Molecular Survey for the Invasive Leafminer Pest Liriomyza Huidobrensis (Diptera: Agromyzidae) in California Uncovers Only the Native Pest Liriomyza Langei

Molecular Survey for the Invasive Leafminer Pest Liriomyza huidobrensis (Diptera: Agromyzidae) in California Uncovers Only the Native Pest Liriomyza langei

Scheffer, S. J., Lewis, M. L., Gaimari, S. D., & Reitz, S. R. (2014). Molecular Survey for the Invasive Leafminer Pest Liriomyza huidobrensis (Diptera: Agromyzidae) in California Uncovers Only the Native Pest Liriomyza langei. Journal of Economic Entomology, 107(5), 1959-1964. doi:10.1603/EC13279

10.1603/EC13279 Entomological Society of America

Version of Record http://cdss.library.oregonstate.edu/sa-termsofuse MOLECULAR ENTOMOLOGY Molecular Survey for the Invasive Leafminer Pest Liriomyza huidobrensis (Diptera: Agromyzidae) in California Uncovers Only the Native Pest Liriomyza langei

1,2 1 3 4 SONJA J. SCHEFFER, MATTHEW L. LEWIS, STEPHEN D. GAIMARI, AND STUART R. REITZ

J. Econ. Entomol. 107(5): 1959Ð1964 (2014); DOI: http://dx.doi.org/10.1603/EC13279 ABSTRACT Liriomyza huidobrensis (Blanchard) is a highly destructive invasive leafminer pest currently causing extensive damage to vegetable and horticultural crops around the world. Liriomyza langei Frick is a leafminer pest native to California that cannot currently be morphologically distinguished from L. huidobrensis. We used a DNA-barcoding approach, a published PCR-RFLP method, and a new multiplex PCR method to analyze 664 ßies matching the morphological description of huidobrensisÐlangei. We found no evidence for the presence of L. huidobrensis in our extensive samples from California. In addition to the new molecular method, this work is important because it provides deÞnitive data that the California “pea leafminer” is currently, and has probably always been, L. langei. These data will also be important in the event that the highly invasive L. huidobrensis ever becomes established.

KEY WORDS molecular diagnostics, multiplex PCR, invasive species, crop pest

The polyphagous leafmining ßy Liriomyza huidobrensis sequently corroborated by differences in two inde- (Blanchard) has become a notorious pest of a wide pendent nuclear genes, and the name L. langei Frick variety of vegetable and ßower crops that number in the was revived for the California (plus Hawaii) popula- hundreds (Spencer 1973, Reitz and Trumble 2002). tions (Scheffer and Lewis 2001). Invasive populations Originally described from Argentina (Blanchard around the world sampled to date have invariably carried 1926), since the 1980s, L. huidobrensis has become mitochondrial haplotypes belonging to the South Amer- invasive and has spread to Europe, the Middle East, ican L. huidobrensis (e.g., those from Canada, China, Asia, and North America (Canada; van der Linden Indonesia, Italy, Israel, Japan, Korea, the Philippines, 1990; Weintraub and Horowitz 1995; Head et al. 2000; South Africa, Sri Lanka, and Taiwan; Scheffer 2000; Scheffer 2000; Scheffer et al. 2001, 2006; He et al. Scheffer et al. 2001, 2006; He et al. 2002; Takano et al. 2005 2002). Newly established populations are often dev- as reported in Takano et al. 2008). Most recently, Takano astating, and it is currently considered one of the et al. (2008) found clear evidence of partial reproductive worldÕs most destructive agromyzid pests. isolation between two colonies of ßies molecularly iden- In 1951, Frick described a leafminer from peas in tiÞed as L. langei and L. huidobrensis. Currently, there is California as Liriomyza langei Frick (Frick 1951, Parrella no evidence of these two species occurring in sympatry 1982). Upon further examination, this species appeared in any part of the world. to be morphologically indistinguishable from L. huido- California is one of the primary vegetable-growing brensis and was synonymized with L. huidobrensis by regions within the United States. Although L. langei Spencer (1973). Scheffer (2000) found that South has been known in California as “the pea leafminer” American and Californian L. huidobrensis s.l. differed (at times as L. huidobrensis) and reported as a substantially in mitochondrial cytochrome oxidase se- polyphagous pest since at least the 1930s (Chaney quence data. These molecular differences were sub- 1995), it was not until the 1990s that its pest status in California changed from minor secondary pest to an The use of trade, Þrm, or corporation names in this publication is economically important crop pest (Chaney 1995, Re- for the information and convenience of the reader. Such use does not itz and Trumble 2002). It has been suggested that this constitute an ofÞcial endorsement or approval by the United States change has come about due to the presence of a Department of Agriculture or the Agricultural Research Service of genetically distinct and more pestiferous population in any product or service to the exclusion of others that may be suitable. 1 Systematic Entomology Laboratory, USDAÐARS, Bldg. 005, Rm., the central coastal valleys that is spreading southward 137, 10300 Baltimore Ave., Beltsville, MD 20705. (Morgan et al. 2000, Reitz and Trumble 2002). Given 2 Corresponding author, e-mail: [email protected]. the invasive nature of South American L. huidobrensis 3 Plant Pest Diagnostics Branch, California Department of Food and and the fact that currently it cannot be distinguished Agriculture, 3294 Meadowview Rd., Sacramento, CA 95832. 4 Malher County Extension, Department of Crop and Soil Science, morphologically from L. langei, the question of Oregon State University, 710 SW 5th Ave., Ontario, OR 97914. whether invasive L. huidobrensis is present but un- 1960 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 107, no. 5 documented in California has never been adequately nogram. DNA sequences have been deposited in addressed. The purpose of this study was to conduct GenBank with accession numbers KJ778702ÐKJ778751. a molecular survey to determine whether L. huidob- A subset of voucher specimens for L. langei from this rensis is present but masquerading as L. langei in veg- study have been deposited in the National Museum of etable-growing regions of California. To accomplish Natural History in Washington, DC. this, we used molecular methods to screen extensive Females and PCR Screening of Pooled Samples. samples of adult ßies collected from several regions To screen a large number of individuals, abdomens and crops known to harbor L. langei. We paid partic- from female specimens were pooled in groups of up to ular attention to the populations in the central coastal Þve abdomens before DNA extraction. A maximum of valleys, where the change in pest status of presumed four pooled samples (20 ßies) were screened from any L. langei was Þrst observed (Chaney 1995, Reitz and particular Þeld collection (location by date by crop). Trumble 2002). The head and thorax of each ßy was retained as voucher material to be investigated further if anomalous results, evidence of L. huidobrensis, or both, Materials and Methods were discovered during the screening. Female abdo- Members of the California Department of Food and mens within each pooled sample were ground to- Agriculture collected leafmining ßies by sweep-net- gether for extraction using a micropestle and sub- ting various crop hosts during the period 3Ð24 October jected to the DNeasy tissue extraction procedure. 2006 in 11 counties of California: Fresno, Imperial, Two independent PCR-based screening procedures Monterey, Riverside, San Benito, San Diego, San Luis were used to test pooled female extractions for the Obispo, San Mateo, Santa Barbara, Santa Cruz, and presence of L. huidobrensis DNA. The Þrst method is Ventura. Specimens were preserved in ethanol and a published PCR-RFLP procedure that uses a single stored at Ϫ80ЊC. Because multiple leafminer species pair of primers to amplify a 1,031-bp fragment of COI were present, samples were sorted by S.J.S. to include and COII in both L. huidobrensis and L. langei (Schef- only ßies matching the external morphological de- fer et al. 2001). Restriction digests of the ampliÞcation scription of L. langei and L. huidobrensis. fragment using EcoRV and SpeI in separate reactions Two molecular methods were used to screen ßies result in diagnostic bands of differing lengths for L. for evidence of L. huidobrensis. First, total genomic huidobrensis and L. langei that can be visualized on an nucleic acids were extracted from the abdomens of agarose gel (for full details see Scheffer et al. 2001). individual male ßies, and mitochondrial cytochrome We tested positive controls (several independent oxidase I was ampliÞed and sequenced. In the second DNA extractions of a single L. huidobrensis pooled method, female abdomens were pooled, and the re- with four L. langei) to ensure that any L. huidobrensis sulting pooled genomic nucleic acids were screened present would be detected even at the lowest fre- using two PCR-based screening procedures. Males quency possible using this procedure. Although initial and females were treated differently to save genitalia results using the positive controls were acceptable of individual male specimens for morphological anal- (data not shown), during the course of the actual ysis while at the same time processing a large number screening, we detected unacceptable variation in the of specimens while minimizing the time and expense strength of the positive controls. of sequencing all specimens in the study. For this reason, we developed a second screening Males and DNA Sequencing. Male specimens were procedure using a CO site-speciÞc primer PCR test to dissected and the abdomens removed and subjected to determine whether L. huidobrensis DNA was present the DNeasy tissue extraction procedure (Qiagen, Va- in the pooled samples. This PCR test uses three prim- lencia, CA) without grinding. The essentially intact ers speciÞcally designed to be used together to amplify abdomens were left to soak in the lysis buffer with either a 356-bp piece of COI of L. langei or a 1,798-bp proteinase K overnight and were retrieved as vouch- piece of COI and COII of L. huidobrensis, depending ers following the extraction procedure. Primers TY- on which speciesÕ DNA is present in a genomic tem- J-1461 (forward, Winkler et al. 2009) and TD-N-3862 plate. The forward primer LirHL-F-1 (5Ј-TATAG (reverse, Simon et al. 1994) were used to amplify the TYGAAAACGGAGC-3Ј) was designed to perfectly entire mitochondrial COI and COII region. We se- match a region of COI of both L. langei and L. huidob- quenced only the COI region of Ϸ1,441 bp, including the rensis. The reverse primer Lang-R-2 (5Ј-ATAAAT DNA “barcode,” as previous work has shown that L. GYTGGTATAGG-3Ј) was designed to perfectly match huidobrensis and L. langei are easily distinguished us- L. langei COI sequence and also to be different from ing even shorter COI sequences (Scheffer 2000, Schef- L. huidobrensis sequence. During PCR, the use of this fer et al. 2006). Sequencing was accomplished using an primer in combination with the forward primer, ABI 3130XL (Life Technologies). The resulting se- LirHL-F-1, selectively ampliÞes a 356-bp piece of only quences were assembled and the consensus sequences L. langei COI even in the presence of L. huidobrensis aligned using Sequencher (Gene Codes, Ann Arbor, DNA. A second reverse primer, Huid-R-3 (5Ј-CCTG MI). These sequences were compared with previously GCGTAGCATCAACC-3Ј), was designed similarly to published sequences for L. huidobrensis and L. langei selectively amplify a 1,798-bp piece of L. huidobrensis using Neighbor-Joining in Paup* (Swofford 2001). COI and COII. When all three primers are used in a The closely related Liriomyza bryoniae (Kaltenbach) single PCR reaction containing L. langei, L. huidob- was used as the outgroup taxon for rooting the phe- rensis, or both, species-speciÞc DNA ampliÞcation October 2014 SCHEFFER ET AL.: MOLECULAR SURVEY FOR INVASIVE PEA LEAFMINER 1961

A B C D EFG H

L. huidobrensisbands

L. langeibands

Fig. 1. PCR banding patterns resulting from taxon-speciÞc priming reactions using eight different templates of various species compositions. Lanes A and F: Þve L. langei pooled; Lanes BÐE: four L. langei and one L. huidobrensis, pooled; Lanes G and H: single L. huidobrensis specimens. DNA ladder in outside lanes. takes place. Because of differences in the length of the cabbage, caulißower, celery, cheeseweed, lettuce, mus- species-speciÞc ampliÞcation products, visualization tard, peas, and spinach. Agromyzid samples from several on agarose gels results in size-speciÞc bands that are of the sampled counties, Fresno, Imperial, Riverside, San diagnostic for each species (Fig. 1, L. langei: Lanes A Diego, and Ventura, contained no ßies matching the and F; L. huidobrensis: Lanes G and H). If both species external description of L. langei or L. huidobrensis, and are present in the same samples, both diagnostic bands are not considered further in this study. will be present; we used multiple positive controls We screened a total of 48 male specimens and 616 consisting of one L. huidobrensis to four L. langei to female specimens from the counties of Monterey, San ensure sensitivity of the test to dilute L. huidobrensis Benito, San Luis Obispo, San Mateo, Santa Barbara, DNA (Fig. 1, Lanes BÐE). and Santa Cruz (Table 1). All DNA sequences from male specimens clustered with 100% bootstrap support with those previously obtained from L. langei Results (Fig. 2). Within the L. langei clade, there was no More than 1,000 agromyzid ßies were received from obvious clustering of sequences either by county of the California Department of Food and Agriculture. origin or by cropÐweed association. Uncorrected pair- Field crops and weeds from which the leafminer samples wise distances between the L. huidobrensis and L. were swept included beans, broccoli, Brussels sprouts, langei clades ranged from 4.8 to 5.0%. Distances within

Table 1. California county and crops from which screened ﬂies were sweep-netted during October 2006

No. of females No. of males County Crop screened screened Monterey Broccoli, caulißower, cheeseweed, lettuce 141 10 San Benito Broccoli, cabbage, celery, lettuce, spinach 52 11 San Luis Obispo Broccoli, cabbage, caulißower, celery, lettuce, spinach 129 0 San Mateo Beans, Brussels sprouts, mustard, peas 35 8 Santa Barbara Peas 76 8 Santa Cruz Broccoli, lettuce, spinach 183 11 Total 616 48 1962 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 107, no. 5

Fig. 2. Neighbor-Joining phenogram of L. langei COI sequences from specimens collected in California as well as previously published COI sequences from L. huidobrensis collected in South and Central America and elsewhere. Bootstrap values for the L. langei and L. huidobrensis clusters indicated on the branches. Within the L. langei cluster, previously published sequences are shown in bold. All of the sequences from L. huidobrensis and from L. bryoniae have been previously published; GenBank accession numbers are indicated on the Þgure for each published sequence.

L. langei ranged from 0 to 1.4% and within L. huido- langei. Anomalous bands were obtained in a few cases, brensis from 0 to 0.7%, although this latter estimate and for these samples the voucher heads and thoraxes does not reßect variation across the native range of L. were extracted and sequenced. In no case was L. huidobrensis (S.J.S. and M.L.L., unpublished data). huidobrensis DNA recovered, although in a couple of Limiting the analysis to only the DNA barcode region cases the DNA indicated the presence of an additional, of COI yielded essentially the same results as with the unidentiÞed species (this is not surprising, as several full-length dataset: the two species formed monophy- nonpest agromyzid species in California are externally letic clusters with 100% bootstrap support for each quite similar to L. langei or L. huidobrensis). species. All California specimens clustered with L. langei (phylogram not shown). Discussion Using the taxon-speciÞc priming approach for pooled female specimens, the resulting PCR bands in Our molecular survey of 664 male and female ßies almost all cases indicated only the presence of L. matching the external description of L. huidobrensis October 2014 SCHEFFER ET AL.: MOLECULAR SURVEY FOR INVASIVE PEA LEAFMINER 1963 and L. langei found no evidence for the presence of L. In summary, we used mitochondrial COI sequences huidobrensis in California. The counties from which and two PCR-based screening methods in an exten- leafminer samples were obtained were those known to sive, but temporally limited, investigation of whether or likely to contain “pea leafminers.” That only L. the invasive L. huidobrensis is present in California. langei DNA was found in these counties (with the Samples originated in a number of different counties exception of a few highly divergent sequences from and were associated with a variety of crop species. unidentiÞed species) indicates that L. huidobrensis From our collections made in 2006, no evidence for was not present in these regions in detectable num- the presence of L. huidobrensis was found. This infor- bers during October 2006. Given the multivoltine life mation and the genetic variation we have documented cycle of L. huidobrensis, it remains possible that this provide important baseline data in the event of the species could be present and undetected during yet introduction and possible establishment of the highly unsampled temporal periods. However, our results are invasive L. huidobrensis. consistent with those of Scheffer (2000; S.J.S. and M.L.L., unpublished data) that have found only L. langei sequence data in samples from other times and Acknowledgments locations within California. To date, there is no evi- We thank Joe Cavey (U.S. Department of AgricultureÐ dence that L. huidobrensis populations have ever es- Animal Plant Health Inspection Service) for facilitating this tablished and spread within California. project. Matthew BufÞngton, Allen Norrbom, and Scott Arm- Our samples of L. langei originated in six different strong provided useful comments on the manuscript. Spec- counties and were collected from 11 different crops. imens used in this study were collected by D. Alnawi, C. No evidence for geographic structure or clustering by Barrera, L. Bronson, C. Cannon, J. Dessert, J. Doyle, A. crop was seen, consistent with the general view of L. Gilbert, E. Gomez, L. Irons, S. Khowsudtsi, V. Lopez, E. langei as a polyphagous pest with at least moderate Moscoso, Q. Nguyen, B. Oliver, Y. Redler, S. Schmidt, B. dispersal abilities. Sidmore, J. Sohal, B. Taylor, J. Willems, and T. Wilson. The new multiplex PCR screening method using pooled samples is an ideal approach to screening large References Cited samples of morphologically similar but distinct species (or populations) that can be diagnosed using se- Blanchard, E. 1926. A dipterous leafminer on Cineraria, quence data. By aligning sequences from identiÞed new to science. Rev. Soc. Entomol. Argent. 1: 10Ð11. specimens from both groups, divergent sites can be Chaney, W. E. 1995. The pea leaf miner as a pest of vegetable crops. Monterey Co. Extension OfÞce Crop Notes found from which group (taxon)-speciÞc primers can (October). be designed. The primer in one direction should be Frick, K. E. 1951. Liriomyza langei, a new species of leaf- appropriate for both groups, while primers speciÞc to miner of economic importance in California. Pan-Pac. one or the other of the two groups should be designed Entomol. 21: 81. such that PCR ampliÞcations for the two groups will He, L., Y. Zhang, N. Xiao, J. Wei, and R. Kuang. 2002. Liri- be of different sizes, each diagnostic for one or the omyza huidobrensis in Yunnan, China: Current distribu- other group. One caveat concerning this method is tion and genetic structure of a recently established pop- that false-positive bands may be obtained if nontarget ulation. Entomol. Exp. Appl. 102: 213Ð219. specimens are inadvertently analyzed. Head, J., K.F.A. Walters, and S. Langton. 2000. The com- patibility of the entomophagous nematode, Steinernema Pooling specimens before DNA extraction allows feltiae, and chemical insecticides for the control of the greater efÞciency in terms of both money and time, South American leafminer, Liriomyza huidobrensis. Bio- due to a reduction in the numbers of both genomic control 45: 345Ð353. extractions and PCR reactions; by pooling Þve female Morgan, D.J.W., S. R. Reitz, P. W. Atkinson, and J. T. specimens per extraction, we cut costs and bench time Trumble. 2000. The resolution of California populations by Ϸ80%, thereby allowing us to screen more indi- of Liriomyza huidobrensis and Liriomyza trifolii (Diptera: viduals. In theory, this pooled approach could be used Agromyzidae) using PCR. Heredity 85: 53Ð61. to screen very large numbers of samples at one time. Parrella, M. P. 1982. A review of the history and taxonomy However, in the case of agricultural pests, especially of economically important serpentine leafminers (Liri- omyza spp.) in California (Diptera: Agromyzidae). Pan- when screening for rare or newly invasive species, Pac. Entomol. 58: 302Ð308. there is a need to be able to corroborate any positive Reitz, S. R., and J. T. Trumble. 2002. InterspeciÞc and in- bands resulting from the test, particularly because of traspeciÞc differences in two Liriomyza leafminer species the possibility of the inclusion of nontarget species. In in California. Entomol. Exp. Appl. 102: 101Ð113. our case, we retained the head and thorax of each Scheffer, S. J. 2000. Molecular evidence of cryptic species female such that if a positive L. huidobrensis band was within the Liriomyza huidobrensis (Diptera: Agromyzi- obtained, we could go back to the heads and thoraxes dae). J. Econ. Entomol. 93: 1146Ð1151. for that pooled sample and sequence for corrobora- Scheffer, S. J., and M. L. Lewis. 2001. Two nuclear genes tion. Obviously, the more specimens have been conÞrm mitochondrial evidence of cryptic species within Liriomyza huidobrensis (Diptera: Agromyzidae). Ann. pooled, the more voucher material has to be individ- Entomol. Soc. Am. 94: 648Ð653. ually sequenced for conÞrmation of positive bands, so Scheffer, S. J., A. Wijesekara, D. Visser, and R. H. Hallett. there is a trade-off between the number of specimens 2001. Polymerase chain reaction-restriction fragment- pooled in a given extraction and the ability to conÞrm length polymorphism method to distinguish Liriomyza important results. huidobrensis from L. langei (Diptera: Agromyzidae) ap- 1964 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 107, no. 5

plied to three recent leafminer invasions. J. Econ. Ento- brensis (Blanchard) (Diptera: Agromyzidae). Res. Bull. mol. 94: 1177Ð1182. Plant Prot. Jpn. 41:43Ð46. Scheffer, S. J., M. L. Lewis, and R. C. Joshi. 2006. DNA Takano, S. -I., R. Iwaizumi, Y. Nakanishi, and H. Someya. barcoding applied to invasive leafminers (Diptera: Agro- 2008. Laboratory hybridization between two clades of myzidae) in the Philippines. Ann. Entomol. Soc. Am. 99: Liriomyza huidobrensis (Diptera: Agromyzidae). Appl. 204Ð210. Entomol. Zool. 43: 397Ð402. Simon, C., F. Frati, A. Beckenbach, B. Crespi, H. Liu, and P. van der Linden, A. 1990. Prospects for the biological control Flook. 1994. Evolution, weighting, and phylogenetic of Liriomyza huidobrensis (Blanchard), a new leafminer utility of mitochondrial gene sequences and a compila- for Europe. Bull. SROP 13: 100Ð103. Weintraub, P. G., and A. R. Horowitz. 1995. The newest tion of conserved polymerase chain reaction primers. leafminer pest in Israel, Liriomyza huidobrensis. Phyto- Ann. Entomol. Soc. Am. 87: 651Ð701. parasitica 23: 177Ð184. Spencer, K. A. 1973. Agromyzidae (Diptera) of economic Winkler, I. S., S. J. Scheffer, and C. Mitter. 2009. Molecular importance. Dr. W. Junk B. V., Hague, The Netherlands. phylogeny and systematics of leaf-mining ßies (Diptera: Swofford, D. L. 2001. PAUP*. Phylogenetic analysis using Agromyzidae): delimitation of Phytomyza Fallen sensu parsimony (*and other methods). Sinauer, Sunderland, lato and included species groups, with new insights on MA. morphological and host-use evolution. Syst. Entomol. 34: Takano, S. -I., R Iwaizumi, Y. Nakanishi, H. Someya, and A. 260Ð292. Iwasaki. 2005. Genetic differentiation and morphological comparison between two clades of Liriomyza huido- Received 11 June 2013; accepted 23 July 2014.