Introduction

The Eriophyoidea is a large superfamily of obligate plant-feeding mites, with over 4,000 described species (Zhang et al. 2011). The eriophyoid mites are highly host-specific, with nearly 80% of them known from a single host species, 95% from one host genus, and 99% from one host family; furthermore, non-monophagous species are often found on closely related host species (Skoracka et al. 2010).

Aceria genistae (Nalepa, 1891) sensu lato is native to Europe and has been reported from species of three plant genera Cytisus, Ulex and Genista (Fabaceae) (Amrine & Stasny 1994; Smith et al. 2010). This is an unusually wide host range for a species of Aceria, most of which are monophagous; only some species of Aceria are known to feed on multiple species within one plant genus, and a species of Aceria rarely attacks related plant genera (Lindquist 1996a). It has been suggested that there might be a complex of cryptic mite species or host-specific biotypes associated with the brooms and gorse (Chan & Turner 1998; Syrett et al. 1999; Smith et al. 2010). Castagnoli (1978), working in Italy, showed that the mite resembling A. genistae on Spartium junceum L. (Spanish broom) was a distinct species, A. spartii (Canestrini). Working on the Aceria genistae species complex in New Zealand, Xue et al. (2015) recently described a new species, Aceria davidmansoni, found on gorse Ulex europaeus L., which Manson (1989) had previously misidentified as A. genistae. They confirmed the earlier suspicion of a species complex on Scotch broom and gorse in New Zealand: A. genistae on Cytisus scoparius only and A. davidmansoni on U. europaeus. Xue et al. (2015) presented both qualitative and quantitative descriptions and comparisons for differentiating A. genistae and A. davidmansoni in New Zealand. They also fixed the identity of A. genistae by designating a neotype collected from C. scoparius in Europe. This provides a foundation for clarifying taxonomic problems for mites identified or reported as A. genistae in other areas such as USA (Chan & Turner 1998; Smith et al. 2010; Andreas et al. 2013).

Aceria genistae has become an important species of Aceria because of its potential as a biological control agent against Scotch broom (C. scoparius) in Australia and New Zealand (Syrett et al. 1999). In North America, A. genistae was first discovered on Scotch broom in Portland, Oregon, and Tacoma, Washington in 2005 and was believed to be accidentally introduced (Andreas 2010; Smith et al., 2010; Andreas et al. 2013). It is now widespread between Oregon and Washington in the USA and extends to southern British Columbia in Canada (Andreas et al. 2013), and has been found at many sites in the northern half of California between 2014 and 2015 (Andreas, personal communication). The mite appears to damage C. scoparius in the field, reducing flowering and plant biomass, although quantitative data from controlled experiments are not available (Smith et al., 2010). At high mite densities, stem mortality and even whole plant mortality were observed (Andreas, pers. comm.). Chan & Turner (1998) reported A. genistae from both gorse and French broom in California, USA, but also discussed the possibility of a species complex. Host specificity tests (in both greenhouse and fields) of A. genistae against C. striatus (Portuguese broom), Lupinus densiflorus and Lupinus albicaulis were conducted using mites from C. scoparius in Washington State and a species complex is suspected (Andreas 2010 & pers. comm.).

This study was initiated during our earlier work on New Zealand eriophyoid mites in general (Xue & Zhang 2008) and the “Aceria genistae” complex in particular (Xue et al. 2015). For comparison with New Zealand material, we loaned specimens identified as “Aceria genistae” from Europe and North America, including specimens from California sent by both K. L. Chan and Jim Amrine. J. Andreas also sent specimens of her host specificity tests and collections from western US (Andreas 2010). In this study, we examined these mite specimens suspected to be A. genistae from a series of plant hosts of the tribe Genisteae, including C. scoparius, C. striatus, U. europaeus, L. densiflorus and L. albicaulis, and G. monspessulana from western USA and Canada. This paper aims to identify the eriophyoid mite species from different hosts in USA. We also briefly discuss mite-plant association in relation to phylogenetic relationship among those host species (Genisteae).

Material and methods

Specimens of eriophyoid mites were collected from USA and Canada, with details presented in the material examined section for each species. The morphological terminology used in this paper is that of Lindquist (1996b) whereas the system of classification follows that of Amrine et al. (2003). General methods of study follow Xue et al. (2015). There were not enough specimens from every host plants to allow a full multivariate discriminant analysis; thus a simple analyse of variance (see Table 1) in GenStat 10 was used for measurements and LSD was used to separate means in multiple comparisons at the significance level of 0.05. The sample sizes are listed in specimen examined section for each species. All slide-mounted specimens are vouchered in the New Zealand Arthropod Collection (NZAC), Landcare Research, Auckland.

TABLE 1.

Analysis of variance main morphological distinguished characters among Aceria genistae (Nalepa) and Aceria davidmansoni Xue, Han & Zhang 2015 from different hosts and locations (for female individuals).

Species accounts

Discussions on host relationships

In both Australasia and North America, host specificity tests and biological observations suggested a complex of cryptic Aceria species associated with the brooms and gorse (Chan & Turner 1998; Syrett et al. 1999; Smith et al. 2010). In a previous paper, we showed that in New Zealand, the real A. genistae attacks only C. scoparius, and a related mite species—A. davidmansoni, previously misidentified as A. genistae by Manson (1989)—is present on U. europaeus (Xue et al. 2015). In this study, we showed that the mites from U. europaeus and G. monspessulana in California, USA, previously misidentified as A. genistae (Chan & Turner 1998), are in fact A. davidmansoni. In both cases, A. davidmansoni can be clearly distinguished from A. genistae qualitatively by the pattern on the prodorsal shield and quantitatively by several characters such as the number of annuli on the hysterosoma (Xue et al. 2015; Table 1 of this paper).

Specialist herbivores such as eriophyid mites show strong host-dependent morphological variation in adaptation to different host species and environmental conditions (Skoracka et al. 2002). For example, Skoracka et al. (2014) reported that some characters (the length of the c2 setae, the number of dorsal annuli, and the length of chelicerae) could differentiate host-related species of eriophyoid mites (between the DBM and MT-1 genotypes), resolving the previous taxonomic confusion between wheat curl mite (Aceria tosichella) and dry bulb mite (Aceria tulipae). Our results, both this one and Xue et al. (2015), also showed that such characters as the length of setae c2 and e could differentiate A. genistae and A. davidmansoni, in addition to qualitative features (the pattern on the prodorsal shield) Within species, there is evidence of morphological variation in adaptation to different environmental conditions and host species. For A. genistae on Scotch broom, mites in the field and greenhouses are similar in most characters, but setae c2 were slightly shorter in specimens from the greenhouse (Table 1, line 8 versus line 9). For A. genistae on Cystius, most dorsal setae (sc, c2, d, f) were shorter in specimens from C. striatus than those from C. scoparius(Table 1, line 9 versus line 10). For A. genistae on Scotch broom and L. densiflorus intermixed in greenhouses, divergence in only setal length of e was observed (Table 1, line 7 versus line 8). For A. genistae on Scotch broom and L. albicaulis intermixed in the field, divergence in setal lengths of sc, c2, e, and f was observed (Table 1, line 6 versus line 9).

We summarized the relationships within tribe Genisteae based on the phylogenetic information in Pardo et al. (2004) and Cristofolini and Feoli-Chiapella (1981), and listed different eriophyoid mite species known from these plants (Table 2). Lupinus is distantly related to Cystius, and it is therefore not surprising that A. genistae failed to form galls on Lupinus in host specificity tests in Washington. In contrast, C. striatus is of same genus as broom and A. genistae was able to establish on this host species. Likewise, U. europaeus and G. monspessulana are more closely related and it therefore does not seem surprising that A. davidmansoni was found naturally on both plants in the wild in Daly City, California. It should be noted that most other species of Aceria (except A. spartii) in Table 2 are poorly known and in serious need of taxonomic revision. There are significant gaps in both the knowledge of the relationship among the host plants in Genisteae and the knowledge of Aceria on these related host species. However, there seems little doubt that there is a complex of Aceria species on Genisteae, some being perhaps monophagous and some having slightly broader host ranges (feeding on closely related host species) with minor intraspecific variation.

TABLE 2.

The relationships of host plants, eriophyoid mites from the related host plants and its reference source.