El género Herpisticus Germar, 1823 de las islas Canarias (Coleoptera: Curculionidae: Entiminae: Tanymecini) ; The genus Herpisticus Germar, 1823 from the Canary Islands (Coleoptera: Curculionidae: Entiminae: Tanymecini)

The taxonomic review of the genus Herpisticus Germar, 1823, endemic to the Canary Islands, is undertaken with support of molecular data (mtCOI). The morphological study includes the description of the larva and pupa of this genus. To the 5 previously known species, 20 taxa have been added, comprising the resurrection of H. eremita var. γ lanatus Wollaston, 1864 as valid species, H. grancanariensis Palm 1974 as valid subspecies of H. subvestitus Wollaston, 1864, and the following new taxa: H. famarae n. sp. from Lanzarote; H. betancuriae n. sp., H. jandiensis n. sp. and H. rectipes n. sp. from Fuerteventura; H. subvestitus pseudolanatus n. ssp., H. guanarteme n. sp., H. denudatus n. sp., H. gigas n. sp., H. guayarmina n. sp., H. tasarticus n. sp., H. scopulus n. sp., and H. nanus n. sp. from Gran Canaria; H. daute n. sp. and H. aridicola n. sp. from Tenerife; H. gomerensis n. sp., H. hispidus n. sp. and H. bobadillae n. sp. from La Gomera; and H. hierrensis benahoare n. ssp. from La Palma. Except for the subspecies H. hierrensis hierrensis, which is present on the islands of El Hierro and La Palma, all the other taxa are mono-insular endemics. Identification keys (♂), distribution maps, and photographs of imagos of all 25 taxa are provided, as well as some general comments on the biology of the group, its potential as agricultural pest, and conservation perspectives. urn:lsid:zoobank.org:pub:6DA1115C-9E4C-4E28-A323-E51C14B7D1F3


Introduction
Herpisticus weevils are large Tanymecini endemic to the Canary Islands, where they live in the open dry or semi-dry habitats of the lowlands and mountains, avoiding the shady and humid forest environments. The first known species was described from the island of Tenerife as Curculio eremita by Olivier (1807), without noticing that the same name was already used by Herbst (1784) for a species from Germany, at present considered as nomen dubium (f. Alonso-Zarazaga et al., 2017). Germar (1823) established the genus Herpisticus with type species H. laesicollis from Tenerife (Fig. 1), which was considered a junior synonym of H. eremita Olivier by most authors (e.g. Winkler, 1932; despite Schoenherr (1833) having already pointed to the previous homonymy of C. eremita and the validity of the name H. laesicollis.
T. Vernon Wollaston (1864), who established the foundations of the Canarian Coleopteran fauna (f. Machado, 2006), described two further species from the eastern islands Lanzarote and Fuerteventura: H. oculatus and H. calvus. This British entomologist was much intrigued by the variation of "H. eremita", which inhabited the other islands, and particularly those specimens from the island of Gran Canaria showing great geographical variation in body pilosity. On that island he recognised a var. α typical, a var. ß subvestita, and a var. γ lanata, but he stated that all transitional stages between them could be found, and that "it is impossible therefore to look upon them as more than topographical varieties brought about by surrounding circumstances and the more or less calcareous nature of the regions in which they occur." In the mid 20th century, several authors Lindberg & Lindberg, 1958, etc.) recorded Herpisticus profusely and commented also on its variation, not being fully convinced about the identification of the taxa sampled. Indeed,  considered H. calvus and H. oculatus as extreme cases within the great variation of H. eremita; two synonyms proposed that have been basically overlooked or misinterpreted. Palm (1974), who provided the first identification key for the genus, studied its variation and noticed that specimens of "H. eremita" from Gran Canaria were clearly different from those of Tenerife in the shape of their aedeagus, and he described H. grancanariensis stating also the morphological transition between localities, as Wollaston did (op. cit.). However, he overlooked that the name subvestitus Wollaston, 1864 (type locality Tirajana) was already available for specimens from Gran Canaria (Palm`s type is from Tafira), thus  creating a new synonym that was later noticed and registered by , and followed by later authors.
The Austrian entomologist Herbert Franz (1979) described H. hierrensis based on a single female from Las Playas, on the island of El Hierro, recording at the same time the presence of "H. eremita" in the north of the island. However, in his final compilation of the coleopteran fauna of this island (Franz, 1996) he assumes that all Herpisticus from El Hierro -including also those recorded by belong to H. hierrensis, its type being erroneously referred to as male.  designated lectotypes of Wollaston's taxa and highlighted the urgent need to revise this endemic genus in this and other publications on the curculionid fauna of the islands , 2015, particularly after comparing a few COI sequences from specimens originating from Gran Canaria, Tenerife, and La Gomera. Previously, Schütte et al. (2013) included these sequences in a large ambitious molecular analysis of Curculionoidea, where they cluster with Sitonini and a few Otiorhynchus. However, this was in a node with a Bayesian posterior probability of 56, which is absolutely non-significant (the whole phylogram has low basal resolution for Curculionidae clades). The filiation of Herpisticus with other Tanymecini remains unresolved.
The last published taxonomic framework for the genus Herpisticus is provided by Alonso-Zarazaga et al. (2017), which is identical to that included in the «Catalogue of Palearctic Coleoptera» .
H. oculatus Wollaston, 1864 H. calvus Wollaston, 1864 H. subvestitus Wollaston, 1864 = grancanariensis Palm, 1974= lanatus Wollaston, 1864 H. laesicollis Germar, 1823= eremita Olivier, 1807 (Curculio) H. hierrensis Franz, 1979  The island distribution shown here has been taken from , with the addition of some doubtful records pending confirmation. In fact, the knowledge of the genus, both from the taxonomic and chorological point of view, is far from satisfactory. Therefore, the aim of the present contribution is to revise both aspects of this difficult genus with the concurrence of molecular data obtained from DNA extracted from specimens collected in all islands.

Material and methods
Material exaMined. This study is based upon examination of 2782 specimens collected by the first author or borrowed from several institutions and colleagues. The majority of specimens of Herpisticus in the MNCN (Madrid) do not bear collector names. However, in accordance with the collecting dates or handwriting, they have been attributed to M. Anatael Cabrera, who donated his collection to the MNCN, and others to M. Manuel Martínez de la Escalera (leg. M. Escalera) if the dates fall between December 1920 to June 1921 (see Izquierdo, 2011).
The following island symbols are used (e.g. last digit in voucher numbers or as prefix) to denote island origin of the specimens or of the locality referred to: C = Gran Canaria F = Fuerteventura G = La Gomera H = El Hierro L = Lanzarote P = La Palma T = Tenerife Colour codes used for the same purpose in the phylogram are explained in its legend. The map of the Canary Islands in Fig. 2 shows for each island its age (Ma) in blue (Carracedo & Troll, 2016) and the final number of Herpisticus species in black.
In the list of synonyms and references that follow each taxon, references to figures have been included except when hyperlinks to the same figures appear repeated in several works of the same author (e.g. P. Stüben) as collateral information.
The distribution maps of Herpisticus species have been prepared with the software QGIS V. 3.4.6-Madeira, using the base topographic map 1:20,000 provided by the Canary Islands Spatial Data Infrastructure (www.idecanarias.es/listado_servicios), as well as the geological maps and Canarian network of protected areas used for overlap analysis.
Morphology. Dissections were done according to standard entomological techniques. For the evertion of the endophallus the method of Van Dam (2014) was followed, after maceration of the penis in pancreatin (Alvarez-Padilla & Hormiga, 2008). Photographs were taken with a Canon EOS 6D digital camera equipped with macro-lens MPE65. The program Zerene Stacker was used for stacking the digital photos and Photoshop Elements for final retouching. Drawings were made using a camera lucida attached to a Leitz microscope or to an Olympus SZX12 stereo microscope, and measurements were taken with an eyepiece micrometer. Symbols L, W and H used in ratios (e.g. L/W) refer to length, width and height, respectively. Body size is expressed as length of specimen × maximum width of elytra, in millimetres. Length of specimens is measured without rostrum (s.r.) in dorsal view, from the anterior margin of eyes to apex of elytra. Length of antennal club is measured including the attached desmomere 7. Rostral width was measured perpendicular to its dorsum across the straight-line tangent to the anterior margin of eyes (base of rostrum), and rostral length from this line to its apex (mandibles excluded). Eye convexity is expressed as the percentage of a theoretical complete ellipsoid or globe emerging from the profile of the head capsule; thus, a 50% convexity would mean a hemispherical protruding eye, a 10% a fairly flat eye, etc. It is obtained by observing the eye in profileperfectly tangent to the head capsule-and dividing its maximum height by its length. Length of elytra is measured from their apex to the base of scutellum (tangent with a line uniting both humeri). Abdominal transverse convexity (H/W) is determined by dividing the maximum height between elytra and ventrites in lateral view by the maximum width of the elytra measured in dorsal view (both measurements at metacoxal level). The length of the hair is considered from its base to its most distant part, compared with the length of a tarsal claw as reference. Apex of the penis is considered from the level of the distal margin of the ostium to the distal tip of the tube.
A detailed redescription of the genus is provided, while for each species a brief description suffices, centred in the diagnostical characters. These descriptions are based on males, with comments on the female at the end. The morphological terms used follow Marshall (1916) and Machado (2010).
The description of the preimaginal stages is based on a few larvae obtained in the field and one pupa (reared) of Herpisticus subvestitus pseudolanatus n. ssp. For drawing purposes, the pupa has been stretched slightly and the left pterotheca separated from the body. Terms and abbreviations for the setae in larvae follow Marvaldi (1997Marvaldi ( , 1999b and for pupae Gosik & Sprick (2013). Numbers of paired setae are given with a vertical bar separating the set of each side (e.g. 7|7) and an addition in brackets may reflect the clusters (e.g. 2+2) in one set.
Larval chaetotaxy: als = anterolateral setae (epipharynx) ams = anteromedian setae (epipharynx) as = alar setae  Molecular analysis. DNA was extracted from one leg of ethanol-preserved or just killed specimens using Chelex extraction protocol (Casquet et al., 2012). A fragment of the mitochondrial gene cytochrome oxidase I (COI) was amplified and sequenced following López et al. (2007) with primers and polymerase chain reaction conditions indicated therein. Sequencing was performed by the Sanger DNA sequencing service of Macrogen Europe in Madrid. Sequences were viewed, edited and assembled using MEGA7 software (Kumar et al., 2016). Alignments were achieved using the program Muscle (Edgar, 2004), with default parameters as implemented in MEGA 7 and adjusted by eye. The plausibility of the alignment was verified at the amino acid level. The Xia et al., test (Xia & Lemey, 2009) as implemented in DAMBE7 (Xia, 2018) was used to assess substitution saturation within the sequences, with negative results. K2P distances were calculated with MEGA7. Phylogenetic relationships were reconstructed using Bayesian Markov chain Monte Carlo inference (Yang & Rannala, 1997) as implemented with MrBayes 3.2.3 (Ronquist et al., 2012) and applying default parameters. The TIM1+I+G model obtained with jModelTest 2.1.4 (Darriba et al., 2012) as the best nucleotide substitution model for our dataset was considered in MrBayes with the command lset nst = 6 rates = invgamma. Full convergence was tested with Tracer v1.6 (Rambaut et al., 2014). The BI final phylogram was edited with TreeGraph 2 (Stöver & Müller, 2010) and nodes supported with less than 0.90 Bayesian posterior probabilities (bpp) have been collapsed. Our set of Herpisticus sequences (718 bp) are registered in GenBank (www.ncbi.nlm.nih.gov/Genbank) with accession numbers: MN432522-MN432597 and MN709122. Ten additional sequences (labelled with prefix-PST) obtained under "The Molecular Weevil Identification Project" (Schütte et al., 2013), were generously provided by Dr. P. Stüben and A. Schütte. Voucher codes preceded by DNA (highlighted in red) and followed by their Genbank accession numbers are included in the Material Examined section of each taxon. All Herpisticus species are present in the dataset, except H. gigas n. sp. (DNA extraction from museum specimens failed). Some redundant sequences have been discarded from the final dataset, which finally totalled 84 OTUs.
Sitona lineellus (Bonsdorff, 1785) from Finland (downloaded from GenBank KJ962739.1) was finally used as outgroup (genetic K2P distance to Herpisticus 21.0-27.3%). We tested a Molybdotus vermiculosus (C.O. Waterhouse, 1881) a Tanymecini from Yemen (Aloove area, Hassan vill. env., 221 m, 10-11-2010, leg. P. Hlaváč;GenBank MN432595) and Coniatus tamarisci (Fabricius, 1787) from Gran Canaria (La Aldea, El Charco 5 m. 26-4-2019 leg. A. Machado; GenBank MN432594), but they were discarded due to long branch attraction effects, possibly related with their greater genetic distance (24.3 -36.5% and 26.2 -34.9%, respectively). species deliMitation. Tanymecini weevils are known for their morphological variability, which obviously does not facilitate species delimitation under morphological criteria. Therefore, a fragment with 617 base-pairs from the mitochondrial cytochrome c oxidase 1 gene (COI) -widely used in the DNA barcoding initiative (Hebert et al., 2003)-has been chosen to assist in unveiling the species delimitation within Herpisticus. Pentinsaari (2016) found an average 11.99% K2P distance between nearest neighbours in a dataset of 1872 coleopteran species, which dropped down to 10.63% when 363 species were added ( values in other arthropod groups like Lepidoptera were much lower: 5.73%). The latter threshold of 10.63% was initially used to postulate species boundaries in our study, combined with the morphological and geographical analysis. coMMon naMes. Spanish official common names of Herpisticus species have been taken from Machado & Morera (2005), but as there are many more species recognised according to the present study, new common names are proposed following the same naming criteria adopted in the referred work, which include several amendments to lend coherence to the complete set. It is convenient to have vernacular names for weevil species that are large and visible, and may become agricultural pests or be of conservation interest. plant species. Plant species full names are listed in family alphabetical order in the 'Additional remarks' chapter, section of 'Food plants'.

Genetic results
Between group and within group mean K2P distances are compiled in Appendix 1. The phylogram obtained is shown in Fig. 3, with nodes below ppb 0.90 collapsed. Resolution is good at leaf level and almost non-informative at root level. To infer the basal phylogeny of this genus other more conservative markers should be added to the dataset, but this is beyond the purpose of this study focused on taxonomy.
Without further basal resolution, the 83 OTUs cluster first into seven clades that converge at the basal node G, covering clades A, B, C, and D species from the eastern islands, clade E and F two species from Gran Canaria, and clade K -with eight subclades-the rest of species from Gran Canaria and the western islands. For reference purposes, clades have been ordered and labelled from A to T according to the decreasing age of the islands (Lanzarote-Fuerteventura > Gran Canaria > Tenerife > La Gomera > La Palma > El Hierro) as it is likely that with more resolution they will reflect a similar step-wise colonisation pattern from East to West, as stated for several arthropod groups and particularly for other entimine weevils like Cryptorhynchinae (Stüben & Astrin, 2010) or Laparocerus .
The overall mean K2P distance in the dataset is 12.4% and between the recognised groups 13.2%, suggesting an old origin for the genus Herpisticus.
Considering the p-distance and applying the general ratio of 2.6% per My for protein-coding mtDNA (Pons et al., 2010) to the most divergent OTUs (15.86%) between Herpisticus calvus from Caleta de Fuste in Fuerteventura and H. hispidus n. sp. from Benchijigua in La Gomera), a rough age estimate of 6.1 My is obtained for the initial radiation (or colonisation) represented by the basal node G.
The between-group mean distances of most clades are well above the 10.63% threshold, and a few a little less (8.3-9.8%), despite these OTUs coming from different islands, being morphologically distinct and attributable to separate species. Moreover, clades C, H, M+N, P and Q further split into subclades with either high distances (> 8.5%) representing species, or in the range of 5-7%, which ought to be interpreted as subspecies, taking into account that the intra-locality divergence in this genus may be high, as observed in three specimens of H. calvus from Barranco de la Torre, in Fuerteventura (1.5-3.7%).
Clade C from Fuerteventura splits into an almost cryptic species (> 10% distance, H. betancuriae n. sp.) and the Herpisticus calvus subclade with divergent sub populations in the northern, central, and southern sectors of the island (distances 4.8-6%), but they are impossible to distinguish morphologically. Thus, we refrain from establishing subspecific taxa. Surprisingly, the southern cluster of H. calvus includes two sequences of a morphologically completely different and sympatric species, H. rectipes n. sp, possibly due to a mitochondrial introgression or to incomplete lineage sorting (see remarks under that species). A future analysis with nuclear markers would help to unveil this striking case.
Clade K corresponds to Herpisticus subvestitus with four apparently allopatric subpopulations (distances 4.6-7.2%). The nomino-typical subvestitus clusters with specimens that look like H. lanatus, but are not. In fact, H. lanatus¸a former synonym of H. subvestitus, separates with distances of 8.8-9.4% as clade M. Therefore, in order to highlight these relations in the complex of H. subvestitus s.l., we recognise a H. subvestitus pseudolanatus n. ssp., while resurrecting grancanariensis (also a former synonym of subvestitus) as a single subspecies covering the northern sector of Gran Canaria, despite its splitting genetically -4.6% distance-in a northeastern (Tafira type localily) and northwestern subclade (Sardina, Agazal, etc.)  However, a third Gomeran species, H. gomerensis n. sp. (clade R) joins with taxa from El Hierro and La Palma (clade S), the youngest islands. Based on COI, clade Q is indeed the only one present on more than one island. Moreover, the clade S representing H. hierrensis, has a really low internal genetic distance of 2.7%, but splits into two subgroups, one present on El Hierro and La Palma (clade T), and the other exclusively from La Palma (see remarks ahead). In constrast with their meaningless genetic divergence, the specimens can easily be separated by their elytral pilosity. Therefore, we have established H. hierrensis benahoare n. spp. for the latter subpopulation as consistent with the differentiation observed in pilosity in the other Herpisticus species. It is clear that it is a more recent event in this case.
The previous paragraphs reflect the eclectic approach adopted in this base study, in which we have flexibilised the boundaries of genetically delimiting species to an 8.3% divergence in order to be consistent with the morphological evidence and the chorological information available to date. On the other hand, we have established subspecies only if strictly necessary to accomodate taxa already described or for solving a potential source of confusion (H. lanatus / H. subvestitus pseudolanatus n. ssp.).
Having set a taxonomic blueprint for Herpisticus, deeper genetical studies should address a reliable chronogram and the complete phylogeography of this difficult but highly interesting group of Entiminae. Did the many Herpisticus species arise from a single initial star-shape radiation, or was it a step-wise colonisation from East to West accompanied by a progressive insular differentiation? Will the three species from Tenerife (clades O and P) cluster together as the morphology suggests? Are there more species or subspecies genetically supported that merit description? Was La Gomera colonised from Tenerife or from Gran Canaria, which is a far more distant island? Is H. rectipes n. sp. a recent case of rapid morphological adaptation fostered by a habitat shift or is it an old species with the mitochondria borrowed from sympatric H. calvus? Is there hybridization between H. subvestitus and H. tasarticus n. sp. or with H. guanarteme n. sp.? These are challenging questions for future studies.

Taxonomy
Family curculionidae Latreille, 1802 Subfamily Entiminae Schoenherr, 1823   head. Rostrum longer than head and continuous with it, never longer than wider (L/W 0.74-0.97), without additional subsidiary carinae above the scrobes or internally along the epifrons; scrobe very deep and curved downwards before the large eyes (upper margin pointing mid of eye and tangent to it); mandibles plurisetose, with briefly pedunculated oval antero-lateral scars; pterygiae not protruding, in some species scrobes hardly visible from above; epistome V-shaped, epistomal keel little developed; rostral sulcus narrow and deep, more or less prolonged (reaching level of middle or posterior margin of eyes), often partially hidden by scales. Prementum triangular, shortly pedunculated, entirely covering maxillary palps; with 4-6 or more setae (Fig. 5B). Maxilla with small galea area (densely beset with setae), 3-4 lacinial teeth, third and second maxillary palps with 1 and 2 setae respectively ( Fig. 5A). Eyes lateral, large, slightly oval or round (L/W 1.00-1.16), more or less prominent (convexity 16-34%), with posterior orbital margin (as broad as 3-4 ommatidia diameter). Antennae short; scape about  half the length of flagellum, not exceeding middle of eye in backward position, slender and slightly bent at base, thereafter more or less incrassate towards apex (clavate); desmomeres 1-2 subequal, elongate and longer than the others; 3-6 shorter and more rounded distally, 7 much enlarged, obconical and closely applied to the club, forming part of it to some extent (funicle with 6-segmented appearance); club fusiform, maximum twice as broad as desmomere 6 ( Fig. 5C).
Pronotum little transverse (L/W = 0.8-0.9), widest at middle, before it or near base; with more or  less developed collar and basal rim forming narrow grooves, more developed laterally and usually vanishing towards middle; posterior angles more or less clearly toothed; disc usually with two pairs of impressions or pits (sometimes almost obsolete), the first pair more separated; median line usually incomplete or missing; a transverse irregular deep flexuous groove or line on each side of disc, near middle, and in cases a lateral single pit almost on the margin (these sculptural features vary and are often partially hidden by the scales, being less conspicuous).
Scutellum triangular more or less pointed or U-shaped, broader or narrower, usually beset with whitish scales.
Elytra straightly (occasionally concavely) truncated at base, elongate (L/W 1.6 -1.9), more or less oval, blunt or acuminate apically, without shoulders, widest usually at middle; lateral margin briefly but clearly emarginated above the mesocoxae (to accommodate the epimeron), and more or less sinuous above the metaxocae; base not much broader than that of pronotum, without humeral callus, elevated as prominent flange (at least laterally) which protrudes on the sides as a humeral tooth; apical declivity smooth (in some species, apex in lateral view somewhat reflexed); striae marked by large deep punctures (looking smaller if scales present) separated less than one puncture diameter; interstriae flat or subconvex.
Metathoracic wings absent. Ventral parts. Prothorax with basal and anterior sulci distinct; a deep groove in the middle of the flanks above the coxae; procoxae contiguous, twice as separated from posterior margin as from anterior: mesocoxae separated by 1/3 of their diameter and inter-mesocoxal process little elevated, trapezoid and narrow; metepisternite distinct (not fused posteriorly); metacoxae reaching metepisternite and margin of elytra; para-metacoxal sulcus marked, integument more or less abruptly inflated before it. Metendosternite ( Fig. 5E) with furcal arms widely open (150º), about twice as broad than at base; stalk broad, laterally strongly concave; longitudinal flange 0.7× length of furcal arms; crux at 2/3 from base; hemiducti little developed, positioned at apical third.
Legs slender and long (metafemora reaching or exceeding apex of ventrite 2); femora unarmed, moderately clavate (profemora usually thickened); protibiae ( Fig. 5D) with apex only projecting inwards (outer angle blunt) -except in H. rectipes-with strong narrow mucro partially covered by a tuff of thick setae converging on it (triangular appearance); protibiae more or less bent inwards anteriorly, 5-16 short denticles along their inner edge (usually alterning in size and in some species reduced to a crenulation), with setae attached to them not so thick and conspicuous as in Molybdotus; meso-and metatibiae with or without denticles or crenulation; metatibial corbels open, devoid of scales, ascending, with conspicuous large setal fringe on outer edge (0.20 -0.24× length of tibia). Tarsi robust, covered with dense fine setae dorsally and thick soles underside; tarsomere 2 as broad as tarsomere 1 but shorter (about 0.6×), tarsomere 3 as long as 2 but much broader; claws simple, connate at base.
Male genitalia and terminalia. Sternite VIII (Fig. 6F) formed by two transversal arcuate wing-like hermisternites with setae variable in numbers and size, connected by short membrane bearing a bifurcate spiculum relictum, sclerotised only at end of arms; sternite IX (spiculum gastrale) completely arcuate or only at apical third, with short tricuspid basal plate. Aedeagus ( Fig. 6D-E) shorter than half length of abdomen; penis tube as long or longer than temones (0.98-0.60×), in lateral view more or less arcuate and strongly curved towards base; lateral margins parallel at middle (in some species somewhat diverging at ostium level) and moderately depressed (H/W 0.56-0.72); apex more or less reflexed (lateral view); internal sac ( Fig. 7A) short, not surpassing middle of temones in relaxed state, of the "bubble-shaped" type (Arzanov, 2003) when everted, as follows: short tubular basal area beset with an ostiolar sclerite and strongly hook-like protruding bifid ligula, broad medial area globular or slightly bilobed (trapezoidal-shape) without conspicuous bumps, and apical area tube-like (constricted at base) without bumps and bearing distally a cup-like transfer apparatus ( Fig. 7B) formed by two semiclosed capsules with complex pleats inside, which hold a free (detachable?) elongate sclerotised seminal bulb with spermal duct at proximal end and gonopore at thinned distal end. Tegmen (Fig. 57) about as long as penis tube; ring generally narrow with more or less develop short median dorsal projection between parameres (variable); parameres 0.4-0.5× length of terminal apodeme, partially sclerotised along midline; apodeme Y-shaped at base, shortly expanded at apex. Tegminal membrane attached to base of penis tube 1/3-1/2 of its length. Females (Fig. 5H) as males ( Fig. 5G) but larger and broader (body length s.r. 8.4-18.6 mm), with pilosity less developed in many species; elytra proportionally longer in relation to pronotum (♀ > 3.1×; ♂ <2.9×)-one exception H. denudatus n. sp (♀ = 2.9-3.0×)-and much broader than pronotum, usually noticeable enlarged behind middle and more acuminate, with tip of apex sometimes uplifted; punctures of striae often smaller; legs normally a little less robust and protibiae less bent; denticles on inner edge often more reduced, particularly in meso-and metatibiae. Tarsi less robust in many species. Ventrites 1 and 2 usually less depressed at middle; ventrite 4 usually rounded apically, with more or less impressed basal grooves parallel to the lateral margins ( Fig.  5F), similar to those depicted for Geotragus brevidens Ren, Alonso-Zarazaga & Zhang, 2013. Spiculum ventrale of moderate length (Fig. 8D), with thin and straight apodeme; plate fused (more or less sclerotised), roundish or triangular in shape, about 0.5-0.6× length of the entire sternite, beset of abundant hairs on its apical margin. Ovipositor very short ( Fig. 8A-8C), as wide as long, about half length of spiculum ventrale; gonocoxite 1 membranaceous, gonocoxite  2 elongate blade-like and sclerotised, with flat styli distant from apex, bearing 4-5 setae (not surpassing apex). Bursa copulatrix without sclerites, with a fold or in continuum with common oviduct (characterised by microspiculed integument). Union of spermathecal duct simple; spermatheca ( Fig. 8B) U-shaped, with globose body, long curved thin cornu, broad short ramus, and flat  nodulus pointing to the same side; spermathecal gland globose, smaller than spermatheca and duct about as long as its diameter.
Abdominal segments I-VII of equal length, with three dorsal folds; spiracles located above lateral midline (vertical); on each side: 1 prs, 5 pds (1 short + 1short, 1 long + 1 short, 1 long), 1 small sps, 2 eps (short, long), 2 ps (long, short), 1 lsts, and 1 eus. Segment VIII shorter, with two dorsal folds; same chaetotaxy but 4 pds and spiracle pointing backwards. Segment IX (Fig. 9C) of type A (v. Van Emden, 1952), terminal, with 4 slightly sclerotised lobes converging around segment X: two large triangular at each side (with middle carina, almost touching at apex) bearing 1 psx; one dorsal with 3|3 ds, and one ventral squarish, located ventrally, with 2|2 ts (aligned distally, of equal small size). Segment X rather small, placed below midline, with four soft lobes around anus; the lateral lobes with 1 ts. description of the pupa (H. subvestitus pseudolanatus n. ssp., Fig. 11). With the characters of the entimine pupae according to Marvaldi (1997); length 7.7 mm, width 2.5 mm (across the thoracic segments); body elongate, of white-creamy colour, with setae arising from conical elevations of various heights (papillate setae). Most setae are somewhat thorn-like but not very robust and with the tip usually bent; the thinner ones are hair-like, more curved or twisted apically.
Abdominal segments I-VII of almost equal length (except VII, ventrally), narrowing smoothly and with setae increasing gradulally in size and more strongly papillate from IV towards apex. Five spiracles, not pigmented, inconspicuous, difficult to observe. Segment IX hardly visible from above. Chaetotaxy: 1|1 d 1 hairlike located anteromediad (longer and twisted in VII-VIII), 5|5 d 2-6 aligned posteriad (1+2+2); 2|2 hair-like ls located near the tergum-sternum border, and segment VIII terminal with the median seta d 2 missing, and d 5 crowning a rather protruding longitudinal lobe. Ventrally, 3|3 small v (1+2) with the pair of v 1 more distant among them on segments VII and VIII. Segment IX almost ventral, with gonotheca divided (♀ pupa), 1 v at each side of it and 3 |3 pc not very outstanding (pc 1-2 on small papillae).
reMarks. Herpisticus are short-nosed weevils (Entiminae) of the tribe Tanymecini, as they bear a row of vibrissae -although short and not very  conspicuous in some species-along the front margin of each side of the prothorax, which is straight; claws are connate at their base (subtribe Piazomiina), and the second ventrite is longer than the third and separated from the first by an obliterated suture (Morimoto et al., 2006). Morphologically, it can be related to the  Piazomias -Leptomias generic group (Leptomias, Geotragus, Hyperomias, Pachynotus, Xizanomias, Triangulomias, Odontomias, etc.), which is very speciose and mainly distributed in Asia (India, Pakistan, Himalayan countries, China, Siberia, Korea, etc.), but also in the southern Ethiopic region in the case of one lineage of Piazomias (Congo, Angola, Namibia, and South Africa). However, potential relations with  other extended African forms (Polyclaeis, Cimbodes, Beardiella, etc.) should be checked.
Herpisticus is endemic to the Canary Islands and represents a rather isolated disjunct geographical situation. It can be recognised by the crenulated or denticulate protibiae (no spiniform teeth) and the six-segmented looking funiculus of the antennae because the seventh antennomere is cup-like and attached to the club. Moreover, the scrobe is tangent to the eye (separated in Leptomias), its dorsal margin pointing to mid-eye, and the rostrum is dorsally free of carinae. The transfer apparatus found in Herpisticus is likely to be specific to this genus (more complex than in Geotragus), and may be important in separating species, but it is a difficult structure to study due to its size and the obscure scheme of folds inside it. The presence of a spiculum relictum in the male sternite VIII suggests an old lineage for this genus.
In habitus, Herpisticus resembles most Molybdotus vermicularis (C.O. Waterhouse, 1881) from the island of Socotra, possibly due to convergent adaptation to similar semi-arid habitats. Molybdotus has antennomere 7 broader than antennomere 6, but is clearly separated from the club, and the tibiae have thick stiff spine-like setae attached to each strong denticle (soft setae in Herpisticus and denticles less developed).
The morphological variation in shape and vestiture within species of Tanymecini is a known drawback in taxonomic studies. However, the variability attributed to pilosity in Herpisticus is far from real, and pilosity has shown to be largely coherent with the groups molecularly postulated as different species, being a good character, although mainly useful with males. Pilosity in females is usually less developed and for some species the following keys do not work well with females. Moreover, special care has to be taken in the observation of pilosity, as specimens kept in museums often have their hairs eaten by mites or booklice, and just a stump remains.
reMarks. Herpisticus oculatus has the most protruding eyes in the genus, is one of the smallest in size (together with H. rectipes n. sp. and H. nanus n. sp.), and has the longest protibiae, somewhat resembling Coelositona when walking. When they are being  disturbed on the ground, they abruptly turn up-side down and freeze with their extremities stretched, a behaviour not yet observed on other Herpisticus.
It can be recognised by the sparse uniform pilosity of the elytra, with curved hairs hardly overlapping longitudinally. The other Herpisticus from the eastern Canary Islands have both recurved setae and a less sculptured pronotum. Wollaston grouped this species and H. calvus as having the second desmomere of the funiculus clearly longer than the first, however it can be the opposite and in many cases they are of equal length. This character is not stable, as already highlighted by .
The genetic distance between a sequence from the central part of the island (Teguise) and another from the south (Yaiza) is 5.3%, suggesting a geographical split within this species, but with no appreciable morphological differentiation. distribution and ecology. Herpisticus oculatus is endemic to Lanzarote. It dwells in the arid central and southern parts of Lanzarote on clayey soil, avoiding the young extensive volcanic lavafields and sandy habitats (Fig. 14). In the north of the island it is replaced by H. famarae n. sp. Usually collected below stones and debrie, it is a summer animal and rather uncommon in winter, in comparison to H. calvus on the neighbour island of Fuerteventura. It feeds on Atriplex semibaccata, Salsola vermiculata, Medicago sp. and oder scrubs.
etyMology. The specific term "famarae" is the genitive of the name of the massif north of Lanzarote where the species lives, although not exclusively.

Herpisticus calvus
reMarks. Herpisticus calvus can be distinguished from H. oculatus by its larger size, more contrasting colouration (fore/hind body), less prominent eyes, protibiae not so much longer than metatibiae, smoother surface of prothorax and elytra with more uniform, shorter and denser pilosity (more than 3 recurved hairs across an interval, overlapping longitudinally).
The single record of H. calvus from Gran Canaria (García, 1986) is a misidentification of H. subvestitus pseudolanatus n. ssp., and its presence on Lanzarote is questionable. Heer (1857) recorded H. eremita for Lanzarote and Fuerteventura. Later, Wollaston (1864) refers these records to H. calvus as he was able to inspect several beetles collected by Hartung and identified by Heer, but he received only one Herpisticus specimen from the Swiss entomologist, possibly that collected on Fuerteventura. So, the assumption he makes ) that the Lanzarote specimen or specimens belong to H. calvus is not solid; they could well belong to either H. oculatus or to H. famarae n. sp. The females of H. famarae are much longer than those of H. oculatus and can easily be taken for H. calvus. On the other hand, Uyttenboogaart (1940)-who synonymised H. calvus and H. oculatus with H. eremita-highlighted a series of six specimens with prominent eyes and long hairs ("var. lanata Woll.") found in Fuerteventura. However, in his collection preserved at Leiden (NMNH), there is indeed a series of six specimens from Ch. Alluaud labelled  "Fuerteventura" from which five are true H. calvus, and only one is a hairy H. subvestitus grancanariensis. In the abundant material studied from many localities from the entire island (551 exx), we have found no trace of hairy Herpisticus on Fuerteven tura and, as Alluaud also collected profusely on Gran Canaria (Alluaud, 1891), a labelling error is highly probable.
As previously commented in the genetic results, H. calvus seems to be geographically structured over the island, with a subpopulation on the mid-eastern coast, another widely spread in the centre, and another on the peninsula of Jandía. The K2P distance within these groups is 2.0-2.5% and between them it ranks 5.3-6.6%, which is high enough to consider them subspecies. Nonetheless, this species varies considerably in shape, and narrow tube-like males can coexist with broader more oval ones in the same locality. There are no good morphological characters that help distinguishing the genetic subpopulations, and the molecular sampling (n=14) does not cover the whole island (northerly parts and the islet of Lobos are missing) as to permit delimiting the extent of each subpopulation. Therefore, it is prudent to keep H. calvus as a monotypic species until a deeper study is conducted, but with one exception. A specimen collected in the massif of Betancuria occupies a basal sister position to the rest of the clade and shows an outstanding K2P distance of 9.7-11.6%, so to merit its description as a separatealmost cryptic-species, which follows next. distribution and ecology. Herpisticus calvus is endemic to the island of Fuerteventura and islet of Lobos (Fig. 19). A common animal under stones on arid land with clayey soils, scattered but can be locally abundant ( fig. 49A). Collected at night on Launaea arborescens, Kleinia neriifolia and more common on woody Chenopodiaceae (Salsola, Suaeda, etc.). The parasitic mite Leptus (Leptus) maxorata Haitlinger, 2009 (Acari: Prostigmata: Erythraeidae) was discovered by the first author on specimens from Malpaís Grande. description. Herpisticus of moderate size (body size ♂ 8.7×3.2 mm -10.5×4.2 mm) and similar appearance as H. calvus, but darker, usually with very few scattered isolated coppery scales (on elytra forming small patches of 3-9 scales); integument black and shiny beset with a dense cover of small recurved setae (1/2 length of a claw), devoid of longer hairs at apex. Antennal flagellum ≥ 2× length of scape; club L/W = 3. Rostrum (L/W 0.8) converging apicad, with dorsolateral margins more sharply rimmed than in H. calvus (blunt), and sides above scrobe less vertical and more visible from above; base broader (ratio rostrum width at base / interocular distance usually > 1.15). Eyes a little oval (L/W 1.12), more prominent (convexity 30%). Pronotum (L/W 0.86) with more or less curved sides, widest at middle or just behind; sculpture as in H. calvus, with mid-line usually marked. Elytra oval-elongate (L/W 1.8), not acuminate, rarely subparallel; striae rather conspicuous with large punctures, interstriae usually subconvex and outstanding. Tibiae with denticles a trifle more developed, but not always; hairs on outer side of protibiae slightly longer. Spiculum relictum of sternite VIII with thin capitated arms. Penis ( Fig. 53D) with apex almost straight (lateral view); endophallus as in H. calvus, slightly less bilobed; tegmen without short dorsal projection between parameres; parameres broad.
etyMology. The specific term "betancuriae" refers (in genitive) to the old island capital and now village of Betancuria, which gives its name to the massif where the species lives. This first settlement was so named after Jean de Bethencourt, the Norman knight that conquered the island in 1405.
reMarks. This species could be considered an almost cryptic taxon, as the morphological differences with H. calvus are subtle, particularly within the wide variation range of the latter. The edged or blunt lateral margins of the rostral dorsum are not easy to distinguish unless specimens of both species are available to compare. Doubtful specimens can be discerned by dividing the rostrum width at its base by the interocular distance. In H. betancuriae n. sp. this ratio usually falls in the range of 1.15-1.17, while it is below 1.15 in H. calvus, and the exceptions are few. Moreover, the antennae are longer in H. betancuriae, and their flagellum is at least twice the length of the scape, while in H. calvus it is less (1.7-1.9×).
distribution and ecology. Herpisticus betancuriae n. sp. is endemic to the island of Fuerteventura, living apparently restricted to the massif of Betancuria, along its interior valley, but not reaching Pájara in the south (Fig. 19). This massif represents the basal complex of the island and is considered the oldest emerged part of the Canarian archipelago (ca. 20 Ma).
The new species dwells in arid and semiarid habitats and is allopatric with H. calvus, despite their sharing the same ecology.   description. Herpisticus of moderate size (body size ♂ 9.7×3.8 mm -10.6×4.2 mm), long-ovate; integument black and shiny, flattish corneolate; general colou ra tion dominantly dark (scales deciduous), elytra with variable patches of scales; scales round, small and separated (pink, coppery, blue, beige, white, etc.); pilosity uniform but inconspicuous, hairs flat, recurved and very small (1.5× scale diameter) on elytra, pronotum and also on external face of femora and tibiae, not or little overlapping. Antennal flagellum as long as twice the scape; club L/W about 2.8. Rostrum of rather trapezoidal appearance (L/W about 0.8), sides convergent apicad; paraepistome more or less abruptly depressed; median sulcus deep and broad, with a shallow longitudinal line on each side (apical half); lateral margins blunt, area between them and upper somewhat carinate margin of scrobes depressed, forming a step. Eyes a little oval (L/W 1.09), moderately convex (24-26%); post-ocular vibrissae short, not longer than width of anterior rim of prothorax. Pronotum (L/W 0.81-0.82) with sides little arcuate, widest at middle; anterior and basal  margins broad, basal constriction brief, with little marked angular tooth; disc without pits or median line; lateral linear impressions deep and long, and some shallow transverse rugosity. Scutellum triangular, blunt. Elytra oval-acuminated (L/W 1.6), widest at middle, about 2.77× length of pronotum, with smooth declivity; striae with large punctures in basal third; external interstriae slightly subconvex; with 4-5 hairs across. Protibiae with small denticles as in H. calvus, little longer (> 1.06×) than metatibiae; meso-and metatibiae crenulated. Spiculum relictum of sternite VIII with arms shorter than base and sclerotised in distal half (Fig. 52F). Penis triangular at apex, tip blunt, almost straight in lateral view (Fig.  53E).
etyMology. The specific term "jandiensis" refers to the peninsula of Jandía in Fuerteventura, where the species lives.
reMarks. The K2P genetic divergence between H. jandiensis and H. calvus is greater (> 14%) than with species from other islands, like H. guanarteme n. sp. (12.1%) and H. denudatus n. sp. (11.9%) from Gran Canaria or H. famarae n. sp. (12.9%) from Lanzarote. In size and shape it resembles the latter, which likewise has a rostrum with weakly defined lateral margins, but less elliptical elytra, slightly longer hairs and longitudinal rugosity all over the rostral dorsum. Herpisticus jandiensis n. sp. is easy to distinguish from the other species living on Fuerteventura by the very small non-protruding recurved setae on femora and tibiae, and by a lateral step on its rostrum. distribution and ecology. Endemic to the island of Fuerteventura, where it lives on the summits of the peninsula of Jandía, in the extreme south, whereas L. calvus lives on the same peninsula but at lower altitude (Fig. 19). These summits of the Jandía massif have sufficient elevation (Pico de la Zarza, 807 m altitude) to receive the direct influence of the clouds formed by the trade wind, maintaining a more humid habitat in strong contrast with the xeric habitats of the rest of the island. There are many endemic species of beetles restricted to this narrow biodiversity hotspot.  description. Herpisticus of small size (body size ♂ 7.0×2.9 mm -8.1×3.2 mm), long-ovate; scaling of light patchy colou r (whitish, beige or pale blue) if scales not missing (deciduous); scales round and polygonal, tangent; pilosity inconspicuous, hairs bent, scarce and tiny, about as long as one 1-1.5 scale diameter, 2-3 hairs across an elytral interstriae; suberect on tibiae and femora. Antennal flagellum 1.9× length of scape; club oval (L/W 2.7). Rostrum slightly convergent apicad (L/W 0.77-84), with flat dorsum and broad median sulcus, surrounded by a middle groove. Eyes slightly oval (L/W 1.13), moderate in convexity (26-28%); post-ocular vibrissae long (clearly longer than width of anterior rim of prothorax). Pronotum (L/W 0.76) with sides arcuate or curved, widest at middle or behind; anterior postmarginal depression marked; posterior angular tooth small; median sulcus fine and complete, no discal pits; lateral groove distinct. Scutellum small, not raised, long-triangular. Elytra broad-elliptical, bulky in appearance (L/W 1.6), widest at middle, narrowed basally, about 3× length of pronotum and 1.5 its width, with declivity starting at middle; anterior rim complete but not very abruptly raised (no basal sulcus and humeral tooth moderate); striae as wide as interstriae. Abdominal transversal convexity about 77%. Legs long and slender; profemora inflated, clearly wider at middle than apically (1.5×), more than in H. oculatus (1.3×); protibiae longer than metatibiae (1.11×), straight apically, somewhat narrowed preapically and slightly bent at middle, truncate at apex with inner and outer angles right or slightly protruding outwards; denticles on inner rim very small (crenulate); meso-and metatibiae with smooth inner rim; tarsi slender with elongated onychium, as long (without claw) as scape. Spiculum relictum of sternite VIII (Fig. 6F) monopodic with Y-shaped distal sclerotisation. Penis (Figs. 6D-E, 53F) with oval apex, not expanded laterally, and tip almost straight; base of transfer apparatus U-shaped. Tegminal ring narrow, with inicipient dorsal projection between the parameres; parameres very broad at base, tappering.
etyMology. The specific epithet rectipes is an invariable Latin adjective composed of the terms "rectus" (straight) and "pes" (foot), in reference to the straight fore tibiae characteristic of this species.
reMarks. Herpisticus rectipes is apparently as small as or even smaller than H. oculatus, although the series available are too limited to confirm this. It is equally long-legged but with broader elliptical elytra. It can easily be separated from any other species of Herpisticus by its elongated onychia and its straight and flattish protibiae, truncated apically, with the outer angle right or even protruding outwards. The hairs on protibiae are stiff and suberect, while in H. calvus they are recurved. The clear coloured scaling is probably related to the calcareous-sandy ground where it lives, a tendency in this genus that was already commented on by Wollaston (1864).
Among all the Herpisticus species inhabiting the eastern Canaries, H. rectipes n. sp. is the most divergent morphologically. Its relationship to any of them is obscure, as it clusters very closely in the COI   phylogram with specimens of H. calvus from the same region, southern Fuerteventura (Fig. 3, clade C). This is probably due to a mitochondrial introgression, as postulated in the section of Genetic Results. An alternative to this hypothesis would be incomplete lineage sorting, if we attribute a recent speciation to H. rectipes and explain its greater morphological disparity by forced differentiation due to habitat shift (elongation of onychium to walk on sand, for instance). In support of this possibility, the aeolian sandy environment of El Jable in the Istmo de la Pared is of Pleistocene and recent origin, much younger than the Miocene shield basalts north and south of it (Carracedo & Troll, 2016). distribution and ecology. Endemic to Fuerteventura and located in the sandy belt that separates the southern peninsula of Jandía from the rest of the island (Fig. 19), whereas its presence in the two other nearest sandy sectors, Jable de Vigocho and Jable de la Angostura, has not yet been checked. Specimens have been found on Ononis natrix, on Launaea spinosa, on Suaeda vera, and under stones (Fig. 49B). Wollaston, 1864 redescription. Herpisticus of intermediate size (body size ♂ 8.5×3.0 mm -11.0×3.9 mm); body elongate and narrow (elytra/prothorax width ratio 1.26-1.32); of testaceous-brown colouration, forebody darker and abdomen (patchy pattern) lighter; scales large, roundish (some occasionally oval), subconvex, and tangent. Pilosity very dense and conspicuous (woolly appearance) formed by a mix of shorter inclined and longer erect silky hairs (1-2.5× tarsal claw), often irregularly shaped or arcuate pointing forwards; 6-7 across elytral interval, long at base but longer towards apex (apical tuft not differentiated); hairs less developed but also protruding on head, pronotum (in profile view), and on legs (hairy tibiae). Antennal flagellum slightly longer than twice the scape; club slender (L/W 3.3 flat with soft declivity towards apex; punctures of striae small. Elytra/pronotum length ratio 2.7-2.8. Protibiae slightly longer (1.1×) than metatibiae; protibiae with 3-4 strong denticles and 1-2 smaller ones in the intervals, shorter on mesotibiae; metatibiae crenulated. Penis tube (Fig. 53H) almost uniformly arcuate; apex moderately curved (lateral view) and tapering (dorsal view), tip slightly twisted to one side; everted endophallus with bulb-shaped median part (Fig. 55C). Tegminal ring between parameres broad, without dorsal median projection. Spiculum relictum of sternite VIII with short thin arms. Females as males but larger and broader (body size 10.9×4.2 mm -12.7×5.0 mm); pronotum wider (L/W 0.79): elytra wider (L/W 1.72) but clearly longer and apically subacuminate (elytra/ prothorax length ratio 3.2m, width ratio 1.47); less hairy, with shorter curved hairs intermixed with erect longer hairs that persist mainly in apical half of elytra. Pigidium covered by dense pubescence. Protibiae usually with only 3 conspicuous denticles. Spiculum ventrale (Fig. 58); spermatheca (Fig. 59D).

Herpisticus lanatus
reMarks. There is a COI K2P distance of 8.3-10.5% between this species and H. subvestitus pseudolanatus n. sp., despite their striking resemblance (Fig. 22). Both have a woolly appearance due to the dense and very long upraised pilosity over the entire elytra, and shorter but equally conspicuous on pronotum, head and legs. H. lanatus tends to be larger, the males not so narrow, the sides of pronotum usually less rounded, and the eyes less convex (20-25% vs 25-28% in pseudolanatus). However, these differences are hard to gauge even with several specimens for comparison, as  they may overlap due to their respective variation. The apex of aedeagus is a little less curved, more acute (in dorsal view) and slightly twisted in H. lanatus, and the mid-section of its endophallus is not cordiform but bulb-shaped, characters that are not easy to inspect. Nonetheless, the geographical location of specimens may be a useful practical criterion. All specimens originating in the older western part of Gran Canaria (9)(10)(11)(12)(13)(14) Ma, Miocene phase with phonolites and trachytes) belong to H. lanatus, while those of H. subvestitus pseudolanatus n. ssp. are restricted to the younger eastern part (Pliocene and Quaternary phase with basanites and alkali basalts), where the eastern margin of the Amurga massif, near the Tirajana ravine, acts as a factual separating border.
Other Herpisticus species of woolly appearance, like H. aridicola n. sp. from Tenerife, or H. hispidus n. sp. from La Gomera can be separated by their penis being slightly bent at the apex, instead of curved as in H. lanatus and even more curved in H. subvestitus pseudolanatus n. ssp. distribution and ecology. Endemic to Gran Canaria and distributed in the south-west sector of the island (ancient part), between the ravine of Tirajana and the ravine of Mogan (Fig. 33). It can be found in diverse kinds of arid and semiarid habitats from the coastal and lower zones, up to lofty elevations bordering the pine-forest in the mountains. It has been obtained by beating Kleinia neriifolia at night, but most specimens were collected from below stones. Occasional damage to pumpkin fields has been reported near Maspalomas. Wollaston, 1864 Figs. 22 redescription. Herpisticus of intermediate to large size (body size ♂ 11.8×4.0 mm -12.3×4.9 mm), oblong in appearance (elytra/pronotum width ratio 1.34). Vestiture dominantly greenish or brown, variable; usually devoid of pink scales on femora; scales polygonal, subconvex, and separated; elytra with cover of depressed bent hairs of different sizes (about as long or longer than a claw), overlapping irregularly, with apical tufts; hairs shorter on head and pronotum; short and hardly overlapping on femora; on outer edge of tibiae not longer than their diameter. Antennal flagellum 2.16× length of scape; club elongate (L/W 3.2). Rostrum (L/W 0.86) slightly convergent apicad, lateral margins not sharply defined and antero-lateral angles blunt. Eyes almost round, moderately prominent (convexity 24%). Pronotum (L/W 0.85) with sides arcuate, widest about middle, base briefly constricted without marked angular tooth; dorsum rather even (less sculptured than in H. grancanariensis), discal pits inconspicuous. Scutellum usually U-shaped. Elytra rather elongate (L/W 1.9), about 2.9× length of pronotum (lectotype); dorsum flat (occasionally slightly depressed at disc); widest at middle or slightly behind, sides arcuate until apical third, wider at base and somewhat acuminate (less acuminate than in H. guanarteme but more than in H. subvestitus grancanariensis); interstriae at disc wider than striae. Abdominal longitudinal declivity starting at apical 1/3; transversal convexity about 77%. Protibiae with many denticles (usually starting apically with four short ones, then four strong denticles each followed by two smaller ones); meso-and metatibiae (not always) with small denticles or granules. Spiculum relictum of sternite VIII with base narrower than length of arms (Fig. 52D). Penis tube (Fig. 53I) only slightly arcuate in middle third, with apex notably recurved and in dorsal view acuminate (tip stronger than in H. subvestitus grancanariensis); mid part of endophallus widely bulb-shaped (Fig. 55D); tegmen with short manubrium short and apodeme between parameres (Fig. 57F). Females like males, but larger (body size 12.3×5.0 mm -14.4×5.9 mm) and with pilosity shorter and less widely spread; pronotum broader (L/W 0.81), elytra wider, more rounded laterally and widest behind middle (L/W 1.78). Ventrite 5 densely pubescent, with a median tuft of strong setae at base. Plate of spiculum ventrale (Fig. 58G) short, with apical margin slightly emaginate in middle third. Spermatheca with short collum (Fig. 59E).

Herpisticus subvestitus subvestitus
reMarks. Wollaston characterised his var. β subvestita as having "long silky erect hairs mostly in apical half of elytra, but not many". However, the lectotype designated and photographed by Stüben  has long silky hairs over the entire elytra, although not so erect and perhaps slightly longer apicad. The Wollaston label of the lectotype depicted by Stüben (op. cit.) reads "Larajana", which can only refer to the village of San Bartolomé de Tirajana, located in the large circus bearing the same name Tirajana, which forms part of the central mountain ridge of the island. Specimens collected in this area belong either to Herpisticus subvestitus subvestitus or to H. guanarteme n. sp., and both have hairy elytra. The former can be recognised by its shorter and slightly convergent rostrum with blunt antero-lateral angles, while the latter has a longer, squarish and parallel-sided rostrum with more sharply angulated dorsal margins, and its suberect pilosity is more uniformly spread over the entire elytra (hairs not twisted).
Having fixed the morphological concept of H. subvestitus, the COI molecular analysis (Fig. 3, clade K) reflects that H. subvestitus from the type locality of Tirajana is related to a complex of genetically differentiated populations from the lower zones, which cluster together and separated from H. lanatus, which was considered one of its synonyms. Figure 22 shows the range of COI K2P distance among them, supporting the resurrection of grancanariensis Palm, 1974 as subspecies of H. subvestitus (previously also considered a synonym), with a 4.5-5.9% genetic distance, and the establishment of H. subvestitus pseudolanatus n. ssp. with a similar  The nominotype subvestitus can be confused with H. subvestitus grancanariensis, but the former is a more robust animal, its elytra are laterally less parallel, their declivity in lateral view is restricted to the apical third (not starting progressively before middle), the long hairs are not limited to the apical half and extend over the entire elytron, the eyes are less prominent (22-24% instead of 27-29%) (Fig. 25A), the penis apex is somewhat more sharply pointed and less twisted laterally, and the females bear a protruding more or less developed middle tuft of setae near the base of ventrite 5, which is characteristic of this subspecies.
Two specimens sequenced from La Aldea (cemetery) -not shown in the phylogram-cluster with H. tasarticus n. sp., which is present in the same basin, but has less curved apex of penis, longer and more parallel rostrum and less developed elytral pilosity. Again, this discrepancy with the COI results suggest another case of mitochondrial introgression, to be confirmed with nuclear markers. distribution and ecology. Endemic to Gran Canaria (Fig. 33). It is distributed at high and intermediate elevations in the central mountain region of the island, including the heads of the watershead of Telde, Tirajana, and La Aldea-Inagua, apparently expanding towards lower elevations in the NW (La Aldea, Guayedra). It has been found under stones and beaten from Teline microphylla or Atriplex halimus (parking parterre). redescription. Similar to nominotypical subspecies but less robust and smaller in size (body size ♂ 7.7 × 3.0 mm-10.6 × 3.9 mm); body elongate and somewhat tubular in appearance (elytra/prothorax width ratio 1.32); usually testaceous dominant colouration with dark patches, often with rosewood coloured scales laterally on thoracic and abdominal sternites; scales large, roundish (some occasionally oval), subconvex, not separated (tangent), and not particularly deciduous. Pilosity much denser and longer, usually irregular in shape and as conspicuous as in H. lanatus (occasionally longer than 2.5× length of claw); on legs more uplifted and clearly protruding (hairy appearance). Eyes more prominent (convexity 26-28%), as prominent as in ssp. grancanariensis. Pronotum (L/W 0.86) with rather evenly curved sides, widest at middle or below it; base with lateral angles blunt (no angular tooth). Elytra narrow, elongate and somewhat subparallel (L/W 1.82), widest about middle; with longer soft declivity towards apex. Protibiae slightly longer (1.1×) than metatibiae; protibiae with 3-4 strong denticles and  1-2 smaller ones in the intervals, shorter on mesotibiae; metatibiae crenulated. Penis tube (Fig. 53J) elongate at middle (only slightly arcuate), with short temones (0.6× length of tube); apex strongly and widely recurved; mid-section of everted endophallus narrow and cordiform (Fig. 55E). Tegminal ring between parameres broad, with a small dorsal projection. Spiculum relictum of sternite VIII with short thin arms.
etyMology. The Greek prefix pseudo-("lying, false") added to the Herpisticus species name lanatus means that the new taxon looks like it but is something else.
reMarks. It is surprising that this subspecies is morphologically more differentiated from the other subspecies of H. subvestitus than from H. lanatus, looking almost identical to it despite the greater genetic distance (Fig. 22). It can be separated from H. lanatus by more prominent eyes and some genital characters already commented; from H. subvestitus subvestitus by its smaller size, more prominent eyes and more developed pilosity over the whole body, including legs; and from H. subvestitus grancanariensis, by having additional long protruding hairs in the basal half of elytra. distribution and ecology. Herpisticus subvestitus pseudolanatus n. ssp. is located in the eastern sector of the island of Gran Canaria, east of the Amurga massif, mainly on the wide arid lowlands, ascending occasionally up to at least 800 m, in what was termed as the "dry-young-sector" by Menéndez et al. (2008), in their geomorphological sectorisation of the island (Fig. 33). Its distribution northward seems indeed  to end after Telde, but there is at least one disjunct record from Las Canteras beach in the city of Las Palmas, within the domain of H. subvestitus grancanariensis; otherwise, they are allopatric. It is not uncommon on clayish barren lands and can be found on several plant species or in their vicinity: Rubia fruticosa, Launea arborescens, Patellifolia patellaris, Atractylis preauxiana, etc. The flatland areas where it dwells are subject to intense agriculture, and this weevil has become an occasional pest of several vegetables and tropical trees. The populations in organic agriculture greenhouses (e.g. pumpkins, pepper) can reach very high numbers. Palm, 1974  redescription. Herpisticus of intermediate size (body size ♂ 9.2×3.4 mm -11.8×4.2 mm); long oblong in shape (elytra/prothorax length ratio 2.85), less robust than ssp. subvestitus; with variable mottled pattern of brown, grey, green and coppery scales; femora usually with a wide ring of pink scales; scales roundish, separate. Pilosity moderately developed, on elytra with recurved and curved hairs mixed on apical half with longer suberect separated hairs (apical tuft present); inconspi cuous on pronotum, a few short erect hairs on head, moderate on legs, and on tibiae longer than its diameter. Antennal flagellum twice as long as scape; club slender (L/W 3.1). Rostrum slightly convergent apicad (L/W 0.77-0.82); eye convexity 27-29%. Pronotum (L/W 0.82-0.84) usually with curved or arcuate sides, with marked collar and without sharp angular tooth at base, widest near middle; lateral grooves marked, the four discal pits generally well marked; disc uneven. Elytra oblong, somewhat cylindrical (L/W 1.9), about 2.9 times longer than pronotum; broad at base, sides subaparallel, widest about 2/3, then curved, blunt apically; striae as broad as interstriae, the latter often subconvex; punctures deep. Legs robust, profemora incrassate; protibiae largely arcuate, slightly longer than meta tibiae (1.07×), denticles less developed than in ssp. subvestitus, reduced to granules on mesotibiae, usually absent on metatibiae. Spiculum relictum of sternite VIII (Fig. 52E) with base about as broad as length of arms. Penis tube (Fig.  53K), elongated with apex recurved like in ssp. pseudolanatus, slenderer and acuminate in dorsal view, but not as much as in H. lanatus; endophallus as in ssp. subvestitus, but distal part longer; tegmen with strong median dorsal projection between parameres and long manubrium (Fig. 57E).
reMarks. Palm (1974) described H. grancanariensis with abundant material from Gran Canaria, lumping all specimens of H. eremita auctt. with the apex of penis strongly curved upwards, despite their variation in sizes, shape and pilosity. Fortunately, he designated a holotype from Tafira that helps to fix its concept. It is difficult to separate from H. subvestitus, but its size is on average smaller, the eyes are more prominent with a convexity of 27-29% instead of 22-24%, pronotum more rounded laterally and constricted basally, and the elytra are more parallel and convex, with blunt apex. Some females are not much broader than males (Tafira 12.7×4.4 mm), whereas the difference in shape between sexes in   Palm, 1974 ♂ (Gran Canaria, Tafira Baja) H. subvestitus subvestitus is more pronounced. In this latter subspecies, the pilosity in the basal half of elytra is formed by small curved hairs mixed with longer ones, and females have a median tuft or field of longer hairs on ventrite 5, which is not the case in H. grancanariensis (only small hairs at base, uniform pilosity on ventrite 5).
reMarks. Herpisticus guanarteme is present in part of the distribution area of H subvestitus subvestitus, but can be recognised by its generally larger size, longer and parallel rostrum with sharply marked lateral  margins and squarish anterior angles, more narrowed elytra in basal third and acuminated in apical third; the pilosity is usually longer, more uniform, and of a somewhat golden tint, and the tip of the penis is moderately bent upwards, not strongly recurved like in the group of subvestitus-grancana riensis-lanatus. The general appearance of this species is rather stable and can be recognised with some training, despite its varying in details. Some females have a rather arcuate or even subtruncate elytral apex (excluded from the type-series). The long pilosity of elytra can be equally or even more developed than in typical H. lanatus, also bearing long twisted hairs; however, hairs are recurved on the head, never upstanding or protruding as in the latter species or H. subvestitus.
Specimens of H. guanarteme n. sp. from the Tamadaba massif and west of it have less developed pilosity in the basal half of the elytra, and the rostrum is not so parallel, like in H. subvestitus (doubts can be solved by inspecting the penis apex). They are likely to belong to a differentiated subpopulation, a hypothesis supported genetically by the 5.0-5.8% K2P distance obtained with two available sequences. Nonetheless, we prefer to wait for more chorological and molecular data before formally establishing another new taxon. description. Herpisticus of large size (body size ♂ 11.9×4.8 mm -13.5×5.6 mm); robust, somewhat navicular in appearance; scaling rather deciduous (frequently only scattered white, cinereous, bluish or pink scales remain), more persistent on legs and head; integument piceus, shiny, with a superficial granular texture; scales polygonal and separate. Pilosity on elytra rather dense (4-5 hairs across interstria); hairs suberect and bent, variable in length, intermixed, but shorter than a tarsal claw; on head and pronotum depressed, inconspicuous. Antennal flagellum 2× length of scape: club elongate (L/W 3.4). Rostrum long and parallel-sided (L/W 0.97), with lateral margins sharply angulate: eyes slightly oval (L/W 1.07), not overly protruding (convexity 19-22%); central furrow reaching mid eye-level. Pronotum (L/W 0.84) less constricted posteriorly than anteriorly; slightly emarginated at base (middle third), sides widest about middle or before, then straight towards base; posterior angle not rounded (near right), with marked angular tooth; integument rather uneven, collar conspicuous, median line, postdiscal and lateral depression deeply marked; without discal points. Scutellum triangular and narrow. Elytra ovoid-acuminate (L/W 1.71-1.73), widest at middle, broad at base and rather pointed apically (more strongly acuminate than in H. guanarteme), 2.64× length of pronotum; striae as broad as or broader than interstriae, punctures large and deep, and interstriae moderately convex (highly conspicuous as they are usually free of scales). Legs slender, protibiae thin, 1.08× length of metatibiae, with 5-7 small denticles decreasing in size posteriad, tiny in mesotibiae, absent in metatibiae; protarsi broad. Spiculum relictum of sternite VIII with arms longer than base (Fig. 52K). Penis tube (Fig.54B) with apex ogival-pointed, a trifle reflexed; base robust; endophallus globose-elongate, with strongly bent apical part (Fig. 56A), transfer apparatus with emarginated base (Fig. 7C-D).
Females larger (body size 13.7×6.0 mm -18.6×7.6 mm), but not so different in appearance than males; ratio lengh elytra/pronotum = 2.9-3.0 (in males 2.8×). Pronotum slightly more elongated (L/W 0.89); elytra widest behind middle, more expanded (L/W 1.61-1.66) and with more rounded (not pointed) apex; 2.7-2.8× length of pronotum. Hairs on elytra shorter and less dense, remaining longer in apical third but not concentrated at apex (tufts very reduced). Protarsi narrower. Spiculum ventrale with apically arcuate plate. Spermatheca with prominent collum and broad nodulus (Fig. 59I). etyMology. The specific term "denudatus" is a Latin adjective that means "naked" in reference to the normal appearance of the insect bearing usually very few scales.
reMarks. The series of the specimens studied stand out -including some with the deciduous mandibular processes-for having very few scales on their dorsum, showing their naked brilliant integuments, the sculpture of pronotum very pronounced, and the elytral striae strikingly marked by the large deep punctures. However, one specimen from near the type locality depicted by  shows part of the complete scaling, with a beautiful mix of whitish and bluish scales. Being a light-coloured insect, it turns blackish by losing the scales, but this could well be an exception. It is worth studying if the scales are really deciduous or if they are just just incompletely formed, as we suspect.
In H. denudatus n. sp. the differences in shape between males and females are less marked, in regards to other Herpisticus species, and as an exception the ratio elytral length / pronotal length in females may fall below 3.0 (2.9-3.0). distribution and ecology. Endemic to Gran Canaria and present at moderate elevations (200-400 m) in the western margin of the Amurga massif, in the South of the island (Fig. 33). Further data should confirm if this species inhabits the entire massif (difficult access to interior) or if it extends westwards into the neighbouring 'barrancos' (gullies). Amurga is another separate geomorphological sector identified by Menéndez et al. (2008), in relation to age of volcanic materials and dominant wet or dry zones. Herpisticus denudatus n. sp. has been collected at night feeding on Kleinia neriifolia, mainly in winter. Some imagos emerge already in November.
etyMology. The specific term "gigas" is a Greek noun in apposition which means "giant" in reference to the large size and robustness of the species.
reMarks. This species is easily recognised by its large size and inconspicuous very short pilosity of elytra free of longer protruding hairs -except the apical tuft-, which is the only case in Gran Canarian Herpisticus. The female has well marked basal lateral grooves on ventrite 5. Herpisticus scopulus n. sp. is also large with short pilosity on elytra, but bears some isolated longer hairs scattered on apical third.
In the Cabrera Collection, kept at the MNHN in Madrid, there is a series of 6 males (14.2×5.5 mm -1.49×5.6 mm) and 10 females (16.6×7.0 mm -18.3×7.8 mm) with a printed label reading: "Tenerife /La Laguna XI. 1928". Cabrera used labels handwritten by him or his daughter, and not printed labels. Similar printed labels were common use in curatorial work at the Madrid Museum, and Canarian coleoptera specimens, for instance those of Escalera or Bolívar-Bonet, are at present mixed with the Cabrera specimens in his Collection. Part of these materials (including Cabrera's) were not prepared, and it would not be the first case of mislabelling errors attributed to an assistant who worked at the museum, particularly in those years (see Machado 1992: 42). There is one more specimen of H. gigas at Madrid, from Barranco de Azuaje in Gran Canaria, dated 5-5-1934 (anonymous collector), a locality in which the species was found at least by Uyttenboogaart in October 1927 (two females kept at Leiden, in the NMNH). Uyttenboogaart visited and studied "most of the anonymous Coleoptera" from the Canary Islands at Madrid in 1935(Uyttenboogart, 1937. He lists the locality Azuaje, among others, and makes comments about the considerable individual variation of Herpisticus eremita, from very small males measuring 11 mm, to the largest females of 19-20 mm [head included]. He apparently measured females of H. gigas, but did not say how many specimens there were, or if he was referring to those in Leiden or Madrid, but surely not those from Cabrera (mostly from Tenerife), as his collection arrived in Madrid as a legacy, after his death in 1943. Azuaje, in the municipality of Firgas, was a popular health spa -visited by Uyttenboogaart several times before it closed in 1938-and it is a plausible origin for the mysterious series under discussion. Perhaps it is related with the post-congress field excursions of the 6 th International Congress of Entomology followed by about 30 entomologists, including Uyttenboogaart, who visited Tenerife and Gran Canaria in September 1935, and part of the insects collected ended up in the Madrid Museum. In any case, it is unlikely that the series originated in La Laguna, Tenerife, an area very well prospected for decades.
etyMology. The specific epithet guayarmina is a noun in apposition corresponding to the name of an aboriginal princess of Gran Canaria, daughter of the 'guanarteme' (king) Tenesor Semidan, in the 15th century.
reMarks. The elytra of H. guayarmina n. sp. are more acuminate at their apex and beset with scattered long hairs in both sexes, never as dense as in H. subvestitus or H. grancanariensis, two species of similar appearance and size that can safely be separated by inspecting the penis characterised by a strongly recurved apex (moderately reflexed in H. guayarmina). It is more similar to its sister species H. tasarticus n. sp., particular in pilosity, but it is more narrow-cylindrical, the rostrum is a little longer and more parallel, the pronotum more curved laterally, elytra more elliptical and acute apically, and crenulation of meso-and metatibiae is more developed.  These differences in the external morphology are not impressive despite the 10.8% COI K2P distance among them (Fig. 3, clade H). However, the median lobe of penis in H. guayarmina is much shorter and arcuate (long and straight in middle third in H. tasarticus) and the apex is less curved and blunt at tip (not sharp-pointed). Both have large tarsi. distribution and ecology. Herpisticus guayarmina n. sp. is endemic to Gran Canaria, having been found to this day only in La Isleta, at the NE extreme of the island (Fig. 33). There is also another spotendemic entiminae weevil species, Laparocerus franzi Machado, 2012, that is apparently restricted to this same islet of about 850 ha and connected to the main island by an isthmus 4.2 km long by 200 m wide, occupied at present by the harbour and city of Las Palmas. Half of the islet is nowadays a protected area.
H. subvestitus grancanariensis is also present in La Isleta, at least on the western beach of El Confital, where it shares habitat with H. guayarmina n. sp. However, the former is found occasionally in early summer while the latter is rather common in late summer and widely spread in the entire islet. description. Herpisticus of large size (body size ♂ 12.1×4.5 mm -13.8×5.2 mm); oblong shape (elytra/prothorax length ratio 2.88); blackishbrown in colour, with variable mottled pattern of brown, pinkish and glaucous scales; scales round, separate, apparently rather dehiscent (integument more or less flattish corneolate); pilosity on elytra formed by small recurved overlapping hairs increasing in size apicad, mixed with separated erect long silky ones (about or longer than tarsal claw) conspicuous on apical fourth (apical tuft little developed); hairs inconspicuous on head and pronotum; little protruding on tibiae (smaller than its diameter). Antennal flagellum 2.1× length of scape; club L/W 3.1. Eyes oval (L/W 1.12), little protruding (convexity 18-21%). Rostrum subparallel (L/W 0.82-0.88) with lateral margins blunt. Pronotum (L/W 0.84) with arcuate/curved sides, widest about middle; angular tooth small, inconspicuous; dorsal sculpture usually well marked. Scutellum broad ogival. Elytra oblong, moderately acuminate, widest slightly behind middle (L/W 1.8), 2.6× length of pronotum; punctures of striae rather conspicuous, interstriae subconvex, about as wide as striae. Abdominal convexity 75%. Protibiae slightly sinuous basad, as long as metatibiae, with strong denticles; meso-and metatibiae with small denticles. Tarsi broad, robust, with tarsomere 2 a trifle longer than wider (tarsomere 3 as long or shorter than length of club). Spiculum relictum of sternite VIII with base narrower than length of arms (Fig. 52H). Penis apex moderately reflexed, wider than longer, triangular and pointed in dorsal view; temones 0.7× length of tube; tube straight in middle third (Fig. 53M); endophallus mid-section broadly cordiform (Fig. 55G): tegmen thin between parameres, with incipient median projection; manubrium long (Fig. 57G).
reMarks. Herpisticus tasarticus n. sp. can be distinguished from H. subvestitus grancanariensis by its more parallel rostrum, less convex eyes, and less developed pilosity of tibiae (less protruding on outer face), and from H. subvestitus subvestitus, by the small recurved pilosity on basal half of elytra and absence of tufts on abdominal ventrite 5. Additionally, the manubrium of tegmen is proportionally much longer (Fig.  57G) and the apex of its penis is less curved than in these species or H. lanatus, but more than in any of the other Herpisticus inhabiting Gran Canaria, including H. guayarmina n. sp., its sister-species. The latter looks rather similar in shape and shares robust tarsi, but bears (♂) long hairs on basal half of elytra and the rostrum is anteriorly more squarish. Scales in H. tasarticus n. sp. seem to get lost easily, thus dominating the dark colour (integument with corneoles), making the elytral puncturation very conspicuous. distribution and ecology. Endemic to Gran Canaria (Fig. 33)    description. Herpisticus similar to H. tasarticus, but larger (body size ♂ 12.4×4.5 mm -13.2×4.7 mm); with variegated greenish-beige colouration on dark background; with whitish scales usually persistent on inner side of femora; integument moderately corneolate, shiny; scales round and tangent; elytra densely beset with small recurved hairs mixed in apical third with slightly larger curved/ bent hairs, and a few scattered longer arcuate hairs pointing backwards near apex (many more, erect and longer than a claw in H. tasarticus); apical tuft little developed; outer edge of protibiae with suberect hairs (not recurved) smaller than tibial diameter. Rostrum slightly convergent, with straight lateral margins (L/W = 0.85), squarish anteriad; ratio width at base / inter-ocular distance < 1.10 (in tasarticus >1.10); eyes similarly little protruding (convexity 18-22%). Pronotum less transverse (L/W = 0.9), with arcuate sides (not rounded), widest at middle or before, and with postero-lateral tooth little marked; disc uneven; lateral oblique impressions and discal points well marked. Elytra slightly more elliptical (L/W 1.85), less convex at base; 2.7× length of pronotum; basal rim not very  elevated near scutellum; apical declivity starting at 1/3. Probitiae with a dozen denticles (alternating in size); mesotibiae with small denticles, metatibiae with granules. Tarsi robust. Spiculum relictum of sternite VIII with broad base and thin, short arms (Fig. 52L). Penis shorter, more arcuate, with apex less reflexed, blunter in dorsal view (Fig. 54D); endophallus mid-section cordiform; apical section slender and bilobed (Fig. 56C). Female as male, but larger and broader (length 15.5×6.2 mm -16.1×6.05 mm); scales more separated; elytral recurved pilosity smaller, less dense, with protruding long hairs more spread throughout apical half (almost in rows); ventrite 5 with scarce pilosity and lateral grooves little impressed. Meso-and metatibiae without denticles or crenulation. Spermatheca with flatish ramus (Fig. 59M).
etyMology. The Latin noun scopulus (escarpment) has been taken in apposition to name this species which lives in the almost inaccessible escarpments of El Risco-Faneque, on the NW coast of Gran Canaria.
reMarks. Besides H. tasarticus n. sp., the other Gran Canarian species that can easily be confused with H. scopulus is H. gigas n. sp. However, it is larger and the females are more inflated; the pronotum is a little more constricted anterior and basally, and the lateral impressions are shorter and less oblique, almost transverse; the elytral apical declivity starts at the middle and the pilosity is less developed than and inconspicuous, almost devoid of protruding hairs on apical third (very few in males, none in females). Moreover, the apex of the penis is even less reflexed, almost straight. distribution and ecology. Herpisticus scopulus n. sp. is endemic to Gran Canaria, being restricted to the ancient basalt Tamadaba massif with its impressive cliffs of El Risco-Faneque in the NW of the island (Fig. 33). Other entiminae weevils, like Laparocerus rugosivertex Machado, 2012, or Laparocerus fraudulentus Machado, 2012 are also confined to this region. It is a summer animal that feeds during the night on Salvia canariensis (common) and possibly Cneorum pulverulentum; during the day it was collected under stones or hiding below flat mats of Lotus callis-viridis.
etyMology. The specific epithet "nanus" (dwarf in Latin) refers to the small size of this species in comparison with its congeners on Gran Canaria. It is to be treated as a noun in apposition.
Remarks. This species is easily recognised, apart from being the smallest of the genus on Gran Canaria, by its cylindrical shape, convergent rostrum, protruding eyes, and scarce upstanding long pilosity on its dorsum, including head and legs. It could be confused with H. lanatus, but pilosity in this latter species is much longer and denser, usually irregular (twisted) and the tip of its penis is clearly more curved than in H. nanus n. sp.
The Herpisticus from Punta de las Arenas (30-5-1988) recorded by García & Peña (1996) belongs presumably to this species, but unfortunately this single specimen has been lost. After two failed efforts of Canarian colleagues in May and June 2019 to find it at its locus classicus, success was obtained in September (three specimens). The series of H. nanus n. sp. now available is limited to six specimens (one male with a deformed head). The light colouration makes this animal difficult to spot on sandy calcareous soils.
reMarks. In the Germar Collection kept at the Martin-Luther-Universität at Halle (Germany), there are nine specimens under the heading Herpisticus laesicollis, but only two males and two females belong to this species. The only specimen that bears a label "Teneriffa / V. Rebeuer" is a male here designated as lectotype; the other three as paralectotypes. From the rest of the specimens, two males (one bearing a label "G. Canaria") belong to H. subvestitus grancanariensis Palm, 1974, two females without labels to H. bobadillae n. sp.-which probably originate in the collecting excursion of Walther May to La Gomera, in February 1908(May, 1912 -and one male that could not be identified with confidence. The DNA sequences of specimens of H. laesicollis originating in the NW of Tenerife (Anaga, Santa Cruz, Radazul) cluster separatedly from similarly shaped specimens from the rest of the island, which form two groups: one with long hairs over the entire elytrahere described as H. aridicola n. sp.-in the arid south and west (13.4% K2P distance), and another, H. daute n. sp., on the central north and northwestern slopes of the island (12.8% K2P distance). See remarks in these species.
Specimens from the central windward parts of the islands, at intermediate elevations (Montaña Bermejo; Fuente Fría; Las Lagunetas, etc.), show the pilosity on the apical half of elytra more developed than Anaga specimens, but keeping the laesicollis pattern (half short / half long). However, sequenced specimens from the Orotava Valley cluster with specimens of Teno as H. daute n. sp. This pattern of two related species in Teno (west) -Anaga (east), with somewhat morpholo gically intermediate species/subspecies in the younger terrains in between is common in several beetles endemic to Tenerife (Carabus, Eutrichopus, Laparocerus, etc.) and it is likely to depend on which refuge-area, Teno or Anaga, played a more important role in the re-colonisation of the intermediate parts (v. Machado, 1976: 393).
Specimens from the Valley of Güímar and sothwards of it, on the island leeward side, are in general of paler colouration, less slender, with slightly more covex eyes, and somewhat less developed pilosity. They could represent another differentiated subspecies (4.3% genetic distance) and merit also further study.
distribution and ecology. Herpisticus laesicollis is endemic to Tenerife where it is distributed in the Anaga massif, expanding westwards along the windward north slope of the island towards the Orotava  Valley, and on the leeward side until the El Río ravine at the boundary of Arico municipality (Fig. 37). It lives in xerophytic open environments, mostly in the warm lower and intermediate zones, but easily reaching 1000 m altitude in dryer exposed spots within the forested areas. Found mostly under stones, or beaten from different plants during the night when it is active: Artemisia, Aeonium, Dittrichia, Sonchus, Lavandula, Cynara, etc. Occasionally, it can develop as a pest to crops like avocado trees (Persea americana), mango trees (Mangifera indica) or vineyards (Vitis vinifera).
etyMology. The specific epithet is the name Daute in apposition, one of the aboriginal kingdoms of Tenerife, where the species is present.
reMarks. This species looks very similar to H. laesicollis despite the 12.8% KP2 genetic distance among them. Females are easier to distinguish because the denticles of metatibiae are usually not developed and elytra are beset with only short hairs and no trace of longer ones. The male elytra are more uniformly convex at base and the pilosity may be more or less developed in size, hairs are curved or recurved, not longer than a tarsal claw, and are spread over the entire elytra, increasing uniformly in size towards the apex, whereas in H. laesicollis there is a clear contrasting difference between the curved short hairs in the basal half of the elytra and the longer hairs in the apical half. Moreover, females of H. daute do not bear protruding erect hairs among the layer of recurved / curved hairs. Doubts can be solved by inspecting the apex of the penis (Fig. 54F), which is more acuminate in H. daute n. sp., with sides sinuous like in H. aridicola n. sp., whereas it is ogival with arcuate sides in H. laesicollis / Fig. 54E). distribution and ecology. Endemic to the island of Tenerife, where it is distributed on its northern slope from Agua García (Tacoronte) in the middle to Teno massif in the extreme NW (Fig. 37). It lives from the coast up to intermediate elevations (ca. 1000 m altitude), on shrubby vegetation (Fig. 49C, type locality) or at open spots in forested areas. It is more common in Teno than in other areas, but Masca ravine seems to mark its southern limit, being replaced by H. aridicola n. sp. southwards. It has been collected at night by beating Artemisia thuscula, and less common on Aeonium pseudourbicum, Ditrichia viscosa or Cistus monspeliensis.
etyMology. The name aridicola is a noun in apposition formed by the conjunction of the Latin terms "aridus", an adjective meaning arid or dry, and the ending "cola" meaning inhabitant, in reference to the arid habitats of Tenerife where the species dwells.
reMarks. The hairy appearance of H. aridicola n. sp. resembles in extreme cases that of H. lanatus from Gran Canaria or H. hispidus from La Gomera. From the former it can be distinguished by the denticles of male protibiae being smaller and more uniform, the larger punctures of elytra, and the less recurved apex of the penis; and from the latter species by its more convergent rostrum with blunt apical angles, not parallel with squarish angles.
The other Herpisticus species from Tenerife have no long suberect hairs on the basal third of the elytra. Specimens of H. aridicola n. sp. with somewhat reduced and less silky pilosity from the western slope of the island-entering the Teno massif-get closer in appearance to H. daute. However, the pilosity in H. daute is denser with curved/ recurved hairs shorter than a tarsal claw, and void of protruding longer erect/ suberect hairs (at most a few near the apex in the male elytra). Moreover, the male of one pair from Agua Mansa (15-6-1927 leg. A. Cabrera;MNCN) has long hairs over the entire elytra and should be attributed to H. aridicola, whereas this locality-if correct-is on the humid north-side of the island. The female bears a pre-apical tubercle on each elytron (teratology?). distribution and ecology. Herpisticus aridicola n. sp. is endemic to Tenerife (Fig. 37), where it is distributed in the lower and intermediate calcareous districts (0-550 m altitude) of the southern and western parts of the island, which are the most xeric habitats, characterised by the presence of Euphorbia species and Launaea spinosa. In these orientations, it can reach the limits of the pineforest (ca. 1000 m altitude). It has been collected on Schizogyne sericea, Artemisia thuscula, Kleinia neriifolia, Dittrichia viscosa, Retama monosperma Foeniculum vulgare, Aeonium pseudourbicum, Cneorum pulverulentum, Bystropogon origanifolius and in great numbers on one isolated almond tree.
etyMology. The specific epithet is an adjective that refers to the species inhabiting the island of La Gomera.
reMarks. Herpisticus gomerensis n. sp can easily be separated from H. bobadillae because of its long silky hairs on the elytra, but it is rather difficult to distinguish from H. hispidus n. sp., which has an even more developed pilosity (see remarks on this species). Surprisingly, it clusters (100 ppb) as clade R with clade S from El Hierro and La Palma (7.7-7.8% K2P distances) and not with the Gomeran lineage (clade M) of H. bobadillae and H. hispidus n. sp., showing 11.8% and 12.4% K2P distances, respectively. distribution and ecology. Herpisticus gomerensis n. sp. is endemic to La Gomera and lives in the western half of the island, including the wide valleys of Vallehermoso, Gran Rey, and Alojera, from the coast to the higher parts (1000 m altitude). Specimens were collected in open shrubby habitat, under stones, beaten during the night from Retama monosperma, and sifting dry hanging leaves of Sonchus gomerensis. Some old specimens originate from the valley of Hermigua, where H. bobadillae is also present (western limit). See Fig. 41. description. Body size ♂ 10.5×4.0 mm -12.4×4.6 mm. Much similar to L. gomerensis in shape and colouration, with scales more separated and pilosity of elytra more strongly developed. Long hairs as long as tarsal claw at base, increasing in length (2-3× tarsal claw) and often irregular towards apex, of hairy appearance if raised up, like in L. aridicola n. sp. Antennal flagellum 2× length of scape; club (L/W 3.0); rostrum (L/W 0.87) squarish at apex, parallel-sided or a trifle divergent apicad; eyes slightly oval (L/W 1.1), convexity 20-22%. Pronotum less broad (L/W 0.9); slightly more constricted anteriorly; pre-and post discal depressions conspicuous; lateral impression strongly marked; posterior angle obtuse, without protruding tooth. Elytra (L/W 1.74) less convex at base, with slight posthumeral depression; 2.6× length of pronotum; widest behind middle; declivity at apical third; striae as broad as interstriae; interstriae a little subconvex. Scutellum U-shaped. Protibia longer than metatibiae (1.1×), with strong denticles alternating with small: meso-and metatibiae with denticles. Spiculum relictum of sternite VIII (Fig. 52P) with arms shorter than its base and with apical sclerotisation elongate. Penis tube with tip of apex triangular and blunt, a little reflexed (Fig. 54I): endophallic midsector globose and distally a little concave (Fig. 7H). Tegmen without sclerotised dorsal projection between parameres; ring very broad (Fig. 57C).
etyMology. The specific epithet hispidus ("hairy" in Latin) is and adjective that refers to the appearance of the males.
reMarks. Despite the morphological resemblance of Herpisticus hispidus n. sp. with H. gomer ensis n. sp., there is a high enough COI genetic K2P distance (12.4%) to consider it a separate species, although somewhat cryptic. The genetic distance with H. bobadillae n. sp., to which it is related as sister-species in our phylogram (Fig.3, clade N), is lower (8.3%), but the absence of long pilosity on the male elytra is sufficient to separate the latter. The extreme reduction of the elytral pilosity in the females is noteworthy, being absolutely inconspicuous (bare appearance). All three Gomeran species have the penis apex a little reflexed, and there are no evident differences except for the shape of the median globose part of the everted endophallus, being concave anteriorly and slightly bilobed in this new species (Fig. 7H)-not in H. bobadillae-, and inverted-trapezoidal in H. gomerensis n. sp. (Figs. 56F). Additionally, the tegmen has no median dorsal projection between the parameres (developed in the other two species). In cases, the more squarish and parallel rostrum (sometimes with divergent sides) of H. hispidus n. sp. may help in separating it from rather hairy specimens of H. gomerensis (rostrum slightly convergent). distribution and ecology. Herpisticus hispidus n. sp. is endemic to La Gomeraand seems to be restricted to the southeastern sector of the island (Fig. 41), like Laparocerus benchijigua Machado, 2007or L. dilutus Machado, 2014, which are also apterous Gomeran endemic weevils. It does not overlap with H. gomerensis n sp. or H. bobadillae n. sp., and is distributed from the arid spurge communities near the coast up to the rim of the watershed at 1000 m altitude, close to the forest limit. An aggregation of many specimens was found on a group of sun-exposed stones on a small grassy clearing in the Erica-Morella forest border (Fig. 49F). Other specimens have been obtained by sifting below Rumex lunaria and Cistus monspeliensis, or beating Cistus or Kleinia neriifolia directly at night.    1♀ 9-12-2006;Puntallana (28R 2933Puntallana (28R 31131) 3 exx 7-4-2009Hermigua 1♀ 8-8-1987 leg. P. Oromí (POM) description. Body size ♂ 10.0×3.9 mm -14.3×5.2 mm. Similar to L. gomerensis, often with scaling dominantly darker; with denser cover of small recurved hairs which barely overlap, and only few straighter longer hairs near apical tuft. Antennal flagellum shorter than twice length of scape (1.8-1.9); club (L/W 3.0); rostrum squarish, parallel-sided (L/W 0.90); eyes round, convexity 17.5-20.0%. Pronotum with less arcuate sides (0.82); posterior angle obtuse, without protruding tooth; dorsum uneven. Elytra (L/W 1.70-1.74) less arcuate laterally; interstriae more convex (particularly 3 and 5); punctures smaller. Protibia longer than metatibiae (1.05-1.08×), with at least 6 strong denticles alternating with small ones: meso-and metatibiae with denticles. Spiculum relictum of sternite VIII with arms longer than its base and apical sclerotisation (head) elongate (Fig. 52R). Penis tube with apex broadly acuminate (tip blunt, sides sinuous), slightly reflexed at tip (Fig. 54J); tegmen with sclerotised dorsal projection between parameres; ring broad ( Fig. 57J).
etyMology. Species named in honour of Beatriz de Bobadilla (1462-1501), Spanish Overlady of La Gomera who had sentimental affairs with Christopher Columbus during his provisioning stays on the island before sailing to the Americas.

Herpisticus hierrensis hierrensis
reMarks. Franz characterised H. hierrensis by flattened eyes and by "antennae inserted after mid rostrum (in H. eremita inserted well away from mid rostrum and reaching slightly anterior margin of eye)". This latter observation is clearly erroneous, but the eyes are indeed much flattened.
The genetic COI K2P mean distance between specimens from El Hierro and La Palma is too low (3.5%) to consider them different species. In fact, this seems to be the only confirmed case of a species inhabiting simultaneously two of the main islands, which are actually the youngest of the Canary Islands, with 1.12 Ma and 1.72 Ma, respectively (Carracedo & Troll, 2016). La Palma specimens have slightly more convex eyes (17-19%) and the pronotum is not so depressed behind the front margin as in specimens from El Hierro. Besides these little differences, the other characters match except in specimens from the interior of the Caldera de Taburiente and the north and northeast slopes of La Palma. Their elytra are beset with suberect hairs (about the length of a tarsal claw) instead of being uniformly covered by short curved pilosity as on El Hierro or SW La Palma. On Gran Canaria, Tenerife and La Gomera there are species with and without long hairs, and it seems that this tendency also occurs on La Palma. To recognise this important morphological differentiation, we split Herpisticus hierrensis into two subspecies. Considering that the other subspecies -H. hierrensis benahoare n. ssp.-is restricted to the older parts of the island, it is likely that El Hierro was colonised by specimens originating in the southern and youngest part of La Palma. distribution and ecology. Herpisticus hierrensis hierrensis is endemic to the islands of El Hierro (Fig.  43) and La Palma (Fig. 45). It seems not to be very common, at least on the first island, where it is the only Herpisticus species present. Franz found the type near the western coast, below a trunk of Euphorbia canariensis, and thereafter not many specimens have been collected in the lower arid coastal zone but rather at intermediate elevations (600-700 m), under stones in mountain shrubby habitat with Cistus, Asphodelus, etc.). On La Palma, all recorded specimens originate from the southwest of the island, mostly from leeward arid environments; while the rest and older parts of the island are apparently inhabited by the other subspecies, H. hierrensis benahoare n. ssp., but not at high elevations. It has been collected on Artemisia thuscula, Kleinia neriifolia, Euphorbia lamarckii, Lavandula pinnatifida or just below stones. Herpisticus eremita (pars), in : 61. Herpisticus eremita, in Cosens, 1964: 63. Herpisticus cf. laesicollis, in Stüben et al., 2015 [DNA]. differential diagnosis. Similar to nominotype subspecies, of moderate size (body size ♂ 11.3×3.9 mm -11.9×4.0 mm) but elytra with curved hairs less depressed and slightly longer (2-3× diameter of scale), and with at least the apical third beset with many additional suberect hairs about length of a claw or longer). Rostrum less convergent apicad; eyes a little more prominent (convexity 17-21%); elytral punctures broader and deep (foveiform); protibiae somewhat longer than metatibiae (1.05×), with tiny denticles except 1-2; mesotibia and metatibiae at most with granules (not crenulate). Penis with apex less reflexed (Fig. 54L). Tegmen with arms and ring narrower. Females larger and broader than males (body size 12.5×4.8 mm -14.6×6.0 mm), with interstriae a little more convex. Plate of spiculum ventrale with apical margin slightly angulate (Fig. 58P).
etyMology. Benahoare is the aboriginal name of La Palma, here used as a noun in apposition as the specific epithet.
reMarks. The genetic distance between this subspecies and the nominal subspecies is not much (2.0-3.9%) in relative terms for Herpisticus (see comments under H. calvus). However, the morphological differencesparticularly the development of long pilosity on the elytra-and the allopatry of the two populations justify the subspecies status proposed here.  distribution and ecology. Herpisticus hierrensis benahoare n. ssp. is endemic to La Palma, living at low altitudes (0-300 m) in the oldest part of the island (northern half), including the central giant Caldera de Taburiente (Fig. 45). The records are too scarce to infer its real altitudinal range and its distribution on the windward side of the island. One isolated specimen from the SE coast (Mazo, Salemera) contradicts the allopatry of the subspecies and such status, but having been collected near a manureheap in a banana plantation, an occasional introduction from the north cannot be discarded. When new banana fields are planted in the traditional way using "heads" (basal part of the trunk bearing sprouts), they are brought from a different remote site within the island; moreover, animal manure is also usually produced and sold in the north, where cattle are abundant.

Key to Herpisticus (♂)
In this key all references to pilosity refer to males. If sex is doubtful, length of elytra /length of pronotum <3 applies to males, and > 3 to females (except H. denudatus 2.9-3.0).  Palm, 1979 Additional remarks There is no information published on the natural history of Herpisticus. To undertake such a study would imply at least two years, as it did with the phenology and biology of Laparocerus (cf. Machado & Aguiar, 2005. However, the empirical information gathered from the copious material examined and from our field experience allows us to infer a broad image of the natural history of this weevil. phenology. We summarised the number of collection specimens for each month in Fig. 46, disregarding those collected in 2018 and 2019 in excursions focused on Herpisticus, which are not comparable with previous years in which these weevils were taken collaterally. There is a peak in May, and a minimum in September. However, this pattern may be somewhat biased as insect collecting periods in the Canaries are concentrated in winter and spring, and less in summer (dry season). The absence of young imagos bearing the mandibulary deciduous appendix (Fig. 47) in January and February, with a peak in March, suggests that Herpisticus is not a winter animal as are most Canarian Laparocerus and other Entiminae.
Moreover, our recent field experience reflects that it is after April and mainly in July-August when Herpisticus appears in numbers in the countryside, as to justify its appellative as summer animal. However, this should be taken as a general tendency and may not apply to all species (e.g. H. denudatus n. sp. emerges in winter). In the Canaries, the phenology of insects is not as sharply tuned as in continental habitats, depending greatly on the shifting first and late rains of each year and on the varied mesoclimate conditions within each island. Herpisticus can be found in any month, but more probably in late spring and early summer, after their emergence.
biology. The imagos are strong animals, long-living, perhaps a whole year (kept in terrarium at least 8 months), and can resist without food 12-47 days (average 27, n = 10). They are active mainly during the night, but we have seen them also moving and feeding during the day, mostly in cultivated fields and greenhouses, less in natural habitats. With daylight, they hide under stones (on the ground or clinging to their underside), in the bunches of dry leaves that hang from plants like Sonchus, Aeonium, or Kleinia, or they just keep still, hanging on to the petiole of leaves, small twigs, below loose bark or any crevice where they feel sheltered.
Herpisticus is very active sexually and is commonly found in pairs copulating or just with the male resting on the females for hours. In the Güímar valley (Tenerife) we recorded a popular name for Herpisticus laesicollis as 'el gorgojo follador', that means something like 'the horny weevil', but this would certainly fit all Herpisticus.
During the copula of H. subvestitus pseudolanatus the male introduces 2/3 of the penis and from time to time makes a push, keeping almost parallel on the female, not as oblique as in Laparocerus. We tried to obtain eggs by providing the terrarium with earth, seedlings, plant leaves, sticks and Van Emden fans (v. Machado & Aguiar, 2019), with little success. On 5/12/2018, five white eggs (3+2) were laid in nooks of the plastic cover of the terrarium. In another terrarium with two pairs of H. denudatus n. sp., we found one white egg loose on the plastic bottom; two eggs with a slightly greenish tint adhered in a nook of the cover, two eggs (white and dark) exposed on the earth, and three blackish eggs when washing the Kleinia leaves provided as food. The terrarium was not inspected for a week, and we conclude that the eggs were not laid the same day and they turn from white to black as they mature, as in Sitona and many other broad-nose weevils that drop eggs at random, singly and without cover (Van Emden, 1950;Marvaldi, 1999a). It seems that Herpisticus ovoposites without making any special arrangements, gluing the eggs free on surfaces -category 3 sensu Anderson & Lyal, 1995-but we so far do not know if it is on the plant, the ground or both options. Its very short oviscapt supports this idea. The number of eggs in several inspected females does not exceed ten, five in each ovarium. This is a rather low number for a large-sized weevil, particularly compared with Laparocerus, where eggs can reach a hundred kept in the abdomen. The low number of eggs in    Herpisticus, if confirmed by more observations, is likely to be compensated by more repeated oviposition during their life. This is a topic that merits further studies.
Field larvae of H. subvestitus pseudolanatus placed in November 2018 in a pot with a young pepper-plant, ended in one pupa (24/3/2019, killed for drawing) and an imago (1/4/2019), and there were still two large but inactive larvae in the earth. Four to five months is a plausible development time for a weevil of the size of Herpisticus, in which diapause is likely to play an important role.
parasites. It is not uncommon to find ectoparasitic mites on Herpisticus, especially in the dry summer months. They are normally attached to the underparts of the body and more rarely on the legs or margins of the elytra (Fig. 48). Some heavily infested specimens may hold more than 30 mites on them, but they apparently do not greatly harm the weevil´s activity. Leptus (Leptus) maxorata Haitlinger, 2009 (Acari: Prostigmata: Erythraeidae) was described on specimens obtained from Herpisticus calvus on Fuerteventura, and at present Dr. Haitlinger (Wroclaw) is studying material provided from other islands, which presumably belongs to several new species.
hábitats. Herpisticus species live in open exposed habitats free of the shade of trees or very dense bushes. When they are occasionally found in forested areas or thickets, it is in clearings, along the margins of dirt tracks, or on stony outcrops exposed directly to the sun. Within this limitation and avoiding high humidity, they are rather euritopic. Some species range from the coast to the high mountains (maximum about 1,600 m altitude), where snow can be present in winter. Clayish and calcareous soils are clearly their preference; bare lava or lapilli fields are avoided, but sandy areas count with at least two apparently adapted species: H. rectipes n. sp. and H. nanus n. sp. Figure  49 shows a sample of the varied natural habitats where Herpisticus is present, and several species are not reluctant to enter anthropic habitats (abandoned plots) or cultivated fields where pesticides are not used (organic agriculture).
Herpisticus is not so abundant as Laparocerus, both in number of species -19 versus 216-and of individuals. Their presence in nature is also less uniform,   showing an aggregate distribution: the habitat may be vast and continuous but Herpisticus appears in a limited area, and the next group may be very distant. It is also uncommon to beat down more than a pair of Herpisticus from the same plant, while in Laparocerus there can be dozens of them.
food plants. Entimine larvae dwell in the earth and feed freely on roots of plants, thus its scarce specificity compared with endophytic larvae of other weevils. Palm (1974) reports large Herpisticus larvae in April, feeding on roots of Aeonium canariense in Tejeda, and we have seen larvae of H. subvestitus pseudolanatus eating roots of pepper plants, pumpkin and lettuce.
Adults are moderately polyphageous, but it is not uncommon to find them in a given locality feeding only on one species of plant, while in  another locality they select a different species. This may be related to the roots they fed on when growing up as larvae.
With some practice, Herpisticus feeding marks on the leaves of plants can be distinguished from those made by other phytophagous insects. They never start feeding from the inside of the leaf leaving holes, like chrysomelids or acridids. They normally eat at the margin of the leaf and move quickly, leaving a set of separate notches trimming the leaf (Fig. 50E), and if they occasionally eat their way further inwards in soft leaves (e.g. Capsicum), then the openings are wide and short (Fig. 50B), not as a corridor that may bifurcate or form diverticula, as typical in Laparocerus (Fig. 50A), by far the most common and abundant genus of short-nosed weevil present in the Canary Islands (Machado, 2003). The notches and galleries made by microlepidoptera (Fig. 50D) start from the interior of the leaf or margins increasing progressively in width, and those of Naupactus are much broader and deeper as they stay longer gnawing at the same place, and have usually a characteristic curved digitiform shape (Fig. 50) The list of feeding plants registered for Herpisticus is not so extensive (Table 1) as it is likely to become. This may derive from the fact that collecting at night-when the adult feeds actively-is a recent practice in the Canaries, and in the past most specimens have been found during the day below stones or sifting plant debris, their refuge-niches. Its presence on several cultivated plants speaks in favour of a wider polyphagy (see next section).
In the wild, the plants most palatable to the genus are apparently Kleinia neriifolia, Artemisia thuscula, Patellifolia patellaris, and perhaps some broom species. Herpisticus can be present in barren areas where there is not a single green leaf available. Succulent plants like Kleinia or dendroid Euphorbia lose their leaves in summer, and other plants simply dry out. What Herpisticus eats in such situations is an intriguing question. In the laboratory they can last more than one month without food, but that may be insufficient to overcome a long dry season. Do they eat debris if needed?
Herpisticus as a pest. On occasions, we receive notice of Herpisticus producing damage to agriculture crops (eg. Puntagorda, on La Palma; Güímar on Tenerife, Ingenio on Gran Canaria, etc.). There are also some newsletters from the Agriculture Service where Herpisticus is depicted as a pest of vines, mango, or avocado trees. However, in more technical publications (e.g. Rodríguez López, 1994) it is not included, as it probably generates more alarm than real damage; at least in these crops. A few leaves being eaten by this weevil does not really harm large trees or their fruits, and if it eats the inflorescence of mango trees, for instance, it may even trigger the generation of more flowers or favour less but bigger fruits. Nonetheless, real damage may happen during the first year when trees are newly planted and the weevil destroys the few sprouting buds or leaves. A severe case happened in productive mango orchards where weeds (mainly Patellifolia) were intensively removed, thus concentrating the weevil's attention on the mango flowers with ca. 90% crop loss (C. Suárez, pers. comm.). The farmers apply pyrethroid insecticides (e.g. cypermethrin) to kill them, until they become convinced it is not worth combatting them in successive years or weeding so thoroughfully. The same in vineyards (e.g. Candelaria in Tenerife), on citrus trees (Telde, Gran Canaria), or almond trees (Puntagorda, La Palma).
The situation is different on vegetables, particularly if they are managed as organic crops. We have visited some fields, usually in greenhouses, where Herpisticus thrives in high numbers and causes real damage to watermelon, green bean, cucumber, pumpkin and pepper plants. In Arinaga (Gran Canaria) adult H. subvestitus pseudolanatus was destroying the pepper seedlings (5-10%), basically by feeding on their soft stalk 1-2 weeks after planting, and presumably the larvae were also contributing by attacking the roots. These installations were embedded in the arid habitat of the weevil and with Patellifolia patellaris bordering the green-houses. In Maspalomas, H. lanatus larvae destroyed recently planted pumpkins so completely that the farmers abandoned their cultivation.  concentrated on the leaves and is merely an aesthetic problem.
Herpisticus has not been reported from ornamental plants in public or private gardens, except one case in a parking parterre, on Atriplex halimus. distribution. With the exception of Herpisticus hierrensis, all species are mono-insular endemics (islets not considered). Within each island, most taxa are or tend to be allopatric, even when living in the same kind of habitat. There are, however, cases where two different species can be found in the same place, mostly on Gran Canaria (e.g. H. guanarteme n. sp. and H. subvestitus subvestitus; H. gigas n. sp. and H. subvestitus grancanariensis, or H. denudatus n. sp. and H. lanatus). Apart from the topo-climatic influence (windward/leeward), the geological imprint in the distribution of this genus is remarkable and likely to be related to the volcanic history of each individual island. Gran Canaria has by far the most complex volcanic history and that could be a plausible explanation for its record of nine Herpisticus species in total.
The island maps compiled with all the material studied provide a reasonable idea of the distribution of the genus in the archipelago (Fig. 14 Palma). It is easy to recognise the almost empty forested areas in the central parts of the western islands, and a belt around where Herpisticus shows more presence. In our opinion, the scheme here presented is not definitive, and further new taxa could well appear in the future.
conservation. Herpisticus, like most phytophagous insects, have large populations with high recovery capacity, if enough habitat is available. Predators of Herpisticus are mainly spiders -large Gnaphosidae, Lycosidae and Latrodectes-, but also very likely vertebrates such as large Gallotia lizards, the Canarian shrew Crocidura canariensis Hutterer, López-Jurado & Vogel, 1987 (Eastern islands) and birds, despite only few cases have been recorded: red-billed chough Pyrrhocorax pyrrhocorax barbarus Vaurie, 1954 (Pais & García, 2000) and the introduced Mediterranean tree-frog Hyla meridionalis Boetter, 1874 (Cott, 1934). Other introduced animals like the cat, rats, snakes, and the continental shrews could be more problematic, but these animals usually stay around anthropic environments. Medina & García (2007) refer to Herpisticus as an incidental prey in feral cats on La Palma.
A risk for Herpisticus species, if any, could come from habitat restriction. The Canary Islands, with a surface of 7,747 km 2 , more than 2.1 million inhabitants (ISTAC, 2018 census), plus a floating population of 15.6 million tourists/year (average stay ca. 8 days),  The genus Herpisticus from Canary Islands is a densely populated archipelago that has been and still is intensively transformed. At the same time, the Canaries are considered a biodiversity hotspot, with a high number of endemic species. Many of these are obviously of conservation concern, mostly due to loss of natural habitat. This is not a plausible risk for most Herpisticus species, which have wide distribution areas and adapt well to environmental changes (e.g. agriculture). Moreover, an overlap analysis of the species-presence maps above commented with the network of protected areas of the Canary Islands (IDE Canarias) shows that all Herpisticus species, except H. denudatus n. sp. and H. oculatus, are present in at least one protected area. These two exceptions have no potential problems of habitat availability. Nonetheless, there are two species with very limited known habitat, which deserve comment: Herpisticus nanus n. sp. may be linked to the sandy habitat of Punta de Las Arenas, on Gran Canaria, and if this is the case-more prospective effort is needed-its distribution area would be only 0.3 km 2 . The area is in a good natural state, remote, with difficult access (only on foot or by sea), part of the Tamadaba Nature Park, and apparently protected from human occupation. But the area is small and exotic species like the recently introduced California kingsnake (Lampropeltis getula californiae (Blainville, 1835)) or rats could put the species at risk of extinction if they reach Punta de Las Arenas. It would be a justified measure to prevent visitor access to this area in the management plan of the Park.
Herpisticus guayarmina n. sp. seems to be restricted to the peninsula of La Isleta, like Laparocerus franzi Machado, 2012. The available natural or seminatural habitat for H. guayarmina n. sp. is less than 6 km 2 , half of it -the interior of La Isleta-is military property and used as such. La Isleta is at present a Protected Landscape, a conservation category that permits several uses. The coastal platform areas outside the military zone are continually visited by the public, since the port and city of Las Palmas already occupies the isthmus and also part of La Isleta. There is at present enough pressure on the habitat (litter, car driving, military manoeuvres, etc.) to recommend these two species-Herpisticus guayarmina and Laparocerus franzi-to be assessed for inclusion in the Canarian Catalogue of Protected Species, perhaps under the category Vulnerable, according to the Canary Islands' Law 4/2010. 15