Bystracoccus mataybae Hodgson, Isaias & Oliveira, 2013

Hodgson, Chris, Isaias, Rosy M. S. & Oliveira, D. C., 2013, A new gall-inducing genus and species of Eriococcidae (Hemiptera: Sternorrhyncha: Coccoidea) on Sapindaceae from Brazil, Zootaxa 3734 (3), pp. 317-330 : 318-329

Bystracoccus mataybae Hodgson, Isaias & Oliveira , spec. nov.

Material examined. Holotype adult female: BRAZIL, Uberlândia, Minas Gerais, in leaf gall on Matayba guianensis (Sapindaceae) , 9.xii.2011, D. Oliveira ( UFMG): 1/1 young adult female (good – pharate adult female also on same slide). Paratypes: State of Goias, Morrinhos Municipality, leaf galls, Oct. 2010, Gabriella L. Peres ( BMNH, UFMG, USNM): 9/13 ad ff + 3/24 1st-instar nymphs; also same site, 9.ii.2012, leaf galls, Denis C. Oliveira ( UFMG, USNM): 3/4 ad ff + 2/4 2nd-instar ff + 1/7 1st-instar nymphs. Also: Reserve of Natural Patrimony, Universidade Federal de Uberlândia, Minas Gerais State, in leaf galls on M. guianensis , 1.x.2011, Denis C. Oliveira ( BMNH, USDA): 2/6 ad ff; as previous 9.xi.2011 ( UFMG, USNM): 2/4 ad ff, 3/6 2nd-instar ff + 1/6 1st-instar nymphs; as previous, leaf galls, 20.iv.2012 ( BMNH, UFMG): 3/6 ad ff + 2/8 1st-instar nymphs; same site, leaf galls, 24.v.2012 ( UFMG): 1/2 ad ff; same site, leaf galls: 22.viii.2012 ( UFMG): 1/1 2nd-instar ff; same site, leaf galls, 11.ix.2012 ( UFMG, BMNH): 2/2 2nd-instar ff; same site, leaf galls, 14.ix.2012 ( UFMG): 1/1 2ndinstar ff; same site, leaf galls, 18.ix.2012 ( UFMG, BMNH): 2/6 2nd-instar ff; same site, leaf galls, 20.ix.2012 ( BMNH): 1/5 2nd-instar ff. Also same site but in stem galls, 22.viii.2012 ( USNM): 1/1 1st-instar nymph and June 2013 ( BMNH; UFMG): 2/8 1st-instar nymphs.

ADULT FEMALE ( Figs 1–3 , 6–8 ). Unmounted material. Inducing galls on the leaf of Matayba guianensis (Sapindaceae) , with gall opening on upper (adaxial) leaf surface ( Figs 1, 3 ). Each gall with a funnel-shaped outer chamber and a roundish inner chamber. Gall opening slightly recessed on upper leaf surface but forming a round convex mound on lower leaf surface ( Figs 2 , 6 ); each mature gall 5–6 mm wide. Body of insect flask-shaped with a flat “bottom” and a short “neck”, latter with heavily sclerotised dorsal plate on apex that fills inner cavity opening. Body red (but white when stored in alcohol).

Mounted material ( Figs 7 , 8 ). Total length 0.65–2.03 mm, width 0.5–2.15 mm. Approximately round in dorsal view but margin indented laterally near anterior spiracles and almost flat ventrally. Derm mainly membranous apart from large mediodorsal sclerotised plate, latter 200–350 µm long, 230–380 µm wide. Venter expanding greatly during maturity, probably forming most of derm in oldest specimens.

Dorsum. Segmentation obscure although transverse lines of sclerotisation present anterior to dorsal plate, probably indicating segments. An almost circular, heavily sclerotised plate-like structure present medially, probably composed of all of dorsum of abdominal segments; structure of dorsal plate discussed in detail in Discussion below. Dorsal setae other than those on dorsal plate all spinose, each 7–9 µm long but varying somewhat in basal width; frequent throughout on head and thorax but absent from dorsal plate apart from a pair of spinose setae on margin just laterad to anal opening; dorsal plate with 2 pairs of setose setae towards anterior margin and with 2 rows of pores along each posterolateral margin (see discussion below); other dorsal setae absent. Macrotubular ducts absent. Microtubular ducts also probably absent but frequent laterally near “margin” (probably all ventral). Other dorsal pores absent. Anal lobes apparently incorporated into dorsal plate. Anus present medially near posterior margin of dorsal plate, apparently imbedded under sclerotisation of dorsal plate; anal ring with 3 pairs of small setae but no pores; anus about 35 µm long and 30 µm wide.

Margin. Margin not demarcated and without marginal setae. Eyespots apparently absent.

Venter. Derm membranous apart from around each spiracle which lie in a more or less oval sclerotised plate, and an area of lesser sclerotisation associated with each middle and (particularly) each posterior coxa (possibly even part of posterior coxae); “margin” of mounted mature specimens indented near each anterior spiracle, with a lightly sclerotised radial line extending medially to near prothoracic leg. Setae: setose setae frequent posteriorly on abdomen around vulva, otherwise very sparse; broad-based spinose setae, bigger and broader than those on dorsum, each about 11–13 µm tall and 10–11 µm wide at base, in sparse groups associated with each spiracle and frequent in a band across mesothorax and sparse across metathorax. Macrotubular ducts absent but microtubular ducts abundant throughout; each small, about 6.5 µm long and about 1 µm wide; apparently without a dark, longitudinal inner line. Loculate pores each about 5 µm wide with mainly 5 loculi, in small groups associated with each spiracle; occasionally 1 or 2 also noted between mouthparts and anterior spiracle.

Antennae probably at least 2 or maybe 3 segmented, segments mainly ring-like, total length 23–30 µm; basal segment with a large fleshy seta; apex with at least 3 fleshy setae + some setose setae. Tentorium proportionately large in youngest specimens but appearing small on older, fully-expanded adults; about 160–170 µm long, but becoming embedded in a strongly sclerotised area forming two wings extending anteriorly, each 115–160 µm long; labium probably 2 segmented, with possibly only a single pair of setae near apex. Spiracles each close to “margin” on mounted material, each with a large oval sclerotised plate around peritreme; each plate about 40–43 µm long and 33–38 µm wide; peritremes small, each 16–17 µm wide. Legs complete (5-segmented) but poorly developed; each coxa sclerotised but remaining segments membranous and often somewhat distorted; claw small, about 6.5 µm long; presence of denticle uncertain; tarsal digitules capitate and slightly longer than claw; claw digitules not definitely detected but thought to be short and capitate and longer than claw; translucent pores absent. Vulva present more or less medially between metacoxae.

Comment. In having the dorsal surface of the abdomen heavily sclerotised, Bystracoccus shows some affinities with Aculeococcus Lepage , another eriococcid genus found in South America (and China) ( Hodgson & Miller, 2010). However, on the latter, the sclerotisation also covers the dorsum of the thorax and most of the head as well. In addition the division into head, thorax and abdomen is reasonably clear on Aculeococcus whereas this is obscure on Bystracoccus . Other apparent similarities between adult females of these two genera are: (i) venter becoming much larger than dorsum on mature adults; (ii) legs complete but mainly malformed and obviously nonfunctional; (iii) metacoxae either becoming much enlarged or with a sclerotised area of venter associated with each metacoxa; (iv) absence of macrotubular ducts, and (v) mouthparts with a large pair of apodemes extending anteriorly from tentorial box. Bystracoccus differs from Aculeococcus , however, in the complete absence of multilocular disc-pores on the abdomen, whereas they form broad bands across several of the more posterior segments in the latter species.

As indicated in the Introduction, the sclerotised dorsal plate of the adult female of this species bears a close resemblance to that of Danumococcus parashoreae , currently included in the Beesoniidae . For further comment, see under Discussion below.

SECOND-INSTAR FEMALE ( Figs 9 , 10 ). Unmounted material. Gall as in generic diagnosis. Insect similar to adult female but smaller; body more or less round when young but becoming flask-shaped with a rounded “bottom” and a short “neck” when older. Body red in life.

Mounted material. Total length about 0.5 mm, width 0.46 mm. Derm mainly membranous apart from anal lobes and around anus. Venter expanding during growth, so that dorsum lies medially in oldest specimens.

Dorsum. Derm membranous but body changing shape as nymph matures, with development of a dorsal mound that includes all abdominal segments and which becomes mildly sclerotised; at maturity, dorsum clearly much smaller than venter ( Fig. 10 ). Segmentation obscure but indicated by positions of marginal and dorsal setae, at least on abdomen. Dorsal setae all short and spinose with parallel sides, each 6–7 µm long, present as follows: (i) in a submarginal line, one seta per segment on abdomen but more random anteriorly; (ii) in a submedial line one seta per segment on abdomen, but more random anteriorly, and (iii) with a pair medially just anterior to anal ring, 3 or 4 pairs medially just dorsad to antennae and 2 or 3 pairs more or less dorsad to mouthparts. Macrotubular ducts absent. Microtubular ducts sparsely distributed throughout. Other dorsal pores absent. Anal lobes well developed and quite heavily sclerotised, each with a long apical seta about 80–90 µm long, a stoutly spinose seta on lateral margin, each about 13–15 µm long, plus a smaller spinose seta more anteriorly, about 12 µm long; ventrally, each lobe with a setose seta on inner margin about 12 µm long and a more spinose seta about 10–12 µm long near outer margin; anal lobes joined together dorsally by a small sclerotised dorsal plate, and ventrally by a narrow sclerotised bar. Anal ring about 23 µm long and 21 µm wide, with 3 pairs of setae (each about 25–30 µm long) and an outer ring of pores.

Margin. Margin clearly demarcated by a line of spinose setae, each about 8–9 µm long but most posterior setae slightly larger, about 10 µm long; one per segment on abdominal segments but apparently more frequent on thorax and head. Eyespot present, about 16 µm wide.

Venter. Derm membranous throughout but expanding considerably as second-instar nymph matures, becoming much larger that dorsum. Setae all setose, with particularly long pairs of setae mesad to each meso- and metacoxa, each 45–50 µm long; with 3 pairs of longer setae in a longitudinal line between antennae; and single pairs of setae on about abdominal segments VIII and IX plus 2 pairs on V–VII; shorter setae infrequent but also in a sparse submarginal row. Macrotubular ducts absent. Microtubular ducts small, each about 5.0 µm long and about 1 µm wide, present sparsely throughout. Loculate pores each about 5.5 µm wide with mainly 5–8 loculi; with 2 or 3 laterad to each spiracle.

Antennae 3 segmented, basal 2 segments ring-like, total length 45–50 µm; basal segment with II setose setae, preapical segment with a single setose seta, and apical segment with 3 long stiff setae, 3 fleshy setae + a few setose setae. Tentorium about 60 µm long; labium 2 segmented, with one pair of setae on basal segment and 3 pairs on apical segment. Spiracles small, each peritreme about 8 µm wide; apodeme 16–18 µm long. Legs well developed; length of metathoracic leg (µm): coxae 43–45; trochanter + femur 68–75; tibia 40–42, and tarsus 45; tarsal digitules capitate and subequal in length to claw digitules; claw long and thin, about 16 µm long with a distinct denticle; claw digitules capitate.

FIRST-INSTAR NYMPH (presumably female) ( Fig. 11 ). Unmounted material. Body reddish about 0.25 mm long and oval. Wintering population inducing pit galls on twigs.

Mounted material. Total length about 0.25 mm, width 0.18 mm. Derm membranous apart from anal lobes and around anus. Venter perhaps slightly wider than dorsum.

Dorsum. Segmentation distinct on abdomen. Dorsal setae of 2 types: (i) onion-dome-shaped spinose setae: a pair present on each thoracic segment, each 7.5–9 µm tall and 5.6–6.0 µm wide at base; and (ii) smaller, parallelsided spinose setae, each about 5.5–6.0 µm tall, present as follows: (a) a single pair just anterior to anal area; (b) single pairs submedially on abdominal segments I–III inclusive; (c) medial pairs on abdominal segment I plus 2 pairs dorsad to antennae; and (d) submarginal pairs on each thoracic segment; setose setae absent. Macrotubular ducts absent. Microtubular ducts present as follows: (a) submarginally on abdominal segments VII, V, IV & II, and on each thoracic segment plus 3 pairs on head; and (b) submedial pairs on abdominal segments II and each thoracic segment. Other pore types absent. Anal lobes well developed and quite heavily sclerotised, each with a long apical seta about 40 µm long; 2 large, stout, spinose setae on either side of anal ring, each 15–17 µm long, and a oniondome-shaped seta on outer margin, about 11–12 µm long. Anal lobes joined together dorsally and ventrally by sclerotised bars but dorsal bar probably with a median plate. Anal ring about 16–17 µm long and 15 µm wide, with 3 pairs of setae (each about 25 µm long) and an outer ring of pores.

Margin. Margin clearly demarcated by a line of onion-dome-shaped spinose setae, each about 8–9 µm long; one seta per segment on abdominal segments but more frequent on thorax and head, totalling 20–22 on each side. Eyespot ventral, about 12 µm wide.

Venter. Derm membranous throughout. Setae all setose, with particularly long pairs of setae mesad to each meso- and metacoxa, each 30–33 µm long; with 1 pair of longer setae posteriorly and 3 pairs of very short setae submedially on abdomen, plus 3 pairs of moderately long setae between antennae; submarginal setae present on thorax and abdomen. Microtubular ducts small, with 1 laterad to each procoxa and another laterad to each metacoxa. Loculate pores each about 3.0–3.5 µm wide with 5 loculi; with 1 or 2 laterad to each spiracle. With a pore of unknown homology lateroposteriorly to each metacoxa, each about 3.0–3.5 µm wide.

Antennae 3 segmented, basal 2 segments ring-like, total length 35 µm; basal segment with 2 setose setae, preapical segment with a single setose seta, and apical segment with 3 long stiff setae, one with a capitate apex on many specimens, 4 fleshy setae + a few setose setae. Tentorium about 33 µm long; labium 2 segmented, with one pair of setae on basal segment and 3 pairs on apical segment. Spiracles small, each peritreme about 2.5–3.0 µm wide. Legs well developed; tarsal digitules offset. Each capitate and slightly longer than claw digitules; claw long and thin, about 11–12 µm long with a distinct denticle; claw digitules capitate and longer than claw.

Males: no differences could be detected between the second-instar nymphs and no adult males have been collected and so it is considered that this species is parthenogenetic.

Species name derivation. The species name mataybae , genitive, refers to the generic name of the host plant, Matayba Aubl.

Comments. Other genera of South American Eriococcidae that have onion-dome-shaped setae in the firstinstar nymph are Aculeococcus Lepage and Tectococcus Hempel ( Hodgson & Miller, 2010). The first instar nymph of B. mataybae differs from the latter in having two types of dorsal setae (only one type, in a submedial row, in T. ovatus ) and from that of A. morrisoni Lepage in having only 3 pairs of onion-dome-shaped pores medially whereas A. morrisoni has six. The distribution of the non-onion-dome-shaped setae is also quite different, those on A. morrisoni being restricted to medially on abdominal segments III, V & VI. Of the known first-instar nymphs of South American eriococcids, that of B. mataybae appears to be most similar to that of A. morrisoni .

Discussion. Dorsal plate structure. The structure and development of the sclerotised dorsal plate on the mature adult female is extremely difficult to discuss without reference to the earlier instars, particularly the second instar. The anal lobes on the first-instar nymph ( Fig. 11 ) are fairly well developed and mildly sclerotised, and the anal ring has both setae and pores. However, unlike with later instars, the marginal setae are all onion-domeshaped and segmentally arranged, and the submarginal setae are few, located on abdominal segments I–III.

At the first moult, the young second-instar nymph still has well-sclerotised anal lobes ( Fig. 9 ) and the anal ring still has both setae and pores. However, the onion-dome-shaped marginal spinose setae have now been replaced by more normal, rather blunt, spinose setae (still segmentally arranged) and there are distinct longitudinal lines of rather parallel-sided and blunt submarginal and submedial setae as well. As the second-instar nymph matures ( Fig. 10 ), a roundish membranous area becomes prominent anterior to the anal lobes which eventually become slightly sclerotised, somewhat convex and quite pronounced. This area includes the marginal, submarginal and submedial setae of all of the abdominal segments, with the marginal seta on either side of the anal lobes slightly larger than the rest. The remaining marginal and submarginal setae on the thorax and head come to lie “submarginally” due to the expansion of the venter so that the venter becomes about twice as wide as the dorsum, and the anal lobes plus roundish membranous area is pushed medially. It is assumed that the roundish area relates to the eventual dorsal plate of the adult, which therefore includes the dorsum of all abdominal segments.

In the young adult female ( Fig. 8 ), the dorsal plate is not as heavily sclerotised as when more mature and so the actual structure of the plate is visible. This appears to be rather complicated and the following description may not be entirely accurate! The uppermost surface appears to be fairly smooth and more or less round, with a flat surface, but with a narrow marginal indentation posteriorly, which is probably the anal cleft, and this extends ventrally to the anal ring which has setae but no pores. The anal ring, thus, appears to lie beneath the upper surface of the dorsal plate. The upper surface of the dorsal plate has two pairs of small setose setae towards the anterior margin and two lines of small pores along each posterolateral margin. These lateral lines of pores almost certainly equate to the marginal and submarginal spinose setae on the second-instar nymph because the distribution appears to be identical and so may be setal sockets. In addition, there is a single larger spinose seta on either side of the anal cleft, which almost certainly equates to the larger, posterior-most marginal seta of the second instar. Thus, as suggested above, the dorsal plate does look as though it has been formed by fusion of the dorsums and margins of all of the abdominal segments. However, lying “beneath” the upper surface of the dorsal plate is what appears to be another layer! This layer has a series of large, heavily sclerotised apodemes arising from it which also appear to be segmentally arranged (i.e. there are about seven pairs) and extend more or less laterally. In the mature adult female, these apodemes are very well developed and extend a long way out from beneath the margins of the dorsal plate. This is all reasonably clear. However, in addition, there is another pair of sclerotised areas which extend posteriorly from the anterolateral margins of the dorsal plate ( Fig. 8 , D). These also appear to lie beneath the upper surface but each has a seta. It is difficult to explain how setae might develop beneath the sclerotised plate. On the mature adult female ( Fig. 7 ), almost none of this structure is visible (apart from the apodemes) due to the very heavy sclerotisation of the dorsal plate.

A structure virtually identical to the dorsal plate in B. mataybae is also present on Danumococcus parashoreae (where it was referred to as the “hump” ( Takagi & Hodgson, 2005)). In D. parashoreae , the “hump” is about 100 µm wide, more or less flat, and surrounded by further sclerotised derm and with the anus hidden beneath the posterior margin. In addition, there are 2 pairs of strong internal scleroses associated with the anterior margin, very similar to the 6 or 7 similar structures on B. mataybae , here considered to be apodemes. Indeed the structure of the “hump” and dorsal plate on these two species appears almost identical, although other details, such as the setal distribution, presence/absence of legs, etc., on the adult female and the structure and distribution of the setae on the first-instar nymphs are different. Whilst it is clear, therefore, that these two species are not congeneric, they show very similar adaptations to their galling way of life. Takagi (in Takagi & Hodgson, 2005) speculated that the hump of D. parashoreae was used to plug the gall orifice as a defence against predators and parasitoids. Thus, the two strong scleroses anterior to the hump on D. parashoreae where considered to be for the attachment of muscles for moving the hump so as to allow the male’s aedeagus to reach the vulva, for the elimination of honeydew (from the anus) and dispersal of the crawlers. Precisely the same speculation can be used for the function of the dorsal plate on B. mataybae , but in this case there are six or seven strong scleroses or apodemes Fig. 8 ). Despite this great similarity in the adult females, Bystracoccus is here considered to belong to the Eriococcidae because of the similarity of the firstinstar nymphs to those of other eriococcids, and thus the similarity between the dorsal plates of Danumococcus and Bystracoccus appears to be a good example of convergent evolution.

At least three other eriococcid genera from the Neotropics have a heavy sclerotization on the dorsum ( Hodgson & Miller, 2010; Hodgson et al., 2011). Dromadaricoccus hansoni Hodgson & Miller, off Astronium graveolens (Anacardiaceae) from Costa Rica, has a highly convex but flat topped “hump” or boss that is believed to be formed medially on the metathorax. This species also differs significantly from B. mataybae in having a long narrow abdomen. Another species, Eriogallococcus isaias Hodgson & Magalhães , off Pseudobombax grandiflorum (Malvaceae) from Minas Gerais, Brazil, has the dorsal part of the abdomen sclerotised but it does not form a dorsal plate and is otherwise clearly different, having loculate pores throughout the dorsum (Hodgson et al ,, 2004). Lastly, Aculeococcus Lepage , another eriococcid genus found in South America (and China) has a sclerotised area on the dorsum but this covers the thorax and most of the head as well ( Hodgson & Miller, 2010). In addition, Aculeococcus has broad bands of loculate pores dorsally across several of the more posterior segments. Another species, the Australian Madarococcus megaventris Hardy & Gullan ( Hardy et al, 2008), has a sclerotized dorsum very similar to that of A. morrisoni , but it is clear that B. mataybae is not closely related to either of these genera.

Structure of first-instar nymphs. Two genera of Eriococcidae from South American have onion-dome-shaped setae in the first-instar, Aculeococcus and Tectococcus Hempel ( Hodgson & Miller, 2010). However, the first instar of B. mataybae differs from the latter in having two types of dorsal setae (only one type in a submedial row in T. ovatus ) and from that of A. morrisoni Lepage in having only three pairs of onion-dome-shaped setae medially whereas A. morrisoni has six. The distribution of the non-onion-dome-shaped (but still sharply spinose) setae is also quite different, those on A. morrisoni being restricted to medially on abdominal segments III, V & VI, whereas those on B. mataybae are present sparsely throughout. Onion-dome-shaped dorsal setae are known also in two genera currently included in the Beesoniidae — three Gallococcus species and Mangalorea hopeae ( Raman & Takagi, 1992; Takagi, 2001). These setae are morphologically very similar to those on B. mataybae , although their distribution is quite different.

Whilst the dorsal plates on adult females of the beesoniid D. parashoeae and B. mataybae are very similar, the structure of the first-instar female nymphs is significantly different. Those of B. mataybae have a marginal line of about 46 onion-dome-shaped setae ( Fig. 11 ) whereas these are entirely absent on D. parashoreae . In addition, the crawlers of B. mataybae have a well-developed anal ring with setae and pores, surrounded by sclerotised plate-like anal lobes, also with onion-dome-shaped setae. These are entirely absent on D. parashoreae .

The female first-instar nymphs of these beesoniids also have dorsal microducts very similar to those in B. mataybae and many other “eriococcids”. It is clear that the phylogenetic relationship of these two taxa is close. As indicated in the Introduction, it has become clear that the “eriococcids” are non-monophyletic and fall into several clades ( Hodgson, 2001; Cook et al, 2002; Cook & Gullan, 2004; Gullan & Cook, 2007; Hodgson & Hardy, 2013). Without the added morphological data from adult males of B. mataybae , it has proved very difficult to place this species in a family and its placement here within the Eriococcidae is tentative bearing in mind the close similarity of its dorsal plate to that of D. parashoreae .

Biology and ecology. Matayba guianensis Aubl. (Sapindaceae) is a deciduous shrub that occurs in Central and South America, from Trinidad and Tobago in the north to Rio Grande do Sul, Brazil, in the south and is found in all forest types except semi-deciduous woodlands. However, M. guianensis is most abundant in riparian forests and cerrado ( Sano et al., 2008). It has compound leaves that emerge between the end of August and October and it flowers from September to November, with mature fruits in February.

At the Ecological Station of Panga, Uberlândia municipality, Minas Gerais State, Brazil, M. guianensis is heavily infested by the leaf galls of B. mataybae (72% of the leaves) ( Figs 1–3 ). Gall induction takes place in the youngest leaves, most commonly in September and October. The leaf galls are light green with red spots, are glabrous and intralaminar, with a narrow projection through which the gall opens on the adaxial surface. The galls are all similar in shape and size, each with two chambers ( Fig. 6 ), an outer chamber which appears to be empty and an inner chamber in which the insect lives. There is just one insect in each gall. Heavy infestations, such as those noted at Panga, i.e. leaflets with 5 or more galls, cause a significant reduction of about 50% in leaf area (mean area of non-galled leaflets 6.61 cm 2, galled leaflets with more than 5 galls 3.2 cm 2) (N = 50). However, the number of generations per year is unclear.

Small galls are also abundant on the twigs of M. guianensis during the dry season (winter) (June–August). These are pit galls ( Gullan et al., 2005), about 2 mm across, each with a single first-instar nymph in the centre ( Figs 4, 5 ); when very abundant, the galls tend to fuse ( Fig. 5 ). These nymphs appear to be morphologically identical to those from the leaf galls and are considered to be female and so no male stages are currently known. It is assumed that these twig nymphs remain in the first instar during the winter (dry season), after which they moult at bud-burst and disperse as second-instar nymphs to the young leaves in August and September, when only the second instar is known. Although adult females are found from October onwards, it is not certain how many more generations there are except that clearly crawlers must be produced in the autumn (May–June) when adult females with eggs have been noted. These then disperse from the senescing leaves onto the twigs. This is the second species of Eriococcidae in Brazil diapausing through the dry season in stem galls, as this strategy is also known for Pseudotectococcus rolliniae Hodgson & Gonçalves ( Gonçalves et al., 2009).

Figures

FIGURE 6. Cross section through galls showing double chamber. A = young gall; B = older gall.

FIGURE 7. Old, fully-expanded adult female B. mataybae. Unless otherwise started, on this and subsequent figures A = dorsal spinose setae; B = dorsal setose setae; C = dorsal microtubular duct; D = “ventral” view of dorsal plate; E = ventral spinose seta; F = posterior leg; G = posterior spiracle; H = ventral microtubular duct; J = 5-locular disc pore; K = ventral setose seta, and L = antenna.

FIGURE 8. Young adult female of B. maytabae prior to full maturity. For letters, see Fig. 7. Note that D is the dorsal plate as seen from underneath.

FIGURE 9. Young second-instar female of B. mataybae, prior to the commencement of dorsal plate development. For letters, see Fig. 7, but where D1 and D2 = dorsal and ventral views of anal area.

FIGURE 10. Mature second-instar female of B. mataybae, where the abdominal segments have started to form the dorsal plate.

FIGURE 11. First-instar nymph of B. mataybae. For letters, see Fig. 7, but where D = anal ring; M = onion-dome-shaped marginal setae; N = labium and P = pore of unknown homology.

FIGURES 1–5. (1) Upper surface of leaf of Matayba guianensis with young galls of Bystracoccus mataybae Hodgson, Isaias & Oliveira, showing outer gall orifice. (2) Lower surface of leaf of M. guianensis with fairly well-developed galls of B. mataybae. (3) Upper leaf surface of M. guianensis with old galls of B. mataybae. (4) Pit gall on twig of M. guianensis with first-instar nymph of B. mataybae in centre. (5) Twig of M. guianensis with many pit galls induced by B. mataybae, some fused together, many with nymphs in their centres. [Photographs D. Oliveira].