An Analysis of US State Regulated Weed Lists: A Discordance between Biology and Policy

Abstract

Weedy and invasive plants threaten our food supply, native biodiversity, and the structure and function of ecosystems. The number and impact of these damaging plants are expected to continue to grow with ongoing global change. Some of the most common policy tools to help mitigate this threat are regulatory weed lists, which limit the importation and movement of listed plant species, but there has never been a comprehensive analysis of plants regulated in the United States. We analyzed US state regulatory lists (e.g., noxious, invasive, prohibited) to evaluate their composition, patterns of listing, congruities with weed distributions, and limitations. In total, 46 states maintain regulatory weed lists that include 3210 total listings of 1249 unique species; 48% of them are introduced, 40% are native, and 12% are not yet found in the United States. Overall, the listed species are not a good reflection of the weeds in each state, and listing appears largely reactive, regulating species after they become widespread. We highlight patterns and incongruities among lists and discuss their implications, especially the large number of regulated species native to the United States.

Weeds have been a feature of the human condition since the dawn of agriculture and continue to play a consequential role in shaping global ecosystems and economies. The impacts of weeds are broad, dramatic, and, in some cases, irreversible (Horvath et al. 2018). The future holds little hope for relief, because the introduction of new species continues seemingly unabated (Seebens et al. 2017), and climate change is expected to worsen their impacts (Hellmann et al. 2008). We find ourselves at a watershed moment of rising globalization, emerging economies, an exploding global population, and worsening climate change with the need to mitigate the growing weed threat across ecosystems.

Our best protection against weeds, both economically and ecologically, is to prevent their introduction and to subsequently minimize their spread once they are introduced (Keller et al. 2007). These are the circumstances under which noxious weed laws were conceived, becoming one policy lever that jurisdictions have to limit the introduction and spread of pest plants. Many countries, states, and localities maintain noxious weed policies that vary broadly in their details (McCubbins et al. 2013) but share their motivation in protecting agricultural and natural resources from weeds.

Although pest plants were elements of previous regulations, it was the US Federal Noxious Weed Act (FNWA) of 1974 that formalized and expanded the regulation of weeds (McCubbins et al. 2013). The FNWA established a federal noxious weed list, defining noxious weeds as plants that can directly or indirectly injure agriculture, navigation, fish and wildlife resources, or public health. In 2000, under the Plant Protection Act (PPA) Congress repealed the FNWA and related legislation, consolidating US Department of Agriculture (USDA) authority over noxious weeds and plant pests into a single statute. The PPA revised the original definition of noxious weeds to include injury to the environment: defining a noxious weed as “any plant or plant product that can directly or indirectly injure or cause damage to crops (including nursery stock or plant products), livestock, poultry, or other interests relating to agriculture, irrigation, navigation, natural resources, public health, or the environment.” The PPA focuses on preventative measures to prohibit further spread of noxious weeds, but does not require eradication of established weeds.

Currently, the US Federal Noxious Weed List includes 107 individual species plus all species within five genera, which cannot be moved into or within the United States. Although the federal listing process was previously fairly opaque, listing was recently improved by requiring a weed risk assessment, ensuring a science-based approach that increased transparency (Koop et al. 2012). Despite this, the list changes infrequently, and as Lodge and colleagues (2006) noted, many of the listed species were well established when they were listed, limiting the effectiveness and intent of listing. As McCubbins and colleagues (2013) argued, the states may be better able to tailor their laws to the conditions and circumstances they face; they are, perhaps, more nimble than the federal government. Most state noxious weed laws emerged in the 1970s, but some were created in the early 1900s (McCubbins et al. 2013). This has led to a patchwork of variable noxious weed listing structures, processes, and authorities among the states, which McCubbins and colleagues (2013) and Quinn and colleagues (2013) found to underregulate invasive plants, focusing more on protecting agriculture. Therefore, states have the opportunity to make quicker and more locally relevant species regulations than the federal government, but they should not be viewed as mutually exclusive, because both have their roles.

Previous analyses of state noxious weed lists were focused on identifying the most listed species and their potential for biological control (Skinner et al. 2000) or were conducted with an eye toward understanding their ability to handle the pending invasion threat posed by the then-meteoric expansion of introduced biofuel crops (McCubbins et al. 2013, Quinn et al. 2013). However, these analyses were either simple compilations (Skinner et al. 2000) or compared noxious weed lists to associated invasive plant lists without examining the listed species themselves (McCubbins et al. 2013, Quinn et al. 2013). Given the severity of the growing threat of weedy and invasive species and the possible mitigating force noxious weed lists present (McCubbins et al. 2013), we analyzed US state weed regulatory lists to address several complementary questions: What are the taxonomic, life history, and nativeness status of all state listed weeds? Do nearby states share similar weed flora and weed lists? Are states being proactive or reactive in their listings? What does listing such a large number of species native to the United States mean for policy and management? Addressing these questions will not only improve our understanding of US weed regulations but will necessitate a close look at how we approach a weedier future.

Noxious weed data and species distributions

We obtained regulated species lists for each state from the summary of plant protection laws and regulations documents available on the National Plant Board website (http://nationalplantboard.org/laws-and-regulations), which is a clearinghouse for such lists that is regualrly updated, and the federal list was accessed from the Animal and Plant Health Inspection Service website (www.aphis.usda.gov/aphis/ourfocus/planthealth/plant-pest-and-disease-programs). Not all states regulate plants under a “noxious weed list,” some use other terms such as invasive, pest, prohibited, nuisance, and regulated. Despite terminological variation, all lists are regulated plants. For simplicity, we will refer to these collectively as noxious weeds, the most commonly used term. The composition of these lists is likely based on a variety of local, biological, economic, political, and other factors. We were not able to determine the motivation or criteria of listing, nor who is doing the listing. These analyses would make an excellent complementary follow up study. All noxious weed species referenced within each document were recorded into a central database. Noxious weed seeds were not included in this process as these occur in separate weed seed laws. Each noxious weed species’ scientific name was then cross referenced with the Integrated Taxonomic Information System database (www.itis.gov) and updated if the currently accepted and appropriate scientific name was not used in the summary while also eliminating synonyms. When all species in a genus were listed (e.g., Cuscuta spp.), the database was populated to include all of the accepted species within the genus according to the ITIS database.

The life history (i.e., annual, perennial, biennial), life form (i.e., forb, graminoid, shrub, tree, or vine), and nativeness status (i.e., native, introduced, and not here) for each species was retrieved from the USDA PLANTS database (USDA 2019). Nativeness was assessed at the country scale (the 50 US states, not including territories), because state-level nativeness information is not readily available or is unevenly characterized across states, and nativeness is often represented at the country level (see USDA 2019, GISD 2020). Next, we recorded each species as either present or absent in each state within the United States using BISON (BISON 2019) and the rbison package (Chamberlain 2019) in R (R Core Team 2019). The species listed as present in the BISON database were coded by the PLANTS database nativeness determination, whereas those species listed as absent were coded as not here. For the final data set, all federal noxious weed species were removed from each state's list, because the federal list supersedes state lists (McCubbins et al. 2013). For consistency, all of the analyses were conducted at the species level, even when state policies identified caveats that further divided species (see table 1).

Table 1.

Caveats of species listings, showing examples of when a subset of a taxa is listed, not the entire taxon.

Category	Explanation	Examples
Varieties	Only regulating a certain variety of a species	Cucumis melo var. dudaim (Arizona, California), Foeniculum vulgare (except bulbing fennel, F. vulgare var. azoricum; Washington)
Cultivars	Either excluding or specifically targeting a specific cultivar of a plant	Hedera helix (cultivar Baltica, Pittsburgh, and Star; Washington); Morus alba (except for male cultivars; Wisconsin), Berberis thunbergii (all cultivars; New York)
Nativeness	Identifying a subset of the taxa by nativeness	Hieracium spp. (all nonnative species and hybrids; Washington), Cuscuta spp. (except for natives; Pennsylvania), Orobanche spp. (except for natives; Pennsylvania), Phragmites australis (nonnative Wisconsin; nonnative genotype; Washington), Lythrum spp. (not native to North Carolina; North Carolina).
Agriculture	Exceptions listed for agricultural reasons	Pastinaca sativa (except for the garden vegetable form; Wisconsin), Nasturtium officinale (except for watercress sold for human consumption without its reproductive structure; Connecticut), Daucus carota (except where commercially grown; Washington), Ulmus pumila (except hybrids and individuals used as rootstock; Wisconsin)
Traits	Identifying a subset of a taxa by a trait	Sorghum spp. (perennial; Utah), Sorghum x almum (with rhizomes; Illinois)
History	Based on the history of the known plant	Tamarix ramosissima (unless intentionally planted prior to 2004; Washington)
Management	Based on the current management	Vitis spp. (when growing in groups of one hundred or more and not pruned, sprayed, cultivated, or otherwise maintained for 2 consecutive years; Ohio),

The current state of state noxious weeds

Across the United States, 46 states maintain a noxious weed list with 3210 total listings of 1249 individual species unique from the federal list. Of these states, the number of listed species ranges between a minimum of 0 (Rhode Island and Georgia only include the federally listed species) and a maximum of 321 (Oklahoma; figure 1a, 1b). Excluding states with no listed noxious weeds, the states list an average of 75 species, although there is clearly tremendous spatial heterogeneity in the number of listed species (figure 1a). This variation could reflect the history, motivation, and structure of how the lists were created and updated (McCubbins et al. 2013, Quinn et al. 2013) but also appears to reveal a lack of coordination among states on noxious weed policy (McCubbins et al. 2013). Despite evidence that weed prevention presents large economic advantages (Keller et al. 2007), four states have no noxious weed lists whatsoever (Kentucky, Louisiana, New Jersey, and Wyoming). Importantly, the current lists are an improvement over the past; in 2000, 16 states had no noxious weed laws (Skinner et al. 2000), which improved to 14 states (Min et al. 2008). Our own state of Virginia currently has a very small list of eight species (only three unique from the federal list) but has recently undergone a revision of its noxious weed law, accompanying regulation, and procedures for listing species. These changes were made to make the listing process transparent and based on science, as well as to update the list to better reflect the weed problem facing the commonwealth.

Figure 1.

An overview of the US state noxious weed lists, with listed genera expanded to include all species within a listed genus. Any state listed species that overlapped with the federal list were removed to allow for state-to-state comparisons. (a) A map of the total number of listed noxious weeds by state. The grayscale represents the relative number of weeds. The blank states maintain no list or only list federal noxious weeds (Georgia, Kentucky, Louisiana, New Jersey, Rhode Island, and Wyoming). (b) The relative number (accounting for each time a species is listed, n = 3210), of introduced (not native and present in red), native (in orange), and weeds not here (not native and not here in yellow) for the 10 states with the longest lists. (c) A breakdown of the life history, life form, and most represented five families for all listed introduced, native, and weeds not here.

In 2000, the 34 states with a noxious weed list included just 506 plants (Skinner et al. 2000). There are now over sixfold more listings across the United States, although Skinner and colleagues (2000) did not expand genera as we did. This suggests enhanced state-level responses to the weed threat in the past 20 years through increased listing. The most commonly listed taxa in 2000 were Cirsium arvense, Carduus nutans, Lythrum spp., Convolvulus arvensis, Euphorbia esula, Acroptilon repens, Sorghum spp., Cardaria spp., Centaurea maculosa, and Sonchus arvensis. Two decades later, some of the same species remain at the top of the listing: C. arvense, C. nutans, C. arvensis, and Centaurea spp. (table 2). Interestingly, the majority of species listed in at least 15 states were unintentionally introduced more than 200 years ago (table 2), clearly demonstrating the legacy of historic introductions (Rouget et al. 2016).

Table 2.

The most frequent state-listed noxious weeds in the United States.

		Number of states
Species	Family	Listed	Present	Life form	Life history	Native range	Introduction pathway	Estimated introduction
Lythrum salicaria	Lythraceae	38	45	Forb	Perennial	Africa, Asia, Europe, Oceania	Accidental	Early 1800s
Cirsium arvense	Asteraceae	28	44	Forb	Perennial	Africa, Asia, Europe	Accidental	1600s
Myriophyllum spicatum	Haloragaceae	26	46	Forb	Perennial	Africa, Asia, Europe	Accidental	Late 1800s
Carduus nutans	Asteraceae	23	45	Forb	Biennial	Africa, Asia, Europe	Accidental	1853
Centaurea stoebe	Asteraceae	21	43	Forb	Biennial	Europe	Accidental	Late 1800s
Euphorbia esula	Euphorbiaceae	21	38	Forb	Perennial	Asia, Europe	Accidental	1827
Fallopia japonica	Polygonaceae	21	35	Forb	Perennial	Asia	Intentional	1850s
Egeria densa	Hydrocharitaceae	20	35	Forb	Perennial	S America	Accidental	Early 1900s
Convolvulus arvensis	Convolvulaceae	19	50	Vine	Perennial	Asia, Europe	Accidental	1700s
Rhaponticum repens	Asteraceae	19	7	Forb	Perennial	Asia, Europe	Accidental	1910
Rosa multiflora	Rosaceae	18	43	Vine	Perennial	Asia	Intentional	1866
Sorghum halepense	Poaceae	18	46	Graminoid	Perennial	Africa, Asia	Intentional	Early 1800s
Lepidium latifolium	Brassicaceae	17	22	Forb	Perennial	Africa, Asia, Europe	Unknown	Early 1900s
Myriophyllum aquaticum	Haloragaceae	17	30	Forb	Perennial	S America	Accidental	Late 1800s
Phragmites australis	Poaceae	17	49	Graminoid	Perennial	Africa, Asia, Europe, N America, Oceania	Unknown	1800s
Trapa natans	Trapaceae	17	13	Forb	Perennial	Africa, Asia, Europe	Intentional	Late 1800s
Pueraria montana	Fabaceae	16	28	Vine	Perennial	Asia	Intentional	1876
Butomus umbellatus	Butomaceae	15	20	Forb	Perennial	Asia, Europe	Intentional	1905
Carduus acanthoides	Asteraceae	15	35	Forb	Biennial	Asia, Europe	Intentional	Late 1800s
Centaurea diffusa	Asteraceae	15	25	Forb	Annual	Asia, Europe	Accidental	1907
Conium maculatum	Apiaceae	15	49	Forb	Biennial	Africa, Asia, Europe	Intentional	1890s
Lepidium draba	Brassicaceae	15	22	Forb	Perennial	Africa, Asia, Europe, Oceania	Accidental	1809
Nymphoides peltata	Menyanthaceae	15	26	Forb	Perennial	Asia, Europe	Unknown	1882

Unsurprisingly, the taxonomies, life forms, and life histories of listed weeds are diverse (figure 1c), representing the breadth of the weed problem and the inherent challenges in management. Overall, perennials were far and away the most commonly listed weed life history (figure 1c). This is, in part, because the breadth encompassed in the perennials grouping, which includes perennial forbs, shrubs, and trees. It is also in line with other analyses showing that perennials represent an outsized portion of weedy and invasive taxa (van Kleunen et al. 2019).

Within introduced taxa, forbs were overwhelmingly the most common life form, whereas Asteraceae and Poaceae were the most commonly listed families. For species native to the United States, forbs and shrubs were listed most frequently, with Grossulariaceae (currants) and Asteraceae the most common. Species that are not yet in the United States were not common relative to introduced and native taxa but followed patterns of life form and family similar to those of natives (figure 1).

One major unanticipated finding was the frequency with which states list all species within an entire genus: 23 states list all species within one or more genera (supplemental table S1), with three states listing six or more (up to nine complete genera in Florida). In total, 41 genera are listed as Genus spp. (table S1), of these, five also occur on the federal list, resulting in a total of 890 individual species (or 773 individual species excluding the federally listed ones). Listing by genus constitutes 63% (890 of 1432) of the total number of species listed, or 62% (773 of 1249) when federally listed species are excluded.

In some cases, the motivation for listing at the genus level is clear, although this is not true for all 41 such listings (table 1). For example, there are three listed genera of parasitic plants, all of which also occur on the federal noxious weed list: Cuscuta, Orobanche, and Striga. This is undoubtedly to protect agricultural crops from these devastating plant pests (Ejeta 2007). Some parasitic plants can cause total crop failures (Linke 1992), forcing growers to abandon parasite-susceptible crops, especially in developing nations where control options are even more limited than in the United States (Parker 2013). There is a particularly alarming example of a cryptic Orobanche invasion into North and South Carolina in the late 1950s (Iverson et al. 2011). The invasion was identified early, quickly followed by a federal quarantine with an ongoing investment by the US government of more than $200 million aimed at eradication, saving untold billions in potential yield losses across the United States, although remnants of the invasion remain more than 60 years later.

Michigan lists all Ribes spp., a large genus of more than 150 species primarily native to the Northern Hemisphere that includes currants and gooseberries. This listing is also why the Grossulariaceae family is overrepresented in the number of listed noxious weeds that are native (figure 1c). The motivation behind listing such a large number of species, especially one that includes so many natives, is likely because Ribes are all alternate hosts for white pine blister rust (Cronartium ribicola), a devastating tree disease (table 1). Black and red currants (members of the Ribes genus) were long banned in the United States to protect the wood products industry (Maloy 1997). It would appear that Michigan took the conservative approach of listing all Ribes spp. as noxious weeds to protect their logging economy.

Overall, there were several aquatic species listed, including the entire genera Myriophyllum, Salvinia, and Trapa. Aquatic plants present a special threat to ecosystems because of their ability to spread easily between water bodies by hitching rides on boats; their rapid growth rates; the difficulty in managing them; the public's aversion to applying herbicides in water bodies; their threat to recreation, water quantity, and quality; trophic dynamics; and their large economic impacts (Lovell et al. 2006, Havel et al. 2015). In fact, aquatic weeds are viewed as such a large threat that the world's leading weed risk assessment gives the largest weight to potential weediness to simply being an aquatic plant (Pheloung et al. 1999). States are therefore well justified in limiting the entry and spread of aquatic weeds into and within their boundaries. More so, aquatic species are commonly transported via recreational vehicles moving among water bodies (Johnson et al. 2001). This is an excellent example of an opportunity for states to coordinate listing species, but also for education and enforcement.

There are several cases of entire genera being listed that encompass common crops, including Allium (e.g., garlic; Oklahoma, Indiana), Brassica (e.g., broccoli, canola, cauliflower, etc.; Oklahoma), Sorghum (e.g., sweet and grain sorghum; California and Utah), and Vitis (e.g., table and wine grapes; Michigan). In addition to being important crop genera, each includes many weedy and invasive plants: wild garlic, wild radish, Johnsongrass, and wild grapes, respectively. The Weed Science Society of America lists 15 species of weedy Brassica alone (WSSA 2019). Therefore, their motivation may be that states are intending to limit large swaths of weedy species and are simply listing entire weedy genera and simplifying identification training. However, an unintended consequence of this policy is that growing any crops in these genera is a violation of the respective state noxious weed laws, although specific exceptions to crops are listed by some states (table 1). We are not aware of any cases where growers have been limited in their ability to grow crops in states that fall within these listed genera, but it does point to seemingly avoidable and unnecessary legal quagmires that could be rectified by listing at the species level, or some other more direct policy. However, this may change with the legalization of hemp, whereas marijuana remains regulated—both are Cannabis sativa and are distinguished only by the amount of psychoactive THC content.

Expanded listing of native species

Excluding federally listed species, 48% (604) of listed species were introduced to the United States, 40% (493) are native to the United States, and a smaller subset (12%, 152) are not known to occur in the United States (figure 1c). The large number of listed native species was surprising, and ranged from 0 to 180 (Colorado), with 80% (40) of states listing at least one species native to the United States. In the 2000 analysis, there were two native species in the top 25 most listed weeds (Skinner et al. 2000), however, they did not evaluate the entire list for natives; we had no native species in the most listed species (table 2). The large number of native noxious weeds could result from the combination of several factors.

First, nativeness is notoriously difficult to define, having received far less attention than defining what is introduced (Pereyra 2020). We evaluated nativeness at the country scale for our analysis, which despite the obvious drawbacks to using this larger scale, countries are a standard unit for evaluating species nativeness (e.g., USDA 2019). Because the United States is so large and ecologically diverse, the number of native species in the United States is enormous. In fact, there are instances where a plant native to one part of the United States is considered weedy in another. For example, the tallgrass prairie native switchgrass (Panicum virgatum) was briefly listed as a noxious weed in California, where it is nonnative, to prevent large scale introduction as a bioenergy crop (Barney et al. 2009, Barney et al. 2012). Ironically, the weed presenting the greatest threat to US agriculture at the moment, Palmer amaranth (Amaranthus palmeri), is native to the Southwestern United States (Ward et al. 2013). In this case, the widespread use of glyphosate tolerant crops and repeated use of the nonselective herbicide both reduced overall weedy competition in agricultural fields and led to widespread and repeated development of herbicide-resistant Palmer amaranth biotypes (Ward et al. 2013). Therefore, newly formed biotypes quickly filled this newly opened niche and led to the greatest single weedy threat to agricultural productivity in generations. Therefore, it is likely that in some cases weeds we considered native to the United States do not occur broadly across the United States but could in fact be introduced to the listing states. However, we lacked the spatial resolution to test this but suggest this as an interesting follow-up analysis.

Second, as with some invasive species, many native species are responding positively to global change by becoming more abundant within their historic boundaries (i.e., infilling) or expanding their ranges (Bradley et al. 2015). In fact, the number of native range expanding species, which Essl and colleagues (2019) named neonatives, is projected to increase and become an important component of shifting community structure. Therefore, it is possible that states are listing native species they currently view as weedy or are being proactive and listing those they expect to be weedy in the future.

Third, a large number of listed natives may be attributed to the 23 states listing 35 complete genera for a total of 773 unique species (not including those already on the federal list), 54% of which are native to the United States. Nearly half (48%) of these genera contain at least one species native to the United States. Therefore, listing entire genera presents a multitude of issues (discussed above), including the listing of native species. This may account for 23 of the 40 states listing at least one native species, the other 17 are likely a combination of the factors described above.

Noxious weed lists are discordant with weed distributions

Given the dramatic spatial variation in the number of listed noxious weeds, we were interested in assessing whether states clustered geographically based on their weed list and whether this aligned with the weeds already present in the state. Given that plant distributions often span multiple states, we hypothesized that state weed lists would cluster on the basis of shared ecoregions and climates. Although the raw number of species varies, we predicted that adjacent states would share listed species more commonly than nonadjacent states. To test this, we conducted separate hierarchical clustering analyses on the matrix of noxious weed species by states listing them and on the matrix of noxious weed species (listed in any state) by states where they are present. The clustering criterion applied was Ward's minimum variance method, which is an agglomerative approach. This means that each state noxious weed list began as its own cluster and was iteratively grouped with other lists by similarity. This method pools within-cluster distances’ sums of squares and is considered to be conservative, meaning that it does not overgroup or create clusters with excessive noise (Hossen et al. 2015). The within-cluster variance was optimized at six clusters in each of the two analyses. In addition to assigning clusters to state boundaries in ArcMap (Environmental Systems Research Institute, Release 10.7.1, Redlands, California), we produced constellation plots of the node interrelationships within and among clusters. Hierarchical clustering was conducted in JMP Pro 14 (SAS Institute Inc., Cary, North Carolina).

Our clustering analysis showed little support for our hypothesis that nearby states would share listed species. Most of the United States, including Hawaii and Alaska, clustered together as being similar in their regulatory noxious weed lists (figure 2a, 2b). A few distinct clusters reflected unique climate or geography (e.g., California) or reflected a list of many complete genera (e.g., Oklahoma). Contrary to our expectations of geographic clustering, adjacent states were no more likely to share listed weeds than nonadjacent states.

Figure 2.

Comparison of the (a) hierarchical clustering and (b) constellation plot of state noxious weed lists to the (c) hierarchical clustering and (d) constellation plot of noxious weed presence. Listed entire genera have been expanded to include all named species and any state listed species that overlaps with the federal list has been removed to allow for state-to-state comparisons. In panels (a) and (c), the colors represent species lists clustered by similarity. Greater branching in constellation plot (d) indicates zones of strong local similarity, compared to their absence in panel (b), where most items branch individually from a single spine.

This differs dramatically from the clustering pattern of the geographic distribution of weedy taxa (figure 2c, 2d). The presence of noxious weeds has a clear spatial pattern that primarily reflects the bioclimatic patterns of the United States (figure 2c): tropical (Florida, Hawaii), Mediterranean (California), etc. There are clear incongruities between which weeds are present in a state and those listed (figure 2a, 2c). However, there are some caveats to our analysis. We only include weeds that are listed as noxious by at least one US state. Although this list included 1249 unique weedy species, there are undoubtedly more weeds throughout the United States that are weedy. For example, the Weed Science Society of America maintains a composite list of weeds that currently has 2847 species listed (WSSA 2019). Therefore, our analysis may overlook a large subset of weedy plants that are not listed, although this would likely only inflate the incongruity between present and listed species. The inclusion of complete genera by many states likely skewed the noxious weed clustering analysis (figure 2a, 2b), because many species were included in these states without regard to their actual spatial distribution or weediness (figure 2c, 2d). Nevertheless, the discordance between current patterns of listing noxious weeds and the current distribution of these weeds, as well as expected geographic clustering, suggests that there is room for improvement and coordination among states.

Are states being proactive or reactive in weed listing?

Alhough states appear to not be coordinated in their listing of noxious weeds, perhaps because of the varied typologies and regulatory motivations of states (McCubbins et al. 2013), we wanted to determine whether state noxious weed lists were being reactive (i.e., species are listed after they are known weeds in that state) or proactive (i.e., listing known weeds before they are present in the state). We evaluated the number of known noxious weeds occurring somewhere in the United States against each state's weed list (figure 3a), which varied between 70% and 99%. In other words, states are only listing 1%–30% of total known noxious weeds present in their state. For most states, there are hundreds of weeds present in the state that are not regulated (figure 3a). This is not surprising, because noxious weed lists are not designed to regulate all weeds, only those deemed to threaten agriculture and the environment. More specifically, most noxious weed regulations are focused on preventing new introductions or limiting existing infestations, especially of agricultural weeds (Quinn et al. 2013). Many weeds are legacies of long ago introductions (Rouget et al. 2016), many of which are well beyond feasible management (Panetta 2009). Therefore, states may be ignoring these widespread weeds, focusing more on species not yet in the state or weeds of limited distribution.

Figure 3.

(a) The total number of noxious weeds occurring in each state that are not listed by that state with shading representing the relative proportion. (b) The number of species listed in each state that are not yet present in that state, but are present in at least one other US state. (c) The number of species listed in each state that are not yet in the United States. When an entire genus was listed, it has been expanded to include all named species, and any state-listed species that overlaps with the federal list has been removed to allow for state-to-state comparisons.

Many states (78%) list species that are known weeds somewhere in the United States but are not yet in that state (figure 3b). This ranges from 1 to 259 species, or 1%–80% of their list. This is somewhat skewed by Oklahoma, which includes several large genera, but overall, many states appear to be proactive by including weeds not yet in their state. Moreover, 74% of states list at least one species that is not yet in the United States (figure 3c), although this is typically only a handful of species. Regulating species before they are introduced is the most pragmatic option for mitigating their impacts. Overall, it appears that few species are regulated before they enter the United States, a much larger number are regulated before they enter the state, but most state lists include species already present. It is possible that some weeds were listed before they were present in the state prior to our analysis, but we lack temporal weed distribution data to adequately test this. Our analysis was unable to determine the motivation for listing species, although this would make an interesting follow up study.

Are states poised to address the coming weedier future?

Across all states, noxious weed lists are restricted, including only a small minority of existing weeds, and are largely reactive, not listing weeds that are known to occur in the United States but not yet in that state. In contrast, there are 152 (12%; figure 1c) not yet known to occur in the United States but listed broadly across states, which could be viewed as some level of prevention. However, this weed prevention is meager at best and leaves much room for improvement. These limitations, in combination with the large number of complete genera listed and the startling number of native species and crops listed raise serious concerns about the motivations, approach, and effectiveness of US state noxious weed listings.

The number of new introductions continues to rise with increasing global trade and wealth and shows no sign of slowing (Seebens et al. 2017). It has been shown time and again that prevention is our most effective tool in mitigating the weed threat and that implementing preventative policies such as weed risk assessments provides a substantial economic benefit (e.g., Keller et al. 2007). Noxious weed lists, in combination with border screening for plants not here yet, and the application of weed risk assessments must be robust, transparent, and nimble to be useful. McCubbins and colleagues’ (2013) analysis showed that the structure and application of noxious weed lists are quite variable, and Quinn and colleagues (2013) demonstrated that these lists underrepresent the invasive flora. Although the extent of lists has improved in the last two decades, many relatively simple improvements can be made. Below, we outline some priorities to be considered for improvement.

States must be coordinated in their listing as weeds obey no political borders and can easily spread to nearby states. One such example demonstrating the coordination among states is that of wavyleaf basketgrass (Oplismenus undulatifolius). Basketgrass was introduced to eastern Maryland in the 1990s and quickly spread locally around the site of introduction, leading to sporadic small populations in Virginia (Beauchamp and Koontz 2013). Virginia worked with Maryland officials to develop a monitoring program and listing basketgrass in Virginia as a noxious weed. It is unclear how common this level of coordination is among states, but we urge such cooperation and information sharing, especially in the context of a nationally coordinated early detection and rapid response network (e.g., Reaser et al. 2020), such as the National Plant Diagnostic Network used for early detection of plant pathogens (www.npdn.org/home). Coordination among states would likely be best achieved through federal level oversight, with the National Invasive Species Council being an obvious candidate.

Native species can no longer be ignored, because they pose increasingly complex challenges to invasion science, management, and policy (Davis et al. 2011, Carey et al. 2012, Nackley et al. 2017). There are nearly 500 species native to the United States regulated as noxious weeds somewhere in the United States—40% of all listed species. This was unexpected and warrants an urgent conversation for the implications and relevance of nativeness as a concept. There may not be a one-size-fits-all definition of nativeness that is applicable to all jurisdictions because of the diversity of noxious weed structures (McCubbins et al. 2013) and the multiplicity of resources they are designed to protect. However, we do believe there are steps that states can take to avoid ecological and regulatory uncertainty. An important step would be explicit documentation of the historic native range, as well as documentation of instances of invasiveness within that historic range. Both would aid in policy and management of the challenging issue of weedy native species (Nackley et al. 2017).

First, states should be intentional and clear in their listing of taxa. For example, we see few instances when listing entire genera is justified (e.g., genera of parasites of crop plants). Although listing entire genera has some advantages (e.g., ease of identification), the drawbacks are numerous and likely outweigh the benefits: regulatory entanglement (e.g., native species, crops, etc.) and management challenges (i.e., science-based screening). Therefore, such listings should be done with trepidation. Each proposed taxon should be considered individually. Second, the regulations should clearly articulate how the nativeness of a species is handled in a way that is best suited to the needs and circumstances of each state. The regulation should also clearly outline how subspecific taxa are covered under the regulations (i.e., cultivars, varieties, etc.) such as those seen in table 1. Third, the listing process must include a transparent and science-based system (e.g., a weed risk assessment sensu Koop et al. 2012) for evaluation that provides legitimacy and limits bias in the listing process but is flexible enough to allow the proactive listing of species not yet present in the state. Finally, states must coordinate and share information on weedy taxa, especially during the early stages of invasion. This may be best done with federal oversight and support while being implemented at the state level. This would also advance the vision of the National Invasive Species Management Plan. While not ignoring all of their existing weedy flora, states would also be well served to focus on known weedy taxa that are not yet in their state but are in nearby states or those that are present at relatively small population sizes. In addition, an analysis of the motivations, implementation, and effectiveness of these regulations would be an excellent next step in our understanding of whether they are serving their intended purpose. The growing weed threat is an inevitable consequence of the Anthropocene, but robust, clearly defined, and coordinated noxious weed lists can be a powerful first line of defense in protecting natural and agricultural resources.

Acknowledgments

The authors appreciate the comments from the reviewers and editors that improved the manuscript. The data are available in Lakoba V, Brooks R, Haak D, Barney J. State Noxious Weed Lists, Species, and Occurrences. University Libraries, Virginia Tech. https://doi.org/10.7294/JV9X-7656.

Author Biographical

Vasiliy T. Lakoba, Rachel K. Brooks, David C. Haak, and Jacob N. Barney ([email protected]) are affiliated with the School of Plant and Environmental Sciences at Virginia Tech, in Blacksburg, Virginia. VTL and RKB contributed equally to this work.