key: cord-0317665-rbdtodxz
authors: Huron, Nicholas A.; Behm, Jocelyn E.; Helmus, Matthew R.
title: Paninvasion severity assessment of a U.S. grape pest to disrupt the global wine market
date: 2022-05-11
journal: bioRxiv
DOI: 10.1101/2021.07.19.452723
sha: 9e1c8103251f556faa9455351c8b95db920c06b9
doc_id: 317665
cord_uid: rbdtodxz

Economic impacts from plant pests are often felt at the regional scale, yet some impacts expand to the global scale through the alignment of a pest’s invasion potentials. Such globally invasive species (i.e., paninvasives) are like the human pathogens that cause pandemics. Like pandemics, assessing paninvasion risk for an emerging regional pest is key for stakeholders to take early actions that avoid market disruption. Here, we develop the paninvasion severity assessment framework and use it to assess a rapidly spreading regional U.S. grape pest, the spotted lanternfly planthopper (Lycorma delicatula; SLF), to spread and disrupt the global wine market. We found that SLF invasion potentials are aligned globally because important viticultural regions with suitable environments for SLF establishment also heavily trade with invaded U.S. states. If the U.S. acts as an invasive bridgehead, Italy, France, Spain, and other important wine exporters are likely to experience the next SLF introductions. Risk to the global wine market is high unless stakeholders work to reduce SLF invasion potentials in the U.S. and globally.

Invasive plant pests cause substantial economic impacts 1 , but most pests and their 27 impacts are confined to specific regions. For a regional pest to become a globally invasive 28 species that disrupts global markets (i.e., paninvasive), ecological and economic factors that 29 determine the pest's transport, establishment, and impact potentials must be aligned at the global 30 scale (Fig. 1, Supplementary Methods) 2 . First, paninvasive pests have high transport potential 31 because they can be easily transported among regions, often through global trade 3 . Second, 32 paninvasive pests have high establishment potential, because their environmental needs for 33 population growth are met in many regions 4 . Third, paninvasive pests have high impact potential, 34 because invaded regions have sizeable agricultural production and industries vulnerable to the 35 pest 5 . If these invasion potentials are correlated across multiple regions globally for an emerging 36 regional pest, there is a high risk of the pest spreading to cause supply crashes in regional 37 markets that cascade to disrupt global markets 6 . 38

Despite the importance of identifying emerging paninvasives, existing approaches lack a 39 cohesive and universal framework for rapidly assessing and effectively communicating such risk 40 to stakeholders 7 . To address this gap, we developed the paninvasion severity assessment 41 framework by adapting the U.S. CDC pandemic severity assessment framework to invasion 42 process theory, which describes translocations of species in terms of transport, establishment, 43 and impact potentials (Fig. 1, 2) 2,8-10 . Although invasive species frameworks are increasingly 44 adapted to understand infectious diseases like COVID-19 11-16 , adapting public-health 45 frameworks to invasion science is novel and leverages an increasingly universal risk vocabulary 46 ( Fig. 2) 17 . Under this framework, we assessed the paninvasion risk of the spotted lanternfly 47 6 grape production (ρ = 0.67, P<0.001, Fig. 5b) , and country wine production (ρ = 0.63, P<0.001). 119

This alignment of potentials is clear in the invasion-potential alignment plots (Fig. 5) . Major 120 grape producing regions fall in the upper-right quadrant of the plots where regions have both 121 high transport and high establishment potentials. 122 123 Paninvasion risk 124 We estimated the risk of SLF to disrupt the global wine market to be an 8 out of 10 ( Fig.  125   6 ). To derive this value, we regressed country grape production on country transport and 126 establishment potentials. Each predicted value from this multivariate regression can be 127 considered an estimate of the risk of SLF to invade and impact a country's grape production. We 128 then rescaled these predicted values from 1-10 and correlated them to wine export market size (ρ 129 = 0.66, P<0.001). To place SLF on a scale of paninvasion severity, we rescaled the correlation 130 coefficient, ρ, from 1-10, so that 1 is a complete negative correlation and 10 is a complete 131 positive correlation between predicted impact and market size. Low values on this scale indicate 132 that the global market is buffered against a paninvasion, while high values indicate that a 133 paninvasion is likely unless mitigation actions are taken to reduce invasion potentials. 134 135

The risk of a spotted lanternfly (SLF) paninvasion is high and coordinated effort should 137 be made to reduce its transport, establishment, and impact potentials globally. In the U.S., efforts 138 to reduce SLF transport potential are primarily through quarantine and inspection of goods. SLF 139 is a regulated plant pest and the U.S. Department of Agriculture (USDA) is working towards 140 implementing consistent, science-based, and nation-wide transport protocols 21,49,50 . We 141 recommend that estimates of SLF transport potential be updated regularly as more states become 142 invaded. Finer spatial and temporal scale data are needed on high transport potential pathways 143 such as rail, landscaping stone, and live tree shipments to better forecast long-distance 144 Reduction to SLF impact potential currently relies on reducing populations with tree-164 band trapping and broad-spectrum insecticides (e.g., carbamates, organophosphates, pyrethroids, 165 neonicotinoids) that have high nontarget mortality 18,23,56,57 . However, existing management 166 practices do not prevent vineyard reinfestations and pesticide application often overlaps with 167 grape harvest when adults move into vineyards 36 . Damaged vines can be pruned but grape yield 168 is reduced 58 . More research on long-term control methods for established populations is critical 169 for reducing SLF impact potential. First, trapping technologies that reduce bycatch must be 170 refined and widely deployed 57,59 . Second, biocontrol agents that specialize on SLF, such as 171 parasitoid wasps and fungus show promise, but more work is needed to understand non-target 172 attack rates 60-62 . Finally, like targeted mRNA vaccines developed to reduce the impact of SARS-173

CoV-2 on human health 63 , SLF-specific RNAi insecticides have the ability to control outbreaks 174 in vineyards and beyond 21,64 . Although SLF impacts to vineyards within its invaded range are 175 significant, to date, SLF has yet to invade a major viticultural area. Its actual impact on such 176 regions with larger, wealthier, and interconnected wine economies is thus unknown. It is also 177 unclear whether market elasticity might weaken or strengthen the disruption of a SLF 178 8 paninvasion to the global wine market. It behooves governments to heed the paninvasion risk of 179 this grapevine pest. 180

When a pest like SLF with high paninvasion risk emerges, coordinated governmental 181 efforts can mitigate global market disruptions. For example, the Great Wine Blight of the late 182 19 th century caused by grapevine phylloxera (Hemiptera: Daktulosphaira vitifoliae) was the 183 largest shock to the global wine market ever recorded 65 . Phylloxera decimated European 184 vineyards, but the market recovered due to pest management solutions whose development were 185 coordinated by high-level officials in the French federal government 65 . 186 For SLF in the U.S., early federal coordination was hampered. The U.S. National 187

Invasive Species Council (NISC) contains high-level federal officials (e.g., Secretaries of State, 188

Interior, Agriculture, Defense) who coordinate reduction of invasive species impact 66 . In 2019, 189 NISC funding was cut and the Invasive Species Advisory Committee (ISAC) was dissolved. The 190 ISAC comprised scientific experts who advised the NISC by producing memoranda and white 191 papers on emerging invasive species issues 67 . These actions decreased U.S. capacity to respond 192 to emerging paninvasive species 68 . In 2021, the ISAC was reinstated, and NISC's funding is 193 expected to be fully restored 69 . Based on our SLF risk assessment, we suggest the reinstated 194 ISAC produce memoranda and white papers on solutions for SLF as soon as possible. SLF risk 195 must be communicated to the NISC, who can mobilize the resources needed to reduce its impact. 196

The paninvasion severity assessment framework is a stakeholder communication tool to 197 assess if an invasive can cause market, environmental, and human-health disruptions at the 198 global scale (Fig. 1 ). It does so by equating pathogen transmission, infectivity, and virulence-199 well-known to the public due to COVID-19 pandemic-with invasive species transport, 200 establishment, and impact potentials (Fig. 2) . Paninvasion assessments produce accessible maps 201 (Fig. 3, 4) , scatter plots (Fig. 5) , and easy to understand risk values (Fig. 6 ). Going forward, 202 invasion potentials for other species are likely to increasingly align and coordinated 203 governmental efforts will be needed to reduce such potentials in the U.S. and internationally. It is 204 prudent that when an invasive species is found, the severity of its paninvasion risk be assessed. 205

Methods 207

Here, we provide a methodological discussion of the paninvasion severity assessment 208 framework (Fig. 2) and its application to spotted lanternfly (SLF, Fig. 1 ). To make the SLF 209 9 assessment easy to refine once new data and insights are available, we provide both an open-210 source R package that includes all data to reproduce all results (https://ieco-lab.github.io/slfrsk/) 211 and a Google Earth Engine application to map SLF paninvasion severity from global to local 212 scales (https://ieco.users.earthengine.app/view/ieco-slf-riskmap). These open-science tools also 213 are adaptable to other emerging regional invasives at risk of paninvasion e.g., 70 . Here, we focused 214 on agricultural and economic data most relevant to assess pest risk, but for non-pest invasives, 215 data on environmental and human health may be a higher priority. 216 217

Although the invasion process can be divided into many stages, the paninvasion severity 219 assessment framework focuses on the three main stages most often estimated in invasion risk 220 assessments 2 and that are analogous to the disease potentials that public-health scientists quantify 221 for pathogens ( Fig. 2 ) 8 . When a pathogen with pandemic risk emerges, public health scientists 222 place it within scaled measures of transmissibility and infectivity (often combined and termed 223 transmissibility), and virulence (clinical severity) to assess its risk 8,9 . For example, when SARS-224

CoV-2 emerged during the COVID-19 pandemic, the initial understanding was that different age 225 groups had similar potentials to transmit and become infected (Fig. 2a, y-axis) , but different age 226 groups varied in their clinical severity once infected (Fig. 2a, x-axis) 8,71,72 . To adapt this public-227 health framework to invasion process theory 2,73-76 , we equated transmission, infectivity, and 228 virulence potentials of a pathogen across different human populations to the transport, 229 establishment, and impact potentials of an invasive species across different regions (see colored 230 arrows between Fig. 2a and b) . For example, in Fig. 2b we placed several hypothetical regions 231 that together indicate strong alignment (i.e., multivariate correlation) among invasion potentials 232 across the regions. In this example, predicted invasion risk (Fig. 2c, x-axis) for these three 233 hypothetical regions is strongly correlated to a measure of their contributions to a global market 234 (Fig. 2c, y-axis) , indicating an overall high paninvasion risk. 235

Paninvasion assessments comprise four steps (Fig. 2d) : 1) estimate invasion potentials, 2) 236 calculate alignment of invasion potentials, 3) quantify paninvasion risk, and 4) articulate caveats, 237 which we describe in detail for SLF below and in the Supplementary Methods. 238

Step 1: Estimate Invasion Potentials 240

Transport potential is a measure of propagule pressure 77 . The prevailing hypothesis on 242 SLF transport potential is that regions that import more tonnage of commodities from the 243 invaded U.S. region also import more total tonnage of goods and trade infrastructure (e.g., cargo 244 containers, pallets, and railcars) that inadvertently transport SLF propagules, such as egg masses, 245 long-distances 18,21,26,34,78,79 . To estimate which states were invaded and identify SLF 246 transportation events, we obtained a database of SLF records from the USDA and aggregated 247 first-find and regulatory incident reports e.g., 38 . As of December 2020, the invaded states were 248

Connecticut, Delaware, Maryland, New Jersey, New York, Ohio, Pennsylvania, Virginia, and 249

West Virginia (Fig. 3) . We estimated transport potentials from the U.S. invaded region as the 250 obtained from GBIF on October 20, 2020. To find the best-fit models that explained TOH and 270 SLF presences, we identified a subset of six covariates, from 22 candidate covariates [83] [84] [85] [86] , that 271 minimized model collinearity: annual mean temperature, mean diurnal temperature range, annual 272 precipitation, precipitation seasonality, elevation, and access to cities. We fit sdm_toh and 273 sdm_slf1 with these six covariates. sdm_toh represents our best estimate of the global distribution 274 of TOH, thus we fit sdm_slf2 that modeled SLF suitability from the predicted values of sdm_toh. 275

As such, sdm_slf2 represents suitability that considers a primary plant host (TOH) 25 that is also 276 invasive but likely not at equilibrium 87 Step 2: Calculate Alignment Correlations

To consider how all three invasion potentials may coincide for regions, we calculated 303 alignment correlations for states and countries separately. Alignment was calculated for each of 304 the two measures of impact potential as Spearman rank correlations between impact potential 305 and the predicted values from linear regressions models of each impact potential regressed on 306 transport and establishment potentials together 94 . We then visualized these multiple multivariate 307 correlations as quadrant plots following a stakeholder-friendly and approachable format adapted 308 from the pandemic severity assessment framework 8,9 . 309 310

Step 3: Quantify Paninvasion Risk 311

To determine if invasion risk for countries corresponds with economic impact to the 312 global wine industry, we investigated the relationship between wine market size and predicted 313 risk of invasion for individual countries. We estimated wine market size for 223 countries 314 (including some that export but do not produce wine) as the value of wine exports corresponding 315 with the years for our trade data (2012-2017, log10 USD) downloaded from the FAOSTAT 316 detailed trade matrix 46 , accessed August 31, 2020. Then, we regressed country grape production 317 on transport and establishment potentials with multiple linear regression. Each predicted value 318 from this regression can be considered an estimate of the risk of SLF to invade and impact a 319 country's grape production. We rescaled these estimates from 1-10 to create an easily interpreted 320 estimate of risk and then correlated these predicted values to wine export market size. To place 321 overall SLF paninvasion severity on a clear scale for both researchers and stakeholders, we 322 simply rescaled the Pearson correlation from 1-10, so that 1 is a complete negative correlation 323 and 10 is a complete positive correlation between country risk and wine export market size. 324

Step 4: Articulate Caveats 326

Paninvasion risk assessments should be performed iteratively as the invasion process 327 continues across regions and responses are mobilized. Early assessments of emergent pests have 328 great utility to support early responses but often come with caveats. The fourth step of the 329 framework is to articulate the caveats of a current assessment to guide future research. These 330 caveats should explicitly consider the assumptions made when estimating invasion potentials and 331 how those potentials and risk could change as the invasion process continues. Below we 332 13 articulate caveats of our SLF assessment to provide a basis for future refined assessments of SLF 333 to disrupt the global wine market. 334 We calculate paninvasion risk of SLF via stepping-stone transport from the eastern US. 335

However, major wine producing nations also heavily trade with China, Japan, and South Korea, 336

where SLF is also established. Total SLF transport potential is thus greater than our estimates, 337 meaning paninvasion risk is higher than what we report here. establishment potential for SLF we report here. Early assessments of paninvasion severity are 357 unlikely to account for plasticity and adaptation that is common for invasive pests, especially 358 when combined with variation in local weather patterns and climate change. Indeed, a recent 359 analysis suggests that SLF will experience increased suitable habitat and a greater impact in 360

China in the future due to climate change 97 . The expected effect of climate change is likely more 361 complicated for SLF, which has a flexible life cycle that can include but does not require, 362 temperature-linked diapause for overwintering in cooler regions 24,98,99 . Survivorship appears 363 greater without such diapause 98 , and thus establishment potential may be even higher than 364 expected in warmer climes. 365

Variation in weather, climate change, host preference, and pest density can influence pest 366 impacts. Such factors often act at different scales and in a spatially heterogenous manner. For 367 example, SLF prefers grapes, but the degree to which it does over alternative hosts near 368 vineyards remains poorly known, which is important, since SLF appear to have their highest 369 densities at vineyard edge habitats 100 . The vulnerability of viticultural regions may be affected by 370 the prevalence of particular grape cultivars or alternative hosts, but additional research is 371 necessary to elucidate SLF feeding presence. Similarly, the relationship of SLF density within a 372 vineyard and density in the surrounding landscape remains poorly known, which is in turn likely 373 to be influenced by weather patterns and plant phenology 36 . As SLF host preference and its 374 relationship to landscape variables become better understood, they should be incorporated into 375 considerations of impact potential. 376

Lastly, to refine the SLF paninvasion risk assessment, future work should calculate 377 invasion potentials for other grape pests like phylloxera to place SLF on an absolute scale of risk 378 severity 101 . Our assessment of SLF relativizes invasion potentials with the assumption that 379 regions with high potentials relative to other regions also have high absolute potentials. SLF has 380 broad environmental suitability, a flexible life cycle, ability to lay many discrete egg masses on 381 numerous substrates (Fig. 1b) , observed rapid spread (Fig. 3) and realized impacts to grape and 382 wine production 21-26,36 , so its absolute potentials are likely very high. However, absolute 383 potentials can only be assessed by comparing multiple paninvasive species, like what is done for 384 pathogens. When a pathogen with pandemic potential emerges, the pandemic severity 385 assessment framework compares the severity of the potentials of the current outbreak pathogen 386 to past pandemic producing pathogens 8,9,71 . The next step towards a mature paninvasion 387 framework is to estimate invasion potentials for current paninvasive species, so that the 388 likelihood of a paninvasion for any emerging regional pest can be placed on an absolute scale of 389 severity. 

The great wine blight

Invasive Species Advisory Committee Products

U.S. action lowers barriers to invasive species

Executive Order on Continuance or 559 Reestablishment of Certain Federal Advisory Committees and Amendments to Other 560

Assessing the ecological niche and 562 invasion potential of the Asian giant hornet

Assessing the severity of COVID-19

Prevent Epidemics. COVID-19 Key COVID-19 metrics based on the latest available science

Invasion biology and biological control

Biological invasion theory: Darwin's contributions from The 572 Origin of Species

Mapping forest canopy height globally 597 with spaceborne lidar

Review on Invasive Tree of Heaven (Ailanthus 599 altissima (Mill.) Swingle) Conflicting Values: Assessment of Its Ecosystem Services

A review of methods for the assessment of prediction errors in 602 conservation presence/absence models

Predicting species 604 distributions from small numbers of occurrence records: A test case using cryptic geckos

Species-specific tuning increases robustness to sampling bias 607 in models of species distributions: An implementation with Maxent

Density of alcoholic beverage, wine, table, all (food)

Alcohol and Tobacco Tax and Trade Bureau

Multiple regression: A primer

Biological invasion risk assessment of Tuta absoluta: Mechanistic versus 617 correlative methods

Forecasting species range dynamics with process-explicit models: 619 Matching methods to applications

Risk assessment of insect pest expansion in alpine ecosystems under 621 climate change

Comparison of the hatch of newly laid Lycorma delicatula 623 (Hemiptera: Fulgoridae) eggs from the United States after exposure to different temperatures 624 and durations of low temperature

Characterizing the spatial distributions of spotted lanternfly 628 (Hemiptera: Fulgoridae) in Pennsylvania vineyards

Biology and management of grape 630 phylloxera

This work was funded by the United States Department of Agriculture Animal 637 and Plant Health Inspection Service Plant Protection and Quarantine under Cooperative 638

Agreements AP19PPQS&T00C251 and AP20PPQS&T00C136; the United States Department of 639

Agriculture National Institute of Food and Agriculture Specialty Crop Research Initiative 640

Coordinated Agricultural Project Award 2019-51181-30014; and the Pennsylvania Department 641 of Agriculture under agreements 44176768

a pathogen across different populations or age groups 8, 9, 11 . For paninvasions (b), invasion 674 potentials for an emerging regional invasive species are estimated (d Step 1) by equating 675 pathogen transmission with transport potential, infectivity with establishment potential, and 676 virulence with impact potential (follow the arrows) across regions (black circles) to construct 677 quadrant plots that depict their alignment based on multivariate correlations (d Step 2; see 678 Methods). Next, paninvasion risk (c) is estimated from the correlation between regional invasion 679 risk estimated from the multivariate regression of invasion potentials (d Step 3; see Methods) 680 and the size of regional markets that could be disrupted. The steps of the paninvasive severity 681 assessment framework (d), culminating by articulating caveats in the current assessment that 682 direct future research (d Step 4) that provide data to inform the next assessment iteration.