Another puzzle piece in the systematics of the chewing louse genus Myrsidea, with a description of a new genus Apomyrsidea

A new avian chewing louse genus Apomyrsidea gen. nov. is described based on species parasitizing birds in the family Formicariidae. Diagnostic characteristics and phylogenetic analyses were used to evaluate and confirm the generic status and merit its recognition as unique and different from Myrsidea Waterston, 1915. Three species previously belonging to the genus Myrsidea are placed in the new genus Apomyrsidea gen. nov. and are discussed: Apomyrsidea circumsternata (Valim & Weckstein, 2013) gen. et comb. nov., Apomyrsidea isacantha (Valim & Weckstein, 2013) gen. et comb. nov. and Apomyrsidea klimesi (Sychra in Sychra et al., 2006) gen. et comb. nov.


Introduction
Chewing lice (Psocodea Hennig, 1966: Phthiraptera Haeckel, 1896 are obligate permanent ectoparasites, meaning they live their entire life cycle on their hosts (Clayton et al. 2015). Many are microhabitat specialists, often occupying specifi c body parts on the host (Johnson et al. 2012). The species parasitizing birds feed on feathers, dead skin, skin secretions, and in some cases blood . At high prevalence, they can cause signifi cant harm to feathers, skin irritation and even possible secondary infections that decrease the fi tness of the bird hosts (Møller et al. 1990;Mullen & Durden 2002).
systematics of this diverse group (Johnson et al. 2004;Smith et al. 2011). The chewing lice genus Myrsidea Waterston, 1915 is among the most diverse and host specifi c genera within the parasitic lice (Phthiraptera), with more than 380 species described worldwide (Kolencik & Sychra, unpublished data). While there are 49 species of Myrsidea described from Neotropical suboscine birds (Kolencik et al. 2018), only three have been described from birds of the family Formicariidae Gray, 1840 (Sychra et al. 2006;Valim & Weckstein 2013). These three species show a set of characteristics that are unique among all species of Myrsidea to the extent that Sychra et al. (2006) and Valim & Weckstein (2013) suggested that they may be placed outside of Myrsidea. Here, we evaluate both morphological and molecular data of these three species along with representatives of the major clades in Myrsidea. We fi nd that the combination of differences found from both morphological and molecular data is strong enough to merit their recognition in a separate genus here named as Apomyrsidea gen. nov.

Morphology
All the morphological descriptions and characters as well as the terminology of chaetotaxy were taken from or follow those from Clay (1961Clay ( , 1962Clay ( , 1966Clay ( , 1969, Sychra et al. (2006) and Valim & Weckstein (2013). Taxonomy and nomenclature of the birds follows IOC Bird World List ver. 10.1 (Gill et al. 2020).

Phylogenetic reconstruction
For the phylogenetic reconstruction we used a fragment of the mitochondrial gene cytochrome oxidase I (COI, 379bp). We amplifi ed and sequenced this fragment from Apomyrsidea klimesi (Sychra in Sychra et al., 2006) (3 nymphs ex Formicarius analis (d'Orbigny & Lafresnaye, 1837) from Zona Protectora Las Tablas on the Pacifi c slope of the Cordillera de Talamanca, 8°54′ N, 82°47′ W; 1300 m a.s.l.; identical sequences; GenBank accession number: MW381016) using the technique described by Martinu et al. (2015). All other sequences were downloaded from GenBank (Table 1) and aligned in SeaView ver. 4.7 (Gouy et al. 2010) using Clustal Omega (Sievers et al. 2011) and checked visually. In total, we aligned 50 sequences from seven chewing lice genera from the suborder Amblycera Kellogg, 1896 and two sequences from the suborder Ischnocera Kellogg, 1896 as the outgroups. To determine the best model fi t to our data we used PartitionFinder ver. 2.1.1 (PF; Lanfear et al. 2016) and compared all models using the corrected Akaike Information Criterion (AICc). We selected two different approaches for a phylogenetic reconstruction to examine the similarity among the methods. First, the randomly accelerated maximum likelihood (RAxML ver. 8.2.12; Stamatakis 2014) method was used with three partitions and the model GTR + I + G for each and estimated 1000 bootstrap trees. Second, MrBayes ver. 3.2 (Ronquist & Huelsenbeck 2003) was used for Bayesian analysis with GTR + I + G model for two partitions and HKY + G for the third. We conducted two parallel runs for 20 000 000 generations with four Markov chains (Huelsenbeck & Bollback 2001), which were sampled every 1000 generations with a total of 20 000 parameter points estimates. Markov chains were examined in Tracer ver. 1.7.1 (Rambaut et al. 2018) to determine that the chains had reached stationarity and 10% of trees were removed as burnin. A 50% majority rule consensus tree with posterior probabilities was generated. The outgroup ischnoceran genus Philopterus Nitzsch, 1818 was used for rooting in both trees.

Results
Here, we fi nd that species of Apomyrsidea gen. nov. are unique and morphologically distinct from all species of Myrsidea. Specifi cally, the simple combination of the presence of the posterior dorsal head seta (dhs) 23 and the dorso-central pronotal setae (dps) 2 ( Fig. 3), sternite I lying inside the wide notch of sternite II (Fig. 5) and different type of fusion in the terminal sternites -male abdominal sternite VIII and female sternite VII both at least partially but conspicuously separated from the subgenital plate, which is formed by a single sternite IX (in the male; Fig. 2) or fusion of sternites VIII and IX (in the female; Fig. 1), make these species distinct. These characters were further defi ned in Clay (1966), Sychra et al. (2006) and Valim & Weckstein (2013).
This separation is also supported with the present phylogenetic analyses of 379bp fragment of COI from 50 amblyceran species across 6 genera:

Diagnosis
Apomyrsidea can be characterized and distinguished from all other menoponid chewing lice genera with the combination of following characteristics: Head -rounded anteriorly, lacking lateral slit or notch, without sclerotized processes (oral spines) arising near the base of maxillary palpi (Figs 1-3); Figs 1-2. Dorso-ventral view of Apomyrsidea klimesi (Sychra in Sychra et al., 2006) . 3); -dhs 18 (outer mid-dorsal head seta by Clay (1966) or dorsal head seta "d" by Clay (1962) is missing; -dhs 22 (outer occipital seta by Valim & Weckstein 2013 or posterior dorsal head seta "f" by Clay 1962) approximately as long as dhs 21 (inner occipital seta by Valim & Weckstein 2013), both surpassing pronotal carina; -dhs 23 (posterior dorsal head seta "e" in Clay 1962) present and anteriorly to line of bases of dhs 21 and 22 (Fig. 3); -head sensilla 3-5 sensu Clay (1961) or c-e sensu Clay (1969) absent; -characteristic gular plate with the greater length and thickness of the posterior pair of setae compared to the rest ones (Figs 1-2). Thorax -prosternal plate well developed with straight anterior margin and two anterior setae (Fig. 4); -pronotum with one pair of minute dorso-central pronotal setae lying near the transverse carina (dps 2 by Clay 1962) (Fig. 3); -pronotum with anterolateral pronotal setae (marginal prothoracic setae 1-3, mps 1-3 in Clay 1962) in following arrangement: mps 1 and mps 3 spine-like, mps 2 fi ne and long; mps 1 and mps 2 located on each lateral corner of pronotum, with mps 3 posteriorly to mps 2 on pronotal margin (Fig. 3); -mesonotum well defi ned with only two anterior setae; -mesonotum without median division, but with a Y-shaped line just below postnotum, not forming a suture or even splitting mesonotum (Fig. 3). This Y-shaped line slightly less evident, but also discernible in good specimens from other host families; -strongly sclerotized ring-like mesothorax -mesothorax with sternum, pleura and tergum fused to form strongly sclerotized ring round the body; -femur III without combs of spine-like setae but with thick or sparse brushes of setae.

Abdomen
-sternite I mostly surrounded by sternite II (it lies inside the wide notch of sternite II) (Fig. 5); -sternite II enlarged with a clutch of heavy spine-like setae at each posterior-lateral margin called aster; -male genitalia as in Fig. 6. -male genital sac sclerite with two roughly serrated spiculated lateral arms (Figs 7-9); -female vulva with smooth posterior margin (Fig. 1); -female ventral anal margin without lateral seta-bearing processes (see Clay 1969); -sternite VII fused with VIII + IX + X, forming female subgenital plate, although with a distinct transverse fenestra distinctly enclosed at lateral sides of subgenital plate where seventh and eighth segments fused (in A. circumsternata and A. isacantha; see Valim & Weckstein 2013: fi g. 11); in the case of A. klimesimale abdominal sternite VIII and female sternite VII both separated from the subgenital plate, which is formed by a single sternite IX (in the male) or fusion of sternites VIII and IX (in the female) (Figs 1-2).

Etymology
The generic name Apomyrsidea is formed by a combination of Greek word 'Apo' = 'from' and Myrsidea, referring that it is separated from the genus Myrsidea, where it was originally placed. The gender is feminine.

Discussion
In the key to the genera of the Menoponidae Mjöberg, 1910 of Clay (1969), Apomyrsidea gen. nov. would key to couplet 33 being closest to Myrsidea and Oculomenopon Price & Hellenthal, 2005. All three genera share unique diagnostic characters including: characteristic gular plate with the greater length and thickness of the posterior pair of setae compared to the other ones; missing dhs 18; well-developed prosternal plate with two anterior setae; strongly sclerotized ring-like mesothorax -mesothorax with sternum, pleura and tergum fused to form strongly sclerotized ring round the body; and general shape of male genitalia. Moreover, enlarged sternite II with aster of heavy setae at each posterior corner is a common feature of Apomyrsidea gen. nov. and Myrsidea, while presence of dhs 23 and the similar size of dhs 21 and dhs 22 is shared between Apomyrsidea gen. nov. and Oculomenopon (Price & Hellenthal 2005).
The most important key characters useful for separating Apomyrsidea gen. nov. from all species of Myrsidea and Oculomenopon are: the presence of dps 2, sternite I lying inside the wide notch of sternite II and different type of fusion in the terminal sternites -male abdominal sternite VIII and female sternite VII both at least partially but conspicuously separated from the subgenital plate, which is formed by a single sternite IX (in the male) or fusion of sternites VIII and IX (in the female).
Here, we provide a key for these three genera that modifi es and extends the fi rst part of couplet 33 of the Clay's (1969)  Valim & Weckstein (2013) also discussed the presence of dps 2 and stated that, according to Clay (1966), the absence of dorso-central pronotal setae (dps 1 and dps 2) is one of defi ning characteristics of the genus Myrsidea. They wrongly interpreted Clay (1969), when writing: " Clay (1969) presented data showing that some species of Myrsidea could have at least one pair of dorso-central pronotal setae (see couplet 32)". When we checked couplet 32 in the key by Clay (1969), there is a note "not more than one pair of dorso-central pronotal setae". Moreover, Clay (1969: 11) wrote: "There are usually two pairs of dorso-central pronotal setae (dps) lying on or near the transverse carina, but in some species-groups (Clay 1962: 237) or genera (Myrsidea) they are reduced to one pair or absent." We have examined many slides across groups and species of Myrsidea and can confi rm that dps are absent apart from the newly erected Apomyrsidea.
Another interesting character is the presence of a cluster of heavy setae on each side of sternite III (Fig. 5). These setae are not as long as the aster, but they are heavier than spine-like setae on the lateral sides of other sternites and they are more conspicuous on females than on males. Despite this, it cannot be used as a basic character of the genus, because it is present only in two species -A. circumsternata gen. et comb. nov. and A. klimesi gen. et comb. nov. An unusual cluster of three spine-like setae situated on one side of sternite III was described by Klockenhoff (1984) from only one female of M. serini (Séguy, 1944 Price et al. 2004). However, this partial division is noticeable only at the lateral margins of the sternites of these species. A medial division or totally divided terminal sternites are only found in species of Apomyrsidea. According to Clay (1969), this is not necessarily a generic character, but in our opinion, it can be useful for separating Apomyrsidea gen. nov. from Myrsidea.
Another interesting character is the length of dhs 21 and dhs 22. While Clay (1966) considered the short length of dhs 22 compared with the long dhs 21 as one of the determining characteristics for the genus Myrsidea, Apomyrsidea gen. nov. shows dhs 22 as long as dhs 21 and both exceed the pronotal carina. Interestingly, the enigmatic genus Ramphasticola Carriker, 1949, which has been diffi cult to place, with different authors suggesting it should be nested within Myrsidea (Hopkins & Clay 1952;Price et al. 2003), has both species with long and short dhs 22, further suggesting a molecular phylogenetic analysis is needed to validate the placement of this genus.
The presence of an aster of heavy setae at each posterior corner of sternite II in Apomyrsidea gen. nov. and Myrsidea may bring some new challenges to determinations. The aster is one of the most visible features of many species of Myrsidea, and easily seen even under a stereoscope. However, in some species of Myrsidea it can be reduced (e.g., some species of Ramphastidae), while in species of Apomyrsidea gen. nov. it is always present. This changes the defi ning characteristic of the aster as the determining morphological feature for Myrsidea and broadens the use of this characteristic to a second genus. Henceforth, if no aster is present in a louse, this does not necessarily mean that it is not a Myrsidea, but the presence of asters will narrow the selection to two genera: Myrsidea and Apomyrsidea gen. nov. This further supports this character as an important one for morphological identifi cations.
Beside unique morphological characteristics, Valim & Weckstein (2013) referred to these species as "Myrsidea from Formicariidae" and suggested they would be phylogenetically distinct from nearly all other species of Myrsidea. Therein, instead of describing the new genus, the authors decided to broaden the set of diagnostic characters used to defi ne the genus Myrsidea. However, in their phylogenetic tree Myrsidea sp. (GenBank KF048123) from Myiarchus panamensis Lawrence, 1861 (the avian family Tyrannidae Vigors, 1825), appears to be in the same clade with both species of Apomyrsidea gen. nov. used in the analysis (Valim & Weckstein 2013). This may be a misidentifi cation of the sample and needs validation. Further, when this sequence was compared to all other COI sequences (a 379bp fragment of mitochondrial gene cytochrome oxidase subunit I) in GenBank, the highest match was Menacanthus sp. (GenBank AF545726) from the tyrannid Attila spadiceus (Gmelin J.F., 1789), with a p-distance of only 0.79%, and Menacanthus sp. (GenBank KJ730539) from the parid Poecile sp. with a p-distance of 1.31%. A comparison between these two species of Menacanthus showed the p-distance to be 1.05%, which is likely low enough to confi rm it as the same species. On the basis of these facts, we believe that sequence KF048123 is potentially an untrustful sequence and we excluded it from our analysis.
When we used a broader range of genera in our study, Apomyrsidea gen. nov. shows a similar pattern creating a well separated clade from all other species of Myrsidea (Fig. 10). The familiar issue with lower support values (mostly in basal nodes) is often present when using only a single gene fragment (e.g., Kolencik et al. 2017). Unfortunately, the sequences, and their quality, for amblyceran species submitted to the GenBank database are limited and mostly only for a 379bp fragment of COI and/ or a 347bp fragment of nuclear EF-1a (elongation factor 1 alpha) gene. Besides that, in many cases both sequences for the same specimen did not overlap; moreover, all three species of Apomyrsidea  (2021) comb. nov., in accordance with their morphology, which was also used in the key for their identifi cation by Valim & Weckstein (2013).
In conclusion, the combination of unique morphological characteristics and the well-separated clade in the phylogenetic tree are strong enough to confi rm Apomyrsidea gen. nov. as a new genus, and a sister taxon to Myrsidea.