Bemisia afer (Priesner & Hosny), 1970

Hernández-Suárez, Estrella, Martin, Jon H., Gill, Raymond J., Bedford, Ian D., Malumphy, Christopher P., Betancort, J. Alfredo Reyes & Carnero, Aurelio, 2012, 3212, Zootaxa 3212, pp. 1-76 : 15-17

Bemisia afer (Priesner & Hosny) sens . lat.

( Figures 13, 14, 16, 17, 19–27, 57, 59, 60, 62–64, 66–71, 78–79)

[Many of these illustrations are referred to more specifically in the accounts of forms A–H, below.]

Dialeurodoides afer Priesner & Hosny, 1934b: 6 .

Bemisia afer (Priesner & Hosny) Habib & Farag, 1970: 8–10 .

hancocki Corbett, 1936: 20 [Synonymised by Bink-Moenen, 1983: 95.]

citricola Gómez-Menor, 1945: 293 [Synonymised with hancocki by Mound & Halsey, 1978: 114.]

Background. Generally, Canarian taxa belonging to Bemisia sens lat . may be assigned to either the B. afer complex or the B. tabaci complex. As currently understood, and was discussed by Gill & Brown (2010) and by Martin (2005), the B. afer complex displays the following puparial characters: vasiform orifice inset from the posterior margin of the puparium by at least its own length (often considerably more), the orifice having emarginate sides; presence of two submedian geminate pore/porette pairings on each side of abdominal segment I between the median line and the first abdominal setae ( Fig. 31). In contrast, the B. tabaci complex can be distinguished by having the vasiform orifice straight-sided and inset from margin by less than its own length; only a single geminate pore/ porette pairing present mesad of the first abdominal seta on each side of abdominal segment I. Other characters that have been variously mentioned as useful to separate afer and tabaci complexes are the presence / absence of basal antennal spines, antennal position relative to fore legs, the proportion of the vasiform orifice length occupied by the operculum and the relative development of the caudal setae ( Gamarra et al., 2010; Malumphy, 2004; Malumphy et al., 2009). However, controversy remains and some Bemisia puparia present in the Canary Islands are not easily placed within one or other of these two main groupings, a situation that also occasionally occurs with some west African populations, where puparia exhibit a mixture of both afer and tabaci characteristics (Malumphy, personal observations, FERA quarantine interceptions). Eight more extreme morphological forms are discussed in detail below, all belonging to the B. afer complex, but not representing all of the variation seen in this group in the Canaries. Although the Canarian Bemisia populations are mostly assignable to either the afer or the tabaci assemblages, worldwide there are many Bemisia species that do not appear to belong to either grouping.

Distribution of B. afer sens lat. in the Canary Islands: GRAN CANARIA: Barranco de Azuaje, Barranco de Larice near Moya, Cruz de Tejeda, Los Tiles, Pinar de Tamadaba. TENERIFE: Barranco de Badajoz, Barranco del

Agua, Barranco de los Cochinos, Barranco de las Moradas, Cuevas Negras, Erjos, Güímar, Las Mercedes, Punta del Hidalgo, Pico del Inglés, Valle Guerra. LA GOMERA: Chorros de Epina, El Cedro, Garajonay. LA PALMA: Los Tilos, Pista de Barlovento. Elsewhere: Palaeartic Region: Corsica, Egypt, England ( Malumphy, 2003), France, Greece, Israel, Italy, Malta, Rhodes, Sicily, Spain, Turkey. Australia: Northern Territory, Queensland; Neotropical Region: Colombia and Perú.

Host plants of B. afer sens lat. in the Canary Islands: Artemisia thuscula , Bencomia caudata , Bystropogon odoratissimus , Cistus monspeliensis , Cistus symphytifolius , Conyza bonariensis , Echium giganteum , Echium pininana , Echium sp. , Echium virescens , Gesnouinia arborea , Laurus novocanariensis , Marcetella maderensis , Ocotea foetens , Persea indica , Phyllis nobla , Rubus bollei , Rubus sp. , Rubus ulmifolius , Salix canariensis , Teline microphylla , Viburnum rigidum . Host plants listed from elsewhere: Mound & Halsey (1978) recorded on 47 plant species belonging to 19 plant families, under the name B. hancocki , a list that was considerably augmented by Bink- Moenen (1983). The B. afer complex as a whole is highly polyphagous.

Comments: Bemisia afer sens . lat. appears to be an assemblage of taxa with a particularly broad range of puparial forms ( Gill and Brown, 2010). Most notable variations are the lengths of dorsal setae (from minute to nearly half as long as the width of the puparia or more) and the differing sizes, shapes and placements of dorsal elevations and protuberances (differing from granular elevations to star-shaped tubercles). These variable morphological characteristics are affected, at least in part, by host plant morphology, including presence or absence of leaf hairs, and differences in the morphology of the upper and lower leaf surfaces that affect the morphology of the puparia. Differing characteristics of whitefly puparia from the the upper and lower leaf surfaces may result in wide variation of puparia on a single plant ( Bink-Moenen, 1983, 1992). Such differences between puparia from upper and lower surfaces of leaves have also been repeatedly observed in the Canarian material of Bemisia afer complex ( Figs 13, 15, 25, 26). Characters that can change between upper and lower surface morphs include presence or absence of tracheal combs and tracheal fold stippling (which is only observed in the very smooth and flat specimens on the upper surface), length of dorsal setae (rather than the number of them) and presence and relative development of tubercles and cuticular sculpturing. Unfortunately, unless field appearance is noted at the time of collecting, it is usually impossible to tell whether unique wax patterns may have existed in life, because these waxes would have been removed in the slide-making process. These, and other problems arise when attempting to determine species limits within the B. afer complex.

It has been shown experimentally that Bemisia tabaci puparia can vary considerably when developing on different hosts ( Mound, 1963). However, several nominal species of Bemisia were placed in synonymy with B. tabaci by Russell (1957) before Mound’s experimental results were published. Neal & Bentz (1999) demonstrated that the development of dorsal structures such as setae, tubercles and papillae in the puparial stage depends on the tactile experiences of the first instar. Prior to our studies in the Canary Islands, it was generally thought that such puparial variation was more extreme in the tabaci complex than in the afer complex. However, intensive field sampling on native macaronesian plants, not only in the Canaries but also in Madeira and the Azores, has revealed more extreme morphological variation within the afer complex than seen in the tabaci complex.

It is believed that the Bemisia afer complex may have originated in Africa but it is now much more widely distributed ( Mound, 1965a). Mound was referring to B. hancocki Corbett , which is currently placed as a synonym of B. afer , although Martin (1987: 307) discussed characteristics which may lead to a reappraisal of this synonymy. Martin (1999) designated a lectotype for B. hancocki . The mooted origin in Africa may well be correct for the apparently endemic forms on the Macaronesian Islands, but this is probably too simple a theory when worldwide populations of the afer complex are considered. There are some morphological differences, of unknown significance, amongst populations of the afer complex worldwide, such as differing puparial lingular shape and moulting suture morphology. If populations of either tabaci or afer were moved across the globe by man alone, it is unlikely that these differences could have evolved in such a very short period of evolutionary time. From an evolutionary standpoint then, the current distribution of whiteflies across the globe, as well as some fossil evidence, suggest that whitefly morphology and species delimitation were well developed between 140 and 80 million years ago, suggesting that many present day genera and species were probably extant before the breakup of the Laurasian and Gondwanan landmasses. It is extremely likely that populations moved with the landmasses and have been isolated for millions of years prior to movement of whiteflies and other living organisms by man, starting around the 14 th century. Nevertheless, movement by man has seriously clouded the picture, as indicated by the movement of the “B” biotype of B. tabaci around the world starting in the early 1980s.

With the recent development of molecular technologies, it is now possible to find non-morphological differences between populations, which should eventually lead us to a better understanding of evolutionary history, phylogenies and species limits with in the Aleyrodidae . Hopefully, molecular expression within populations will not be influenced by the widely variable morphologies that occur, so that we will have a better understanding of the limits of that variability, and thus be able to more accurately define species.

Similar puparial morphs of the afer complex have been observed on different host plants and, conversely, widely differing morphs have been found on very similar host plants, across the northern islands of Macaronesia. For example, Bemisia lauracea Martin et al. (1996) is a member of the B. afer complex associated with the laurel forest in Madeira. It was described from Ocotea foetens , Laurus azorica and Persea indica . Bemisia lauracea can be recognized by the characteristic delineation of the submedian area of the puparium, which is also covered by a thick layer of glassy wax secretion in life. In addition to B. lauracea , several other forms of the B. afer complex have also been discovered on the same lauraceous hosts in both Madeira and the Azores, as well as in the Canary Islands.

We have considered it prudent not to describe all of the Canarian forms as new species, since to do so would require a much more extensive comparative study of populations from a wider area than that considered in this work. However, we have decided to name two of the Canary Islands forms of the Bemisia afer complex as new species, based on supporting adult morphology in one, and highly unusual wax secretions in the other. However, we feel that it is of considerable interest to discuss in detail, and illustrate here, eight of the other Canary Islands forms which we term A to H, without making premature judgements on whether these represent species, subspecies, forms, biotypes or varieties.

The following eight examples of variation within what appears to be the afer complex indicate the intricacy of the situation within the complex on the Canary Islands, and a similar situation exists in Madeira and the Azores. The general morphology of each Canarian “form” seems consistent with the parameters set forth for B. afer by Bink-Moenen (1983) which describe the open, narrow vasiform orifice, operculum not covering entire orifice, and elongate lingula with one pair of subapical setae, reduced seventh abdominal segment at the midline, as well as variable lengths of dorsal setae and protuberances in some individuals. Each form also displays four pore/porette geminate pairs on the first abdominal area between the abdominal setae, two pairs on either side of the median line (as in Fig. 31). Specimens of the afer- complex usually have very short caudal setae as compared to specimens in the tabaci complex. However, many of these Canary Island forms have caudal setae greatly elongated if other dorsal setae are also elongated. Most of these forms do not show the formation of extraordinary visible dorsal waxes but the very ornate appearance of form G puparia suggests that secretions are produced by the tuberculate dorsal protuberances. Since the tentacle-like projections on living and unprocessed specimens ( Fig. 70) do not appear the same as they do on slide-mounted individuals ( Fig 22) they therefore must be waxen extrusions. Differences amongst the eight undescribed forms and the other named afer -complex species in the Canary Islands are found in the dorsal setal lengths, numbers of discernable setae and their locations (particularly on the submargin and cephalic areas), the shapes and placement of the dorsal protuberances, and the production of visible waxen ornamentations. Upper and lower leaf-surface puparial morphologies on the same plant further complicates the situation.