The citrus leafminer, Phyllocnistis citrella Stainton (Lepidoptera: Gracillariidae), was first detected in Florida in 1993 and quickly spread throughout the state (Heppner 1995). Leaf mining causes a decline in leaf photosynthesis and increased susceptibility to citrus canker (Graham et al. 2004; Gottwald et al. 2007). Incidence and severity of P. citrella (CLM) damage have increased in Florida recently, possibly due to resurgence of leafminer populations resulting from intensified spray programs directed against the Asian citrus psyllid (Diaphorina citri). Consequent increases in canker are especially notable in young trees and susceptible varieties such as grapefruit and early season oranges (Dewdney & Graham 2012).

One of the most common methods to monitor CLM abundance is to use a pheromone lure to attract the adult moth. Ando et al. (1985) reported attraction of male P. citrella in Japan to (Z,Z)-7,11-hexadecadienal (diene). Elsewhere, including Florida, the diene alone did not attract males (Jacas & Pena 2002). Leal et al. (2006) identified two additional aldehydes from pheromone glands of female P. citrella from Brazil, (Z,Z,E)-7,11,13-hexadecatrienal (triene) and (Z)-7-hexadecenal (monoene), in a ratio of 30:10:1 triene:diene:monoene. However, inclusion of the mononene in a ternary blend did not increase trap catch of males in Florida compared with a 3:1 triene:diene blend (Lapointe et al. 2006). Moreira et al. (2006) also identified the triene and diene compounds from P. citrella populations in California, USA, and reported an optimal ratio of 3:1 triene: diene to attract males. The 3:1 ratio was confirmed as optimal in Florida (Lapointe et al. 2009). Recently, the presence of a congeneric leafminer native to Florida, P. insignis (Frey & Boll), was observed in sticky traps baited with P. citrella lures loaded with the 3:1 blend (Keathley et al. 2013). Here we provide evidence that the 3:1 blend of (Z,Z,E)-7,11,13-hexadecatrienal and (Z,Z)-7,11-hexadecadienal attracts multiple unrelated and genetically distinct species of Phyllocnistis in southern Florida.

Recent research in Florida has focused on the use of pheromone traps to determine optimal application timing of insecticides, efficacy of insecticides, optimal trap density, correlation between CLM damage and adult trap counts, and number of CLM generations per year (Jones, unpublished). The traps (Great Lakes IPM, IPS-G004) containing an insecticide dispenser (Vaportape™ II, Great Lakes IPM, HC-8500-25) and a lure loaded with a 3:1 blend of (Z,Z,E)-7,11,13-hexadecatrienal and (Z,Z)-7,11-hexadecadienal (IT203 ISCAlure-Citrella, ISCA Technologies, Riverside, California) were monitored weekly between Feb and Nov (2011 and 2012), and bi-weekly from Jan through Dec. Trap locations included large commercial citrus groves in Collier, Hendry, Lee, and St. Lucie Counties, Florida, the U.S. Horticultural Research Laboratory at Ft. Pierce, Florida and a 35,000 acre unmanaged natural area in Okaloacoochee Slough State Forest (Table 1). The latter is a habitat characterized by marsh, cypress, wet prairie, pine flatwoods, oak hammocks, and oakpalm hammocks straddling southwestern Hendry and northwestern Collier counties. A single trap was centrally located in the upper canopy of trees of each selected citrus grove, and 12 traps were placed in the slough at varying distances (1.6, 3.2, 4.8, and 6.4 km) from known citrus. Pheromone lures were replaced every 6 to 8 weeks.

Moths were separated by morpho type and specimens of each type were sequenced for the 658 bp “barcode region” of the mitochondrial cytochrome c oxidase I (CO1) gene following our published techniques (e.g., Kawahara et al. 2013; Kawahara & Rubinoff 2013; Rubinoff et al. 2012). The COI sequences generated were combined with known Phyllocnistis COI sequences from GenBank ( www.ncbi.org) and BOLD ( www.boldsystems.org). We included 4 gracillariid outgroups, Acrocercops astericola, Cameraria ohridella, Phyllonorycter acerifoliella and P. junoniella. Sequences were assembled, edited, and aligned using Geneious 5.4 (Biomatters). We applied the “Geneious Alignment” option with default settings and manually checked the alignment. As we have done previously in our studies (e.g., Kawahara et al. 2011; Kawahara & Rubinoff 2012; De Prins & Kawahara 2012), the final dataset was subject to a maximum likelihood phylogenetic analysis. We conducted 1000 best tree searches and 1000 bootstrap replicates, applying a GTR+I+G substitution model with a random starting tree in the program GARLI (Zwickl 2006). All sequences are available in GenBank ( www.ncbi.nlm.nih.gov).

TABLE 1.

THE NUMBER OF EACH MORPH (A–F) COLLECTED, AS DETERMINED BY PHYSICAL EXAMINATION FOR EACH LOCATION AT COMMERCIAL CITRUS GROVES IN COLLIER, HENDRY, LEE, AND ST. LUCIE COUNTIES, FLORIDA (2011–2012). GROVES 2 AND 3 ARE COMMERCIAL ORANGE. “PHYLLOCNISTIS SP.” REFERS TO UNIDENTIFIED PHYLLOCNISTIS SPECIES THAT FALL INTO CLADES 4 AND 5 IN FIG. 1. “OTHER” REFERS TO NON-PHYLLOCNISTIS SPECIES.

Phyllocnistis citrella, P. insignis, P. vitegenella and 2 unidentified congeners were obtained from traps baited with ISCAlure Citrella™ lures at locations in Collier, Hendry, Lee, and St. Lucie Counties. Based on COI sequence data, at least 5 genetically divergent Phyllocnistis species were attracted to the lure. Two unidentified species (circles 4 and 5, Fig. 1) share a separate origin from P. citrella and constitute a genetically distinct group from P. citrella and P. insignis. Because this preliminary study was conducted at a limited number of sites in Central Florida, it is possible that additional species in the genus might be attracted to the major components of the P. citrella pheromone. Species in the genus often appear morphologically similar based on wing pattern (Kawahara et al. 2009; De Prins & Kawahara 2009; Davis & Wagner 2011), therefore it is likely that past surveys overlooked the presence of multiple Phyllocnistis species in traps. The morphological similarity of these leaf miner species implies that estimates of P. citrella infestation in citrus groves might be influenced by non-P. citrella species, and therefore caution is required when making estimates on damage based on the number of moths attracted to lures.

This study was a preliminary investigation of Phyllocnistis species attracted to pheromone lures. We sequenced only one gene for this initial screening, and we plan to sequence additional samples and loci in the future. A study that utilizes a combination of morphology and multiple molecular markers (e.g., Mitter et al. 2010) will be necessary to conclusively determine the true identities of the clusters we observed from the COI barcode region alone.