Steneotarsonemus ananas (Acari: Tarsonemidae): a complementary description from Australian pineapples and a new pest on Neoregelia spp. (Bromeliaceae) in Costa Rica

The pineapple tarsonemid mite Steneotarsonemus ananas (Tryon), 1898 is associated with pineapple around the world, including Costa Rica. Here, we report its association with Neoregelia sp. (Bromeliaceae) for the first time. These mites damaged the plantlets considerably, affecting the esthetic quality of the product and thus causing economic losses to the local growers. This discovery prompted us to redescribe the adult females and males from the type host and locality (pineapple in Queensland, Australia) and compare it with the original description, subsequent descriptions, and material collected from the USA, Costa Rica, and Colombia on pineapple, Neoregelia, and an unidentified bromeliad. Further specimens from pineapple in Guatemala were identified as the same species and represent the first record for this country. We concur with previous observations that the original description by Tryon is not the same species in subsequent descriptions. We tentatively consider all specimens examined the same as those described as S. ananas in succeeding descriptions, i.e., Steneotarsonemus ananas Tryon sensu Beer.

Costa Rica exports a great number of ornamental plants of commercial interest to different markets around the world, including bromeliads in the genera Bromelia and Neoregelia. Although mites often associate with Bromeliaceae (e.g., Nesbitt 1985; Guerra et al., 2012, including S. ananas from pineapples in Costa Rica (Ochoa et al., 1991), no mite species were reported as pests of Bromelia and Neoregelia in Costa Rica. Here, we provide the first record of S. ananas found associated with the ornamental bromeliad Neoregelia spp. Its discovery raised questions about the identity of the species on both pineapple and the new host plant, leading us to redescribe S. ananas from pineapple so that it could be reliably compared to other material. We also describe the symptoms induced by this tarsonemid mite on Neoregelia.

Material and methods
Like many bromeliad genera, Neoregelia is native to the South American rainforests, specifically from southeastern Brazil (Royal Botanic Garden, Kew Science, n.d.), as are pineapples (Coppens d´Eeckenbrugge and Leal 2003), but the first Neoregelia cuttings were brought to Costa Rica from Europe more than 20 years ago and now mother plants are all produced in Pococí, Limón, Costa Rica. Some of the cuttings are developed to the seedling stage on another farm located in the province of Alajuela (810 masl). When they are ready for production, they are transported to the province of Limón to continue their growth. Mother plants are also produced on the farm in Limón, consequently materials from both places are mixed and grown together. Farm workers are responsible for the selection of mature plants and packing them in boxes for shipment to the different markets around the world.
A commercial plantation in Costa Rica, comprising many different bromeliad plant species, was visited on February 2018 because numerous plants were attacked by a small mite, causing visible symptoms. The damage was negatively affecting the quality of the product for export, particularly bromeliads in the genus Neoregelia. Plants for export should lack any injury and foliage must be free of spots, deformations, or any other visible damage.
The plantation, which utilized a black shade cloth (Saran) system, is located in the canton of Pococí, province of Limón, at 467 m above sea level. This zone is known for having high temperatures (24 to 30 o C throughout the year) and humidity (90-100% during the year), as well as considerable annual precipitation mostly through the rainy season. Eighteen samples, consisting of whole plants, were collected in plastic bags and placed into a cooler. The samples were transported to the Laboratory of Acarology at the University of Costa Rica and processed according to the laboratory protocols.
The samples were observed under a stereoscope-microscope, individual leaves were detached manually; then, the symptoms and mite behavior were recorded. The mites were collected in a Syracuse dish with 70% ethanol and mounted on slides in Hoyer´s medium (Krantz and Walter 2009). The slides were processed in an oven at 50 o C for three days or until the mites were completely clarified.
Once the mites were ready, the slides were sealed with glossy polyurethane varnish and examined with the help of Olympus BX 51 and Zeiss Axio Imager D1 microscopes, both with phase contrast and DIC (Differential interference contrast), as well as an Axio Imager D1 with DIC.
For the material from Costa Rica, five slides were deposited in the mite collection of the Laboratory of Acarology (CIPROC), University of Costa Rica and two slides with male and female specimens, as well as two larvae, at the Smithsonian (NMNH), National Insect and Mite Collection, Washington DC. Additionally, three larvae, two females and one male from Costa Rica were later prepared and deposited in the Queensland Museum (QM). The Australian material of S. ananas was found as an unidentified tarsonemid mite species in the Queensland Agriculture collection housed at the Queensland Museum (QM). This material, from pineapple in southeast Queensland (the type locality), was examined with phase contrast (Leica DMLB) and DIC (Leica DM2500). Other material examined is housed in the NMNH. A photograph of the first record in the USA of S. ananas on Aechmea (Figure 1) is provided. Morphological terminology follows Lindquist (1986).  Beer Steneotarsonemus Beer, 1954: 1229. Type-species Steneotarsonemus hyaleos Beer, 1954, by original designation. Parasteneotarsonemus Beer & Nucifora, 1965: 40, Lindquist, 1986: 274. Neosteneotarsonemus Tseng & Lo, 1980: 127; Lindquist, 1986 Steneotarsonemus are characterized for having: females with stigmata close to setae v1, body elongate-slender, sejugal and poststernal apodemes reduced; males often with round lobes on femorogenu IV; both genders with small pharynx and subcircular to subquadrate gnathosoma (Beer 1954; Lindquist 1986. A full description is provided by Lindquist (1986). (Tryon, 1898) Tarsonemus ananas Tryon, 1898: 458. Steneotarsonemus ananas (Tryon), Beer, 1954Beer, : 1276.

Diagnosis
All life stages: femur II with three setae; tarsi II-III elongate, longer than wide, setae tc′ on tarsus I distinctly proximal to tc″. Female (Figure 2). Posterior margin of prodorsal plate medially concave and crenate; seta v1 positioned at anterolateral corner of prodorsal plate, anterior of stigmata; bothridial seta sc1 capitate, its head obovate; seta c2 positioned far anterior of seta c1; pore ia positioned far anterior of seta d; setae c1, d, e, f slender to slightly thickened, stiff; seta h thickened; c2 about twice as long as c1 and filiform; setae e and h with minute barbs. Paired remnants of sejugal apodeme each highly curved, boomerang-like; apodemes 3 reaching level of seta 3a; apodemes 4 well developed, almost reaching mid-line; seta 2a positioned on or slightly behind apodeme 2; seta 3a short, not reaching trochanter III; leg IV moderately long, extending to or just beyond idiosomal margin; femur I seta d lanceolate, l′ and l′′ thickened; femur II seta l′ thick; genu II seta l′ thick; seta G v′ IV not reaching base of seta Ti v′ IV. Male (Figure 3). Prodorsum with four pairs of setae; seta sc1 at least three times as long as seta sc2; seta c2 not reaching base of c1; seta c1 not reaching base of seta d; coxisternal plates smooth; apodeme 5 well-developed, forked; femur, genu, tibia II-III without sexually-dimorphic spine-like setae l′, v′; leg IV with small abaxial femorogenual flange and large, semi-circular adaxial flange; seta v′ Ti 30-54 long, thick. Larva. Setae v1 short, tips not overlapping; setae h1 long, about twice length of h2. Seeman et al. (2016) brought together the two existing subgeneric classifications of Steneotarsonemus, recognizing six subgenera. Steneotarsonemus ananas is best placed in nominate subgenus Steneotarsonemus. However, it does not match this subgenus perfectly as the small abaxial flange on the male femorogenu IV is present in S. ananas and Parasteneotarsonemus. However, this feature is easily missed, and may be present elsewhere in Steneotarsonemus.

Differential diagnosis
Many descriptions of Steneotarsonemus are rudimentary and sometimes lack even basic illustrations of females (e.g., Cromroy 1958) which are essential for classification. Thus, our proposed diagnostic features are tentative. Nevertheless, Steneotarsonemus ananas is unique in Steneotarsonemus by having females with a crenate posterior margin of the prodorsal plate. However, the crenate prodorsal margin does occur in some females of Tarsonemus such as T. stammeri Schaarschmidt, 1959, as well as on Deleonia aguilari Goldarazena & Ochoa, 2002(Goldarazena et al., 2002. The two other species of Steneotarsonemus found on pineapple are S. comosus and S. perezi. Steneotarsonemus ananas is differentiated from S. comosus by having females with setae c2 about twice the length of c1 (c1 and c2 subequal in S. comosus) and coxal setae 3a and 3b subequal (3a about twice as long as 3b in S. comosus); and a male with leg IV with a large rounded flange (versus a narrower rectangular flange in S. comosus). Steneotarsonemus perezi is distinguished from S. ananas and S. comosus by having males with setae v1 much longer than v2 (versus v1 shorter than v2).
Legs (Figure 7). Leg III longer than legs I, II and IV. Leg IV cylindrical; terminal seta tc″ filiform, about three times longer than subterminal seta v′Ti; subterminal setae stout, finely pilose; femorogenu with genual seta v′G slightly shorter than seta v′F.
Plate CD with setae c2, filiform, two times longer than c1. Setae c1 and d stout, barbed. Plate EF with setae f, stout, pilose. Setal lengths vary between populations: Australian specimens with much shorter setae compared with Costa Rican and Colombian populations, USA intermediate (Table 2). Genital capsule as long as wide, with accessory copulatory structures ps1 spinelike, pointed apically.
Legs. Legs I-II, IV subequal in length; leg III longer than leg IV. Leg IV with large adaxial rounded flange and narrow distal abaxial flange ( Figure 10B).
Legs. Leg setation for legs I-III similar to female except lacking ϕ2 on tibia I, proral setae on tarsus I and pv′ on tarsi II-III. Form of setae similar to female except for: femur I seta d only slightly thickened, peg-like, 2-3, other setae slender, smooth; genu II seta l′ not shorter than other setae; femur II, genu II setae slender, smooth; tibia III with short, thick thorn-like seta l′, 3.

Discussion
Tryon (1898) described S. ananas from mites collected on pineapples showing severe symptoms of fruitlet core rot, which is caused by Fusarium fungi and linked to the feeding of S. ananas by Tryon (1898) and later by several studies (e.g., Petty et al., 2006). Unfortunately, Tryon (1898) did not mention the specimens on which the species is based: Leg IV seta v'G length 8-11 12-14 8-10 9-10 Leg IV seta v'F length 13-14 15-16 12-15 12
they are merely an undisclosed number of mites from pineapples in southern Queensland. Ewing (1939) did not examine type specimens, instead relying on Tryon's manuscript and female and larval material from Hawaii. Beer (1954) considered the whereabouts of Tryon's material "unknown″, again relying upon Ewing´s material from Hawaii. In the past decade or so, several unsuccessful efforts were made to locate these type specimens at the Queensland Department of Agriculture collection, the Queensland Museum, and the Australian National Insect Collection. Tryon's drawings and descriptions are, not surprisingly for their time, rudimentary, however, we doubt that the species described by Tryon (1898) is the same species later called S. ananas by subsequent authors. Tryon (1898) provided ventral views of each sex. The male shows a few significant differences from later descriptions (Beer 1954; Jeppson et al., 1975. First, the flange on leg IV is absent in Tryon (1898), but with a low-powered microscope, it is unlikely but possible that the flange was misinterpreted as a spine-like process. Second, in his original description the male is widest just anterior to seta c1, whereas in subsequent descriptions of the male, it is widest just anterior to seta c2. Third, Tryon (1898) also shows an undivided apodeme 5, but male S. ananas have a posteriorly forked apodeme 5. The female presents further significant differences that are incongruous not only with S. ananas in subsequent descriptions (Ewing 1939; Beer 1954; Jeppson et al., 1975 but also with Steneotarsonemus. Instead of an elongate body as in Steneotarsonemus, it is ovate; rather than short setae h, they are long; contrary to separate apodemes 3, they are joined; instead of lacking a post-sternal apodeme, it is present; rather than separate apodemes 4, they join the post-sternal apodeme. Therefore, the mite described by Tryon is not S. ananas as currently defined. It is not Steneotarsonemus. A similar conclusion is reached in Kolicka et al. (2016), who doubted that S. ananas (Tryon, 1898) was the same species described by later authors as S. ananas. What is S. ananas (Tryon, 1898)? We are unaware of any species of tarsonemid that matches his original description. The only tarsonemid mites described from Bromeliaceae are the six species known from pineapple and records of S. ananas from other bromeliad hosts (Lin andZhang 2002; Kolicka et al., 2016). Thus, as Tarsonemus is the only other tarsonemid genus known from Bromeliaceae, perhaps Tryon (1898) described a species of Tarsonemus rather than Steneotarsonemus. This genus at least has females with an ovate body and more complete apodemes. Unfortunately, the three species of Tarsonemus recorded on pineapples (T. bilobatus, T. buchelerei, and T. merus (Suski 1965; Smiley 1967; Flechtmann 1971; Lin & Zhang 2002) bear no great similarity with S. ananas sensu Tryon (1898). Until material matching Tryon's description is found, the true identity of S. ananas will remain a mystery. Ewing (1939) provided the first description of S. ananas after Tryon (1898). While he copied the description of the male from Tryon, his description of the female was new. However, the specimens examined, two females and a larva ("nymph″), were poorly and incompletely described making the description unsuitable for identification purposes or comparison with Tryon (1898). Thus, the first complete description is that of Beer (1954), who also provided the first description of the male, and we thus consider the species Steneotarsonemus ananas (Tryon, 1898) sensu Beer (1954). Unfortunately, the specimens of Ewing (1939) have not been located, and Beer´s (1954) material in the University of Kansas is not available.
The specimens from Australia, described here, are from the type locality and host for S. ananas, but are clearly not the same species described by Tryon (1898) and instead match S. ananas sensu Beer. Comparisons between females from Queensland pineapple, Costa Rican Neoregelia and pineapple, Colombian bromeliad and Californian (USA) ornamental bromeliad demonstrated not only concordance with diagnostic features but also measurements (Table 1). However, males were less similar (Table 2). While they all agree in general diagnostic features, some setal lengths differ considerably in size with males from Queensland pineapple having much shorter setae than those found on Costa Rican Neoregelia and Colombian bromeliads. As intraspecific variation in males is known in Tarsonemidae (e.g., Tarsonemus waitei Banks, Lindquist (1978)) and other collections from USA pineapples were intermediate in size, we tentatively consider these differences as intraspecific. Thus, they are all S. ananas sensu Beer (1954), pending further studies into the identity of Steneotarsonemus species on pineapple throughout the world.

Description of symptoms on Neoregelia spp. (Bromeliaceae)
Mites were observed grouped in the interior of the plant, especially hidden on the underside of the leaves, near the axils of the younger foliage. The leaves also form rosettes, which compose microreservoirs in the axils, beneath the main phytotelmata of the plant. This provides a good habitat for many organisms, including tarsonemid species, which require warm temperatures, high humidity and low light intensity to develop (Jeppson et al., 1975). The feeding damage shows a light brown color, mostly in the middle of the young leaf surface ( Figures 14A and  14B), while the apex remains green and the base of the leaf white in color, which serves as a depository camouflaging a high concentration of wandering mites and eggs.
When the leaves develop and become more coriaceous, red spots of different diameters appear along the leaves upper side. These spots are the remnants of the earlier feeding damage by S. ananas. The dimensions of the brownish discolorations could be associated with the population density at the time of feeding ( Figure 14C).
It is at this point in the plant's development, when the mites are no longer present, that these symptoms are most likely to be noticed by farm personnel in charge of monitoring the plants. Plants exhibiting such damage are not commercially acceptable for export. Leg IV solenidion ϕ length 8-9 8-9 9-10 9-10 Leg IV seta v'G length 11 9-11 14-18 13-16 Leg IV seta l''G length 13-14 16-17 14-17 13-15 Leg IV seta v'F length 6-7 10-12 10-12 7-10 The damage caused by S. ananas on pineapple occurs on growing plants during the development of the inflorescence, fruit and crown (Jeppson et al., 1975; Py et al., 1987; Rohrbach and Johnson 2003, which is similar to the symptoms caused by this organism on Neoregelia. As noted, the pineapple tarsonemid mite has also been associated with Fusarium and also in the pathogenesis of Penicillium funiculosum Thom, (1910) (Rohrbach et al., 1981). Another important factor is the presence of an entomopathogenic fungus, Hirsutella spp. (Umaña et al., 1990; Zoebisch et al., 1992; Quesada-Sojo and Rivera-Mendez 2016, associated with soil and bromeliads in Costa Rica; species of this fungus genus could be used for the control of phytophagous mites like S. ananas that live in concealed places on their host plants. Further studies on the biology of the mite are required for effective monitoring, focused on the plant phenology, as well as control measures, both chemical and biological. In conclusion, we consider Steneotarsonemus ananas sensu Beer (1954) to be the species found on pineapple and several other bromeliaceous plants world-wide, where it causes damage through both its feeding and association with pathogenic fungi. Its taxonomy remains difficult, in part due to the lack of types, and that at least two species are involved: the species originally described by Tryon (1898) and that described by Beer (1954), Jeppson et al. (1975), and also probably by Ewing (1939). However, the presence of other species of Steneotarsonemus on pineapple, and the intraspecific variation in males noted here, suggest the problem requires a much larger revision involving fresh samples from numerous hosts and countries, combining both morphology and molecular methods.