) we first checked the data distribution for normality, and applied log 10 transformations where appropriate (this was done for plant height, seed mass, geographic extent and flower size).For seedbank longevity, the data contained many zeros, so we added half of the lowest value to all values and then log 10 transformed the data.
We checked the assumption of equal variance visually using a scatter plot and Bartlett's test of homogeneity of variances (p > .05, the null hypothesis that the variances are not equal was rejected).This assumption was met for all variables.We then conducted independent samples t-tests for seed dry mass, plant height, geographic extent, seed (seedbank) longevity and flower size.We tested differences between species pools flower colours using Pearson's Chi-squared test.

| RE SULTS
British species that are present in Australia had significantly larger periods of seed bank longevity compared with British species that are not present in Australia (Figure 2a, t = 9.01, df = 851, p-value <.001).The average seedbank longevity for British species in Australia was ~3 months, whereas the average seedbank longevity for British species that are not in Australia was ~3 weeks.
The mean seed mass for British species that are present in Australia (0.79 mg) was not significantly different to that for British species that are not present in Australia (0.81 mg; Figure 2b, t = −0.11,df = 1179, p = .91).
British species that are present in Australia were significantly taller (mean = 0.94 m) than British species that are not present in Australia (mean = 0.69 m; Figure 2c, t = 4.58, df = 1376, p < .001).
British species that are present in Australia had mean geographic extents in the British Isles larger than British species not present in Australia.British species not present in Australia were found in 263 10 km grid squares on average, compared with 987 grid squares on average for species that are present in Australia (Figure 2d, t = 11.18,df = 1330, p-value <.001).
British species that are present in Australia did not have significantly different flower sizes from British species that are not present in Australia (Figure 2e, t = −0.99,df = 634, p = .32).
The most common flower colour in our dataset was white (28% of the full dataset) followed by yellow (24%), then green (13%).The frequency of the different flower colours expressed in the group of British species present in Australia was not significantly different to the frequency of the different flower colours expressed in the group of British species that are not in Australia (Figure 2f, p-value = .19,df = 9, χ 2 = 12.4).

| DISCUSS ION
We found that British plant species that are present in Australia have significantly greater seedbank longevity, plant height and geographic extents across the British Isles than British plant species that are not present in Australia.This suggests either that a human-mediated introduction bias does exist in the pool of species that are introduced to a new range, and/or that the traits measured in our study affect species' chance of establishing a population in Australia.
Previous studies have found evidence for introduction biases reflected in the traits of species introduced for cultivation, timber production, or use in ornamental horticulture (Maurel et al., 2016;McGregor et al., 2012;Omer et al., 2021;van Kleunen et al., 2007).
Omer et al. ( 2021) compared the cultivated flora of Southern Africa with the global flora and found that the cultivated flora was not a random subset of global flora.Species with larger native range sizes, species from Australasia and freestanding woody species had a greater likelihood of being introduced for cultivation.Subsequently, some of these characteristics were reflected in the naturalised flora of Southern Africa (Omer et al., 2021).Similarly, Maurel et al. (2016) found that introduction biases in plant species introduced for ornamental horticulture in Germany influenced the naturalisation success of these species.They found a positive relationship between naturalisation success and the number of botanical gardens in which a species is grown, minimum residence time since cultivation and native range size.Our work shows that the trends identified in these previous studies apply not just to species deliberately introduced for cultivation, but are also apparent across the full suite of introduced species in Australia.
One important implication of our findings is that studies that compare the traits of native and invasive species (a common approach; see Pyšek & Richardson, 2007) need to be careful to consider the role that traits play in species' introduction and initial establishment as well as their role in conferring invasion success once they are naturalised.For example, several studies have linked plant maximum height to invasiveness (Dehnen-Schmutz et al., 2007;Gallagher et al., 2015;Pyšek & Richardson, 2007), with a common argument being that increased height increases competitive ability and confers an advantage over native species (Díaz et al., 2016;Falster et al., 2018).Our results (Figure 2c) suggest that at least some of the advantage of larger species occurred during transport to/establishment in the new range.
British species that are present in Australia had longer periods of seedbank longevity than those that are not present in Australia (Figure 2a).This could reflect a benefit conferred by a seed's ability to remain viable during long transportation periods, as well as the ability for a species to 'wait' until it is in the right conditions before germination.Previous studies have found that invasive species are more likely to have persistent soil seedbanks (Thompson et al., 1997).Introduced species that can form persistent soil seedbanks also pose an issue for long-term management, due to their ability to withstand environmental changes through the presence of a dormant reservoir of individuals (Gioria et al., 2012(Gioria et al., , 2021)).If species that are reaching and establishing in a new range are more likely to have longer seed bank persistence, then this may facilitate their increased success once in their new environment and confound the effect of seed longevity in studies of invasiveness.
We that British species that are present Australia tend to have larger geographic extents in the British Isles than do British species that are not present in Australia (Figure 2d).This complements studies that have found global or regional native range size to be a good predictor of introduction success (Moodley et al., 2013;Pyšek et al., 2009) and reflects a global trend towards species with larger global native range sizes increasing in abundance, while species with narrower native range sizes decrease (Newbold et al., 2018).A study of over 1 million vegetation plots in Europe found the global native range size of introduced plants was positively correlated with local abundance and habitat breadth (Fristoe et al., 2021).Species with larger global native range sizes and larger niche breadths are more likely to be able to persist when introduced to new environments, while the opposite is true of species with narrower global native range sizes (Sheth et al., 2020).
Plant species with larger geographic extents in their native range  more likely to encounter humans and be brought with them past the limits of their natural geographic ranges.As a result of an increased likelihood of co-existing with humans, species with larger geographic extents are more likely to be better adapted to surviving in anthropogenically altered novel environments, which they are also more likely to encounter post-introduction to a new range (Martin et al., 2009).Owing to the strong positive correlation between geographic extent in the native range, human association and introduction success, the geographic extent of a species in its native range has the potential to be used more often in the screening of introduced species.
We found no evidence that British species with smaller seeds are more likely to be present in Australia (Figure 2b).Previous studies have associated plant invasiveness with small seed mass, in Proteaceae and Pinus species (Moodley et al., 2013;Rejmánek, 1996), and at continental and regional scales (Hamilton et al., 2005).Yet other studies have found the opposite effect, with larger seeded species more likely to be invasive, such as for cactus species on Mediterranean islands (Lloret et al., 2005;Novoa et al., 2016).An explanation for the link between smaller seeds and invasiveness is that small seed mass is inversely correlated with seed output per unit canopy area per year (Henery & Westoby, 2001) and longer persistence times (Thompson et al., 1993), which are characteristics of introduced species that can rapidly spread in novel environments (Groves, 1992).
However, species with larger seeds tend to grow on larger plants with longer reproductive lifespans, and so have similar lifetime seed production to species with smaller seeds (Moles, 2018), potentially balancing out differences in propagule pressure between the two strategies.Similarly, large-seeded species tend to have higher rates of seedling survival (Moles, 2018), which might be an advantage in the early stages of establishment in a new range.However, the available evidence suggests that the early advantage of large-seeded species is likely counterbalanced by longer juvenile periods, with no relationship between seed mass and survival to reproductive maturity (Moles, 2018).Another factor that might contribute to our null result for seed size may be that ornamental and horticultural introductions are the major pathways for species introductions (van Kleunen et al., 2018), and seed size does not influence which species are selected for introduction via these pathways.
The idea that flower characteristics influence whether a plant As historic records of which species were/were not transported to Australia are scarce to non-existent (Dodd et al., 2015), we were not able to distinguish between traits that increase a species' chances of arriving in Australia (the first filter in Figure 1; e.g.introduction bias), and the traits that increase a species' chances of establishing a population once it has reached a new range (the second filter in types of source ranges also affect these trait relationships.Lastly, while we were most interested in functional trait level differences in source and destination species pools, similar analyses with a taxonomic focus could complement our approach.
Our results contribute to a broader body of work showing that the introduced species present in new ranges are often non-random subsets of the source communities.If humans are selecting species for introduction that possess certain traits, either intentionally or unintentionally, and these traits also aid the species in establishing once they reach the new environment, then human-mediated introduction biases could contribute to an increased risk of ecosystem invasion.

ACK N O WLE D G E M ENTS
Thank you to Giancarlo Chiarenza for his time spent providing feedback and assisting with code and to Eva Earle for her contribution to quantifying flower size.

FU N D I N G I N FO R M ATI O N
Author JA was supported by funding through the Australian Government Research Training Program (RTP).
species is introduced to a new range has been raised previously-for example, byCadotte et al. (2006) who said, 'if we were to select a single strong predictor of invasive plant success, it would probably be showy flowers, since most plant introductions are from horticulture'.However, our study presented the first broad empirical test for a relationship between a species' likelihood of being present in a new range and flower 'showiness', indicated through size and colour.Contrary to our predictions, British species present in Australia did not have larger flowers than British species not present in Australia (Figure2e) and the composition of flower colours in the two groups was not significantly different to what could be expected by chance (Figure2f).There are at least two potential explanations for the lack of a relationship between showiness and a species' chances of being present in a new range.First, plants are selected for cultivation in gardens based on characteristics other than the size and colour of their flowers.Popular horticultural plants can be selected due to their winter hardiness(Maurel et al., 2016), being well-matched for the climate(Bayón et al., 2021), edible or ornamental fruit, or other features such as attractive foliage or showy bracts (e.g.Bougainvillea).Second, although ornamental plants are important invaders(van Kleunen et al., 2018), species are also frequently introduced for other purposes, such as forestry, erosion control, agriculture, as biological controls or as landscape 'improvements'(Hulme et al., 2008).These other pathways to introduction, combined with differences in the probability of different species successfully establishing populations once they reach a new range may counterbalance the effect of introductions made based on showy flowers.In different places, people may have preferences for different kinds of plants and different ideas about what constitutes interesting and exotic plants to grow in their gardens for ornamental purposes, which may also decrease the direct relationship between a species' flower characteristics and ornamental value.

Figure 1 :
Figure 1: e.g.environmental filtering).Teasing apart the different ways traits mediate each of these two important steps in the invasion process would be a very valuable direction for future research.Another interesting future direction would be to ask whether the traits associated with establishment in a new range are consistent across different source and destination ranges worldwide.This has the added advantage of testing if different dominant vegetation

Trait name Number of British species that occur in Australia Number of British species that do not occur in Australia Mean sample size for each species-level trait average Unit
plots showing a comparison of British species that occur in Australia and British species that do not occur in Australia.Black dots indicate the mean and lines indicate the standard deviation of the trait for each group, and the violins show the data distribution (the width of the plot corresponds to the data frequency; that is, wide areas are trait values shared by many species, while narrow areas are trait values possessed by few species).Purple represents the group of species present in Australia, and yellow represents the group of species not present in Australia.(f) Bar graph showing the frequency of flower colours for British species that occur in Australia (not hatched) and British species that do not occur in Australia (hatched).A red asterisk indicates significant results.