Introduction records compiled from the Global Register of Introduced and Invasive Species (GRIIS) by the ISSG were used as the underlying data for the analysis of pathway trends. The GRIIS dataset provides verified and annotated country checklists of alien and invasive species (Pagad et al. 2018). In addition to species names, each record includes taxonomy, the environment/system in which the species occurs, the provenance/origin of the species, evidence of impact (yes/no), date of introduction or first record, type of introduction, pathways of introduction, mechanism of impact, and references for source information. GRIIS Version 2016.2 includes draft checklists for all 196 countries that are party to the CBD.

Data for 18746 introduction events, involving 4832 alien species in 101 countries, and occurring between the years 1300 and 2017, were available and adequate to conduct a pathways assessment (Fig. 1). Here we define an introduction event as a recorded introduction of an alien species in a country outside of its native range. Each introduction event included the date of first introduction or first record of a species and contained data on either all or some of the following information types: (1) introduction being intentional or unintentional (i.e. ‘pathway type’); (2) ‘pathway category’ (escape, release, transport as contaminant or stowaway, corridors, unaided or unknown); (3) ‘pathway subcategory’ (further details of specific vectors within each pathway category). The data include Animalia, Bacteria, Chromista, Fungi, Plantae, Protozoa and Virus taxa. The 101 countries cover six regions: Africa, Asia, Europe, North America, Oceania and South America (Figure 1; Suppl. material 3: Table S1). These countries encompass a range of different sizes, development status and climatic regions and thus are geographically representative of global data.

Figure 1.

The 101 countries (orange) used to conduct the global pathways assessment. Red open circles indicate small island nations (n = 9) (https://mapchart.net/; accessed 30 July 2019).

A further 5113 species are known to be introduced to the selected 101 countries via known pathways but were not included in analysis as they do not have authoritative information on dates of introduction or first record. These species were therefore excluded and we concentrated on the 4832 species for which the date of introduction in these 101 countries is known. The total number of introduction events currently estimated is ~ 98422, involving ~ 10800 species (including the 5113 species mentioned above). These events, besides known invasive species, include weeds, agricultural pests and diseases, and other non-invasive aliens for which no pathway information or dates of introduction are known.

Information and data on pathways of introduction were extracted during 2016/2017 from sources used to compile national checklists (see Pagad et al. 2018 for information on the general data collation and entry process). Information sources ranged from scientific peer-reviewed literature, databases, reports both published and unpublished and research data. Textual information describing pathways of introduction were documented and then reviewed for categorisation. These categories were inserted into the data collection templates. Because the CPC is relatively new, some of the information from the data sources used pathway terminology that did not fully align with the CBD framework. In these cases, it was necessary to interpret the pathways within the CBD framework, using literature-based pathway information as a guide. This enabled all data to be compiled using the standard categorisation of pathways endorsed by the Parties to the CBD (UNEP 2014). These categories were inserted into the data collection templates.

Information on dates of introduction or first record and information related to the three levels of the pathway hierarchy for the actual introduction event were recorded - pathway type, pathway category and pathway subcategory. Each introduction event was temporally classified using centuries and decades as classifiers (Suppl. material 2: Appendix S1). First introduction records were aggregated by decade beyond 1800. Decadal scales are appropriate because there is often a lag between detection and reporting events. All records prior to 1800 were aggregated as ‘Pre-1800’. Records from the most recent decade were classified as ‘2011 plus’.

We used the definitions and descriptions of introduction pathways contained in Harrower et al. (2017). This document is the most up-to-date guideline for interpreting the definitions of the CPC and provides examples of the CBD Pathways Categorisation’s application to species information (Harrower et al. 2017). The definitions and descriptions were revised and modified by a panel of experts, using comparisons of the CPC pathway descriptions to descriptions used in (1) the Global Invasive Species Database (GISD), (2) the Delivering Alien Invasive Species Inventories for Europe (DAISIE) database, (3) the Great Britain’s Non-Native Species Information Portal (GBNNSIP) database, and (4) the EASIN information platform (Harrower et al. 2017). Of particular benefit is the distinction between pathway subcategories that appear to overlap. For example, the ‘Contaminant on Plants’ subcategory seemingly overlaps with the ‘Contaminant nursery material’ and ‘Transportation of habitat material’ subcategories. The Harrower et al. (2017) guideline defines and describes the difference between these pathways and treats them in a prescribed order of precedence for category allocation. For example, the ‘Contaminant on plants’ subcategory is defined to contain all contaminants on plants that are not related to the nursery trade, where ‘Contaminated nursery material’ is given precedence over ‘Contaminant on plants’ (Harrower et al. 2017). Despite some shortcomings of the CPC framework, particularly the uncertainty involved in interpreting some subcategories (Faulkner et al. 2020; Pergl et al. 2020), it is a reliable framework with which to report on introduction trends at a transnational level. The CPC framework is still relatively new (2014) and its further development and adoption will facilitate its use as a standardised tool for reporting on alien introductions (Pergl et al. 2020).

For pathway types (i.e. intentional or unintentional introductions), we report trends in terms of both total recorded introduction events for each decade, as well as cumulative introduction events documented between 1800 and 2017. Pathway categories are reported as total number of introduction events per decade for each category. We also report cumulative introduction events for pathway categories, using 1970 as a baseline year. This date was chosen for its comparability with the 1970 baseline used for CBD global biodiversity indicators in Butchart et al. (2010). The dominant pathway subcategories are reported as cumulative introduction events from 1800 to 2017.

We used generalized linear models (negative binomial distribution with log link function) to quantify changes in the recorded number of introduction events over time (introduction events ~ decade). This was conducted at all introduction pathway levels (pathway type, pathway category, pathway subcategory). For subcategories, only the pathways with more than 100 introduction events (n = 18 subcategories) were considered.

Figure 2.

Overview of the hierarchical, standard categorisation of pathways. Six pathway categories and 44 pathway subcategories are broadly categorised into a) intentional transport and introduction of taxa, b) pathways of unintentional introduction and c) pathways by which taxa move to new regions, without direct transportation by humans (i.e. Pathway types). Adapted from Harrower et al. (2017).

There was a total of 8172 (43.59%) intentional and 10574 (56.41%) unintentional documented introduction events of alien species across the 101 countries (Table 1). Since 1800, steady and significant increases in both documented intentional and unintentional introduction events have occurred (Table 2; Figure 3a, b). From 1800 to 1900, both pathway types showed similar cumulative increases in introduction events, but with more documented intentional introduction events than unintentional events (Figure 3b, Table 3). Post 1900, the overall number of documented unintentional introductions per decade was higher than intentional introductions (Figure 3a, b; Table 3). Decadal increases in documented introduction events ranged between 5.79% (1800–1810) and 23.15% (1860–1870) for intentional introductions and between 7.19% (1800–1810) and 24.68% (1890–1900) for unintentional introductions (Table 3). The average decadal increase in intentional and unintentional introductions was 13.12% and 15.29%, respectively (Table 3). The decade of 1991–2000 had more documented introduction events than any other decade in the time series (Figure 3a).

Figure 3.

Decadal changes in the documented number of intentional and unintentional introductions of alien species for 101 countries. Trends in introduction events (n = 18746) involving over 4800 alien species are shown as: a number of documented introduction events, and b the cumulative number of documented introduction events. An introduction event in this figure represents one species introduced outside of its known native range for the first time and into one of the 101 countries in the pool.

The documented number of introduction events for each pathway category has increased significantly per decade since 1800 (Table 2; Figure 4a). The ‘escape’ pathway is the most prevalent pathway by which species introductions are known to occur (37.84%), followed by ‘unknown’ pathway introductions (32.19%; Table 1; Figure 4a). Post 1970 trends show both escape and unknown pathways increased dramatically in cumulative number of introduction events, with 3177 and 2350 additional events, respectively, occurring between 1970 and 2017 (Figure 4b). This is equivalent to 81.38% (escape) and 58.14% (unknown) of the total number of pre-1970 documented introduction events. The remaining five pathway categories had fewer cumulative introduction events compared to escape and unknown pathways, the highest being ‘transport-contaminant’ (1982 events by 2017) and the lowest ‘unaided’ (148 events by 2017; Figure 4c). Of these five pathways, ‘transport-stowaway’ showed the steepest cumulative increase in introduction events post 1970 (Figure 4c).

Figure 4.

Changes in the six main pathway categories (as well as the number of introductions via unknown pathways) a the documented number of introduction events (n = 18746) of alien species per decade since 1800 for 101 countries b–c cumulative number of documented introduction events by pathway since 1970 (note different scaling on y-axes).

The 18 pathway subcategories with more than 100 introduction events since 1800 ranged from 109 to 7203 records (Table 1). The top 18 subcategories were representative of all pathway categories. ‘Unknown’ was the pathway subcategory associated with the most documented introduction events (7203; 38.42%), followed by three subcategories from the escape pathway: ‘horticulture’ (3581; 19.10%), ‘agriculture’ (1193; 6.36%) and ‘aquaculture/mariculture’ (1010; 5.39%; Table 1; Figure 5a). Many of the subcategories showed sharp rates of increase, particularly from the beginning of the twentieth century, including ‘hitchhikers on ships’, ‘ship ballast water’, ‘ship hull fouling’ and ‘interconnected waterways’ (Figure 5b, c; Suppl. material 1: Figure S1a). In comparison, most subcategories from the escape pathway (except horticulture) had slower cumulative introduction rates, including ‘agriculture’, ‘aquaculture/mariculture’, ‘forestry’, ‘ornamental purpose other than horticulture’ and ‘pet/aquarium species’. All but one subcategory (‘release in nature for use’) significantly increased in introduction events per decade since 1800 (Table 3).

Figure 5.

Changes in the dominant pathway subcategories across decades. Cumulative number of documented introduction events (note different scales on y-axes). The 18 pathway subcategories shown are those with most (> 100) introduction events (see Suppl. material 1: Figure S1 for ‘corridors: interconnected waterways/basins/seas’, ‘unaided: natural dispersal across borders’ and ‘unknown’ pathway subcategories not shown).

Prior to the twentieth century, the cumulative rate of increase for both intentional and unintentional introduction events documented were virtually identical (Figure 3b). The beginning of the twentieth century saw unintentional surpass intentional introductions, a trend that has continued up to the present. The increase in unintentional introductions is likely due to the rise in international trade, which is widely acknowledged as an important factor in allowing alien species to successfully establish in novel geographic regions (Levine and D’Antonio 2003; Perrings et al. 2005; Yemshanov et al. 2012; Chapman et al. 2017). In particular, the accidental transport of inconspicuous taxa, such as fungi, microorganisms, pathogens and invertebrates are often associated with global trade, including live plant imports and importation via shipping (Brockerhoff and Liebhold 2017; Chapman et al. 2017; Okabe et al. 2017). Studies from multiple regions including Europe (Chapman et al. 2017; Pergl et al. 2017; Saul et al. 2017), Asia (Hong et al. 2012; Okabe et al. 2017), the US (Liebhold et al. 2012) and the Antarctic (Osyczka et al. 2012; Houghton et al. 2016) have found that these taxonomic groups are more often associated with unintentional pathways. Increases in trade volume and the subsequent rise in accidental introductions of alien species may counteract existing national biosecurity and phytosanitary measures (Brockerhoff and Liebhold 2017). It is therefore important to improve measures for monitoring unintentional introduction pathways to effectively address the ongoing occurrence of accidental alien introductions. Interestingly, although unintentional introductions surpassed intentional introductions, escape (an intentional introduction pathway category) had most associated introduction events (excluding unknown events). This highlights that the prevention and management of intentional introductions are of equal importance to those of unintentional introductions, especially given that the impact realised by alien taxa has been associated more frequently with intentional than unintentional introductions (Pergl et al. 2017).

Our findings corroborate previous studies of alien introduction pathways in several ways. First, ‘escape’ is overall the most common documented pathway category by which alien species are introduced (Turbelin et al. 2017). Second, ‘transport-stowaway’ is becoming an increasingly important introduction pathway, particularly for marine stowaways (Zieritz et al. 2017). Finally, records of introduction events via unknown pathways are prevalent in existing databases and presents an ongoing problem for assessing alien introductions (Katsanevakis and Moustakas 2018). Our global perspective takes into consideration alien species from multiple taxonomic groups but supports similar findings from studies focussing on specific taxonomic groups or regions.

Escape was the most prevalent pathway, with records almost doubling between 1970 and the present (Figure 4b). Escape has been identified as the most frequent introduction pathway across all taxa at global (Turbelin et al. 2017) and national (South Africa; Faulkner et al. 2016) scales, for plants at country- (Czech republic; Pyšek et al. 2011; USA; Lehan et al. 2013) and city-scales (Padayachee et al. 2017) and for both plants (Pergl et al. 2017) and vertebrates in Europe (Saul et al. 2017; Roy et al. 2019) and globally (Saul et al. 2017, van Kleunen et al. 2018). However, the escape introduction pathway has been shown as less important for some taxonomic groups, such as marine species (Molnar et al. 2008; Katsanevakis et al. 2013) or terrestrial arthropods (Essl et al. 2015). Our results highlight the importance of the escape pathway at a multiregional level and emphasise the ongoing need for better containment procedures and greater public awareness of the risks involving escaped organisms, particularly ornamental plants (Ricciardi et al. 2017; Saul et al. 2017, van Kleunen et al. 2018).

Horticulture is the most important pathway subcategory of alien plant introductions (Turbelin et al. 2017, van Kleunen et al. 2018) and was the pathway subcategory with the largest and fastest increase in introduction events (Figure 5a). Agriculture was the second most important pathway subcategory and is also recognised as an important contributor to alien plant introductions (Mack and Erneberg 2002; Richardson et al. 2003). Both horticulture and agriculture are pathway subcategories specific to plants (Harrower et al. 2017) and their combined, high proportion of recorded introductions in the dataset (see Table 1) supports previous studies that show escape (from horticulture or agriculture) is an important pathway for plants.

The importance of escape as an introduction pathway for faunal species is reflected by the high number of introduction events attributed to escape from aquaculture/mariculture (e.g. fish farms) compared with the pet/aquarium trade. Aquaculture/mariculture had the third most introduction events, while records attributed to the pet/aquarium trade remained relatively stable across the assessed time-period (Figure 5a). Aquaculture was found to be the highest contributing pathway to freshwater alien species introductions in Europe (Nunes et al. 2015) and an important pathway for alien invasions of European seas (Nunes et al. 2014). The ecological impacts of invasion via aquaculture can be severe (Naylor et al. 2001; Keller et al. 2011) and given the aquaculture sector is one of the fastest growing global primary industries (Teletchea and Fontaine 2014), it is also likely that alien introductions via this pathway will continue to rise.

Subcategories of the transport-stowaway category were among those with largest growth in alien introductions since 1970 (Figure 5b). In particular, there was a sharp rise in the post-1970 introduction of marine stowaways as hitchhikers on ships, in ship ballast water or as ship hull fouling, which saw 67%, 62% and 67%, respectively. The importance of marine/aquatic pathways is also reflected in the sharp rise in introduction events by interconnected waterways since 1970 (Suppl. material 1: Figure S1a). Interconnected waterways were found to be an important pathway in the Mediterranean Sea, largely via the Suez Canal (Zenetos et al. 2012). These increases in alien introductions are likely due to the continued expansion of tourism and international shipping (Early et al. 2016; Turbelin et al. 2017). The introduction of marine and freshwater alien taxa via shipping-related transport has been confirmed as an important source of ongoing propagule pressure in many parts of the world, including the Mediterranean region, Northwest Europe, the Northeast Pacific and Australia (Tingley et al. 2017; Zieritz et al. 2017; Anil and Krishnamurthy 2018).

A key challenge in attempting to decipher trends in alien introductions is uncertainty in the specific pathways used by species (Katsanevakis et al. 2013; Essl et al. 2015). This is particularly problematic for unintentional introductions via transport contaminants or stowaways, and for smaller organisms such as marine invertebrates that are at a higher risk of going unnoticed or undocumented (Essl et al. 2015; Ojaveer et al. 2017; Zieritz et al. 2017). The results shown here demonstrate the problem clearly: the total number of introduction events where a pathway category was unknown far exceeded all other known pathway categories (Figure 4b–c). The exception to this was the ‘escape’ pathway, an intentional pathway category that surpassed the number of unknown introduction events (Figure 4b). Furthermore, ‘unknown’ was the highest-ranked pathway subcategory in terms of the number of introduction events and was almost double that of the second-ranked subcategory (Horticulture; Table 1). These results corroborate previous studies that have demonstrated and highlighted the risk that uncertainty poses to introduction pathway datasets and trends (Zenetos et al. 2017; Katsanevakis and Moustakas 2018).

There are several reasons why uncertainty in pathway identification and trends occurs. Often, the lack of historical introduction records (i.e. pre-mid twentieth century; Ojaveer et al. 2017) can result in gaps in datasets that can particularly impact the interpretation of introduction temporal trends (McGeoch et al. 2010; Katsanevakis et al. 2013; Galil et al. 2018). Usually this occurs due to decreased scientific effort or reduced awareness of the need to record alien species introductions (Ojaveer et al. 2017).

In many cases, multiple pathways are equally tenable as the cause of alien species introductions to a new region (Minchin 2007). This makes assigning the correct pathway difficult and decisions may be entirely based on the interpretations or assumptions of assessors (Zenetos et al. 2012). In other instances, the species’ ecology may be used to infer an introduction pathway (Zenetos et al. 2012). A representative example of this is the introduction of marine species into the Mediterranean Sea via the Suez Canal. Several pathway subcategories could feasibly be responsible for new introductions into the Mediterranean, including species as hitchhikers on ships, through ship ballast water or hull fouling, or through natural dispersal through the canal (Katsanevakis et al. 2013). These types of uncertainty can potentially over- or under-emphasise certain pathways, causing trends to be misrepresented at both global and regional scales. Using a confidence score in allocating pathways may provide a cautionary approach to the compilation of pathway data that helps identify which species, pathways or regions are particularly susceptible to uncertainty (Essl et al. 2015). A focus on improving monitoring of these identified species, pathways or regions may aid efforts to alleviate uncertainty in pathway data. Confidence scores have been successfully integrated into other alien-focused, standardised frameworks, such as the Environmental Impact Classification of Alien Taxa (EICAT; Blackburn et al. 2014; Hawkins et al. 2015) and have recently been used in assessing alien introduction pathways in Europe (Pergl et al. 2020).

The compilation of pathway data from multiple countries or regions can also be a source of uncertainty. Data is often unavailable in many countries, due to a lack of adequate monitoring, data collection efforts or funding (Latombe et al. 2017). Compiling data at national or regional levels usually requires a well-established network of contacts and managing these networks can expend a great deal of time and effort (Zenetos et al. 2017). Furthermore, having multiple pathway data sources will result in multiple ways in which the data is formatted, leading to discrepancies between data. Enacting a standardised framework such as the CPC to filter and arrange pathway data will ensure that trends in introductions of alien species are reported accurately. This is crucial if pathways of introduction are to be considered as an accurate indicator for alien species invasions (Wilson et al. 2018). Given that trends in pathway introductions change over time and across regions, the accuracy and standardisation (or lack thereof) of data can greatly benefit or hinder monitoring and biosecurity efforts (McGeoch et al. 2016; Latombe et al. 2019).

Developing indicators from standardised pathway data is necessary for accurate reporting of alien introduction trends. These indicators can then be used to identify the shortcomings in invasive alien species management and policy targets and help improve legislation for dealing with biological invasions (McGeoch et al. 2010; Hulme 2015). Predictive tools such as risk assessments and horizon scanning can incorporate pathway indicators to better estimate the susceptibility of regions to invasion and identify those species that will pose the greatest introduction threat (Hulme 2015; Rabitsch et al. 2016). The continual input of new pathway data will be needed to ensure that indicators remain up to date and to prevent policy decisions relying on historical pathway patterns (Latombe et al. 2019). Given that the Strategic Plan for Biodiversity 2011–2020 is coming to an end, and the 2021–2030 phase is about to begin, the development and testing of pathway indicators for tracking invasive alien species trends becomes increasingly urgent (Rabitsch et al. 2016).