Three cryptic species in Asecodes ( Förster ) ( Hymenoptera , Eulophidae ) parasitizing larvae of Galerucella spp . ( Coleoptera , Chrysomelidae ) , including a new species

Three morphologically very similar species of Asecodes Förster (Hymenoptera: Eulophidae) are reviewed. Asecodes parviclava (Thomson) is removed from synonymy under A. lucens stat. rev., and differentiated from A. lucens (Nees) and A. lineophagum sp. n. All three species develop as gregarious endoparasitoids in larvae of Galerucella spp. (Coleoptera: Chrysomelidae), but each species has its own unique host range. Asecodes lineophagum attacks only Galerucella lineola (Fabr.) and A. lucens only G. sagittariae (Gyllenhal), whereas A. parviclava parasitizes G. tenella (L.), G. calmariensis (L.) and G. pusilla (Duftschmid). The Asecodes species are similar but display small though distinct morphological differences, and are distinguished also through molecular differences. The genetic distance in mitochondrial CO1 ranged from 2.3% to 7.3% between the species. Five names, one valid and four synonyms, were available for this group of species, but none of them was linked to a primary type. To promote stability of nomenclature, primary types are designated for all five names, neotypes for Eulophus lucens Nees, Entedon mento Walker and Derostenus parviclava Thomson, and lectotypes for Entedon chthonia Walker and Entedon metagenes Walker. Entedon mento, E. chthonia and E. metagenes remain synonymized under A. lucens. JHR 30: 51–64 (2013) doi: 10.3897/JHR.30.4279 www.pensoft.net/journals/jhr Copyright Christer Hansson, Peter A. Hambäck. This is an open access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC-BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. REsEARCH ARTiClE Christer Hansson & Peter A. Hambäck / Journal of Hymenoptera Research 30: 51–64 (2013) 52


Material and methods
The colour photos were made with a Nikon SMZ 1000 stereomicroscope and a Nikon DS-5M camera.To eliminate reflections from the metallic and shiny body, a domelight manufactured as described by Kerr et al. (2008), was used as the light source for photography.Photos were taken at different focus levels and Helicon Focus Pro version 4.75 was used to merge them into a single image.The photography of wing interference patterns is described in detail by Shevtsova and Hansson (2011).The SEM photos were made from uncoated specimens on their original cardboard mounts.These were taken in low vacuum mode on a JEOL ® JSM 5600LV SEM microscope.
Morphological terms follow Gibson (1997).For illustrations of the morphological terms see www.neotropicaleulophidae.com.
The genetic distances were estimated from pairwise comparisons of 784 base pairs corresponding to the three prime end of CO1 (as used in Hambäck et al., unpublished).In the analysis, 7 individuals of A. lucens, 23 individuals of A. parviclava and 3 individuals of A. lineophagum were used.The analysis was done in PAUP* ver.4.0.a125(Swofford 2002), and genetic distances were calculated under the Kimura-2-parameter model (K2P) with pairwise deletion of missing data.Additional mitochondrial and nuclear genes data were used for reconstructing the phylogeny (Hambäck et al., unpublished), but were not included in the estimate of genetic distances due to a less extensive data set than for CO1.

Designation of primary types
Five names were available for this complex of species prior to this investigation, with one name, A. lucens, considered as valid and four other names as synonyms under the latter name.However, none of the names were fixed to a primary type and three of the names lacked type material altogether.For nomenclatural stability names must be fixed to a type specimen, and we therefore designate primary types, neotypes or lectotypes, for all five available names.

Neotypes
Eulophus lucens was described by Nees from a single female caught on Robinia pseudoacacia in Sickershausen [Bavaria, Germany] June 22, 1812, which was placed in his own collection.Specimens in the collection of Nees no longer exist apart from specimens sent to Westwood, now in the Oxford University Museum of Natural History (Graham 1988).No material of E. lucens exists in the Oxford collection (Graham 1988) and it is thus safe to assume that the holotype of E. lucens was destroyed.A female from Hungary, Vas County, Szakonyfalu, collected 23.vi.1960, agrees well with the description of E. lucens and is here designated as neotype.The locality in Hungary is the closest to the type locality where it has been possible to find material of this species.
Entedon mento was described by Walker from an unspecified number of males from near London and from Belfast.There is no material that agrees with the original description, either in the general collection or in the type collection of the Natural History Museum, London, where the material Walker based his descriptions is kept.In the type collection, the box supposed to contain some type material of E. mento is emp-ty, but with a note "mento??".Thus, it appears the material on which Walker based his description of E. mento is now lost.A female from England, Middlesex, Southgate, collected 6.vi.1972, fits the description of E. mento and is here designated as neotype for Entedon mento.The species was allegedly described from males, but males and females of this species are very similar.Further, Walker frequently misidentified the sex (Graham 1963), and a female is selected for neotype to make this species comparable to the other species of this complex, all of which are represented by females.
Derostenus parviclava was described by Thomson from an unspecified number of females he collected on Öland, an island in the Baltic Sea, and by G.F. Möller in Holmeja, a locality in Skåne, the southernmost province in Sweden.The collection of C.G. Thomson is in the Lund museum, and the collection of G.F. Möller is in the Natural History Museum in Gothenburg, both in Sweden.However, neither collection has any specimens from the type localities of D. parviclava.There is a female under the name D. parviclava in the G.F. Möller collection from Bökeberg (labeled "Bök"), which is a locality very close to Holmeja, one of the type localities.Because Thomson was very specific concerning localities from the province Skåne, where he lived and worked, it seems unlikely that he interchanged "Holmeja" with "Bökeberg".However, the female from Bökeberg agrees well with the original description of D. parviclava and it is from a locality very close to one of the original type localities.This specimen fulfills the criteria for a neotype for Derostenus parviclava and is designated as such here.

Lectotypes
The descriptions of Entedon chthonia and E. metagenes do not have information on the number of specimens used, and neither has been fixed to a primary type.The type collection of the Natural History Museum, London, has a specimen each of E. chthonia (type no.5.2603) and E. metagenes (type no.5.2604).These specimens fit the original descriptions and are here designated as lectotypes.

Biology of the parasitoids
The Asecodes species included here are gregarious koinobiont endoparasitoids of beetle larvae (Stenberg and Hambäck 2010).Females lay their eggs in the early larval stage and successful parasitoid development leads to a mummification of the host larva.The host larval mummies are black and morphologically resemble larvae, whereas unparasitized larvae form soft yellow pupae typical of chrysomelid beetles (photos in Hambäck 2004).After successful development the parasitoid larvae pupate inside the mummified host larva.The number of parasitoid pupae in the mummies is highly variable, from 1-14 within one mummy.The emergence of adult parasitoids typically occurs in intervals, which probably reflects separate egg-laying events and thus indicates superparasitism as a common trait.The number of parasitoids in each host affects both the sex ratio and the adult body size of emerging offspring.Single emerging parasitoids are invariably females, but at high parasitoid densities the sex ratio is male biased (up to 80%) (Stenberg and Hambäck 2010).It also seems that density dependence in sex ratio and adult body size are correlated with host species, as indicated by a comparison of parasitoids emerging from Galerucella tenella and G. calmariensis (Stenberg and Hambäck 2010).
Parasitism rates may at times be very high, close to 100%, but may at other times be very low.In Sweden, where this study was performed, there seems to be a latitudinal shift in parasitism rates, at least for some hosts.The parasitism rates for G. calmariensis and G. tenella in northern localities, close to Umeå, are typically very high, between 50% and 100%, but less than 10% in more southern localities.The genus contains both strictly monophagous species and oligophagous species and the different Galerucella species often occur in the same localities, but on different wetland plants.There are often large differences in the parasitism rates between Galerucella species within the same locality.For instance, parasitism rates may be very high on G. lineola and very low on other Galerucella species in one locality, whereas parasitism rates are high in another species in another locality.The different parasitism rates are not likely to be due to phenological differences or spatial distributions within localities because host plants colonised by different larval species may occur on neighboring plant individuals.

Identification
For identification of the species treated here the following additions can be made to the latest key to European species of Asecodes in Askew and Viggiani (1978).
Antenna as in Fig. 10.Frons below frontal suture with weak reticulation, above suture smooth; antennal scrobes join on frontal suture.Vertex with very weak reticulation inside ocellar triangle, smooth outside triangle.

Discussion
The separation of A. lucens into three species based on molecular and morphological evidence is supported by biological data.Asecodes lucens and A. lineophagum were reared from only one host species, Galerucella sagittariae and G. lineola, respectively, whereas A. parviclava was reared from three host species, G. tenella, G. calmariensis and G. pusilla.These observations have also been confirmed with independent observations of parasitoid behaviour in the laboratory, where females were found to attack the respective host species, but not other species (L.Fors, unpubl.data).The delimitation of three species is also supported by observations in the field, where one species of Galerucella larvae may be heavily parasitized and another is not attacked in the same locality.Such parasitism patterns have, however, not been observed for G. tenella, G. calmariensis and G. pusilla.In fact, earlier studies show strong correlations in parasitism rates between G. tenella and G. calmariensis among localities (Hambäck et al. 2006).Moreover, field observations suggested that parasitism rates on G. tenella were consistently higher when this species was sympatric with G. calmariensis.The reason for this pattern was not resolved (Hambäck et al. 2006), but the pattern suggest that the parasitoid population may mediate indirect interactions between its hosts, as is known for other host-parasitoid systems.The current information on the species delimitation within Asecodes was important to identify pairs of host species where such effects would be likely.Based on previous information, we could have expected similar indirect effects also for other species pairs but the novel information on population differentiation among parasitoid individuals suggest this not to be the case.In view of our findings, the previous host records of Galerucella nymphaeae and Lochmaea suturalis for A. lucens need confirmation.Investigation of Asecodes specimens reared from these hosts might quite possibly reveal additional cryptic species in this group.