Ctenolepisma almeriensis n . sp . of Lepismatidae ( Insecta , Zygentoma ) from south – eastern Spain

Ctenolepisma almeriensis n. sp., from the south–eastern part of the Iberian Peninsula is described. This species was determined previously as Ctenolepisma lineata (Fabricius, 1775), which is widespread over the south–western Palaeartic region. The main difference between the two species is the setation of thoracic sternites. In each bristle–comb of the mesosternum and the metasternum, macrosetae are arranged in a single row in C. lineata and in two parallel rows in C. lmeriensis n. sp. In the prosternum, the first species shows 1–2 irregular lines of macrosetae per comb, and the new species shows 2–3 lines. Based on other parameters of setation, a discriminant analysis was carried out to separate a group of Spanish specimens of C. lineata from another group of specimens of the new species. This analysis demonstrated the validity of the occurrence of double or single lines of macrosetae in thoracic sternites to distinguish between thetwo species.

To compare the new species with C. lineata, a total of 100 specimens were selected, 50 from each species (25 males and 25 females), each from a different sample and locality.Statistical analysis included six quantitative variables in each specimen, resulting in a total of 600 data.
A standardized principal component analysis was performed to obtain combinations of the six variables which account for most variability in the data.The most useful variables were then included in a discriminant analysis to determine whether C. lineata and C. almeriensis were significantly different.Mann-Whitney U-tests were also conducted to compare medians of the four groups.
Variables and their abbreviations used in the discriminant analysis for the differentiation between C. lineata and C. almeriensis: N-notum.Number of macrosetae of a posterolateral comb of the metanotum; N-pros.Number of macrosetae of a posterolateral comb of an antedistal comb of the prosternum; N-meso.Number of macrosetae of a posterolateral comb on the mesosternum; N-meta.Number of macrosaeta of a posterolateral comb on the metasternum; N-uro.Number of macrosetae of a lateral comb of the urosternite IV; D/a [Mt].Ratio distance between lateral combs of urosternite IV / width of a comb.

Introduction
Ctenolepisma lineata (Fabricius, 1775) is a widespread species of Lepismatidae native to the south of Europe and introduced in other regions and continents.Following revision of material that had been determined as this taxon from different countries, notable variability has been detected, to the point that it is reasonable to state that C. lineata is not one but a group of species.Several forms have been found within the Iberian Peninsula, the most widespread considered here as the typical C. lineata.A different form which occurs in the south-east Spain is described here as a new species.

Material and methods
As usual in this order, specimens were fixed in alcohol, and many were dissected and mounted in Tendeiro medium for microscopic observation to verify identification.The studied material is deposited in the following institutions: MNCN.Museo Nacional de Ciencias Naturales (Madrid, Spain); UCO.Dept.de Zoología, Univ. of Córdoba (Córdoba, Spain).

Habitat and distribution
This species is found under stones and bark, at the base of pine-trees or Juniperus shrubs.The habitat is similar to that of C. lineata in Spain, but the new species appears to tolerate a higher degree of aridity.It is found from sea level to 2,000 m at Sierra de Gádor and Sierra de los Filabres in the province of Almería.
C. almeriensis n. sp. is mainly seen in the arid bio-geographic province named "Murciano-Almeriense" (Rivas-Martínez, 1987) in south-eastern Spain, being more frequent in the south (province of Almería) where the rainfall is lower.It spreads over the Spanish provinces of Almería, Murcia, Alicante and Valencia, always on the Mediterranean slope.The species takes its name from the province where it is most abundant.

Description
Body length of females up to 13.2 mm, males up to 12 mm.Fusiform and relatively robust body, thorax slightly wider (up to 3.5 mm) than the abdomen base.Faint to distinct epidermic pigment, usually violet-brown, with a variable pattern of distribution; this pigment can be more intense on the hind part of body, on the basal or distal parts of the articles of the appendages, and on the head, or it can be nearly uniformly extended (except for a lighter tonality ventrally).Scales dorsally brown, yellowish brown, dark greyish, silvery grey or greyish-brown, darkish and often with iridescence; they can draw an almost uniform pattern of distribution or can be arranged in alternately light (yellowish brown) and dark (greyish) longitudinal lines, as in other species of the genus.
Setation of head as usual for the genus.Eyes black, composed of about 12-13 ommatidia.Antennae longer than body, up to 15 mm (maximum preserved).Maxillary palp with long articles, the distal one 0.9-1.2times longer than the antedistal and 4.7-9 times longer than wide (fig.1).Distal article of the labial palp more or less unilaterally dilated, shorter to slightly wider at the apex than long; it always bears five sensory papillae arranged in a single row (fig.2).
Pronotum with 8-9 + 8-9, mesonotum with (9)10-11 pairs and metanotum with 9-10 + 9-10 lateral bristle-combs of 3-7 macrosetae each.Trichobothrial areas of the nota situated on the last and penultimate lateral combs in meso-and metanotum, and on the last and the antepenultimate combs in the pronotum.Posterolateral bristle-combs usually with 7-12 macrosetae each.Thoracic sterna as shaped in figs.3-6, very similar to those of C. lineata, except for the features observed in their fields of macrosetae.The word "field" of macrosetae is used here instead of comb, to emphasize that the setae are not arranged in a single row.In the prosternum they are arranged in 2-3 irregular almost parallel lines (fig.3), and on the meso-and metasternum there are usually two more or less parallel and very close rows (figs. 4-5).The number of setae per "comb" is variable, but on the average it ranges from 8 to 20.The highest numbers of macrosetae are often observed in the antedistal combs.The number of combs on each sternite is also variable; there are usually 4-5 pairs in the prosternum, 2-3 pairs in the mesosternum and two pairs in the metasternum.However, the extension of these combs can produce a juxtaposition of the contiguous fields of macrosetae and therefore the number of perceptible combs is reduced.Consequently, in the metasternum it is possible to count only 1+1 large combs with more than 20 macrosetae each (fig.6).
Tibiae I (fig.7) 2.2-3.4 times longer than wide; metatibiae 3.4-4.5 times.Apart from usual setae, there are some plumose macrosetae whose length is shorter or equal to the diameter of the tibia.The number of such macrosetae is usually 2-4 dorsal and 3-6 ventral in all tibiae.On the inner side there are many lanceolate scales that are absent on the outer side.These scales have also been detected in the proximal article of the tarsi.Hyaline short spines are usually present ventrally on the outer side of the article (figs.7-9), as in many specimens of C. lineata from the Iberian Peninsula, although these spines are more numerous and even shorter and stronger in this new species.Up to 25 spines have been observed on a hind tibiae (fig.9), but in other specimens the spines are absent.
Apex of the outer side of the femora covered by elongate and lanceolate scales, the inner side with many scales shortened and with truncate or emarginated apex.
Urotergite I with 1+1 combs, II-VII with 3+3 combs and VIII with 2+2 combs.Submedian bristle-combs with 7-9 macrosetae each, lateral combs with 6-10 and sublateral with 7-15 macrosetae.Urotergite X subtriangular, short, with a rounded apex that can show an angled or rounded point, more or less prolonged (figs.10-13).Urosternites I and II without setae, III-VIII with 1+1 lateral bristle-combs usually with 14-27 macrosetae each; young specimens may bear fewer than 14 and the largest specimens may bear more than 27 macrosetae, but this is not usual.In relation with the high number of setae per comb, the distance between the lateral combs of a urosternite is 2.7-5 times wider than the width of a comb.
Both sexes with two pairs of stylets.In males the inner process of the IXth coxite is slightly longer than wide (ratio length/width = 1.1-1.2) and about 3-3.5 times longer than the outer process (fig.15).These two ratios are 2.5 and 2.5-4 in females.The stylets IX are about 2.6 times longer than the inner process of the coxites IX in males and about 2.3 times in females (fig.14).Ovipositor very long, with 55-57 segments, reaching beyond the apex of the stylets IX up to 2-2.5 times the latter's length (fig.16).Apices of gonapophises unsclerotized.Caudal filaments as long as body length or slightly longer (maximum preserved in a paracercus: 13.5 mm; cerci a bit shorter).The main feature to distinguish this species from the other Ctenolepisma Escherich, 1905 from Europe, is the occurrence of double combs, i.e., fields of macrosetae on the thoracic sternites that are composed of two more or less parallel rows of plumose setae; in the prosternum these rows are more irregular and three rows can be observed in some fields.Some previously known species from other continents show these "double combs"; this finding is mentioned in the descriptions of South African C. weberi and C. pretoriana by Wygodzinsky (1955), and in that of C. saxeta (Irish, 1987), but no taxonomic significance was given to this feature.This character has been also detected in undescribed taxa from North Africa and the Near East (probably identified as C. lineata, without a fuller involving dissection of the specimens).C. almeriensis sp.n. is the only one of the available specimens from the West-Palaeartic region with double combs that bears only two pairs of stylets in both sexes.However, the occurrence of a double or single row in the aforementioned combs is specially useful for distinguishing between C. almeriensis n. sp. and C. lineata (respectively), two very similar taxa from the Iberian Peninsula (see figs. 17 and 18).The validity of this feature for distinguishing between both species is confirmed with a discriminant analysis (see below).More attention should therefore be paid to this feature for the future diagnosis of species of this genus.

Comparison of Ctenolepisma lineata and the new species
In comparison with the other European species belonging to the lineata-group, the new species described here is most similar to C. lineata as its tenth urotergite has the same shape, its legs show the same cover of scales and its abdominal setation is similar.For this reason, a detailed comparison of these two taxa is carried out here to elucidate the validity of the aforementioned character that is used to separate these two species.
Other differences can be found between specimens with double and single sternal combs, such as the number of macrosetae on the lateral and posterolateral combs of the nota, and on the combs of urotergites, thoracic sterna and urosternites.Even   18), with a double row (mainly insertions of macrosetae can be seen).
are significantly different on the basis of these variable features, thereby proving the validity of the "single/double rows of macrosetae" in the sternal combs.
The variables selected are detailed in Material and methods and were measured in 50 specimens of C. lineata and the same number of specimens belonging to the new species.
the ratio distance between combs / width of a comb in the metasternum and in the urosternites may differ.However, in these features the margins of variability overlap, so in some cases it is difficult to ascribe a certain specimen to C. lineata or to C. almeriensis if we do not see the combs of the thoracic sternites.A discriminant analysis may demonstrate whether C. lineata and C. almeriensis  These conclusions justify the differentiation between the two species and the validity of the double combs of macrosetae on thoracic sternites as a clear character to distinguish between them.
The specimens belonging to the new species have never been found within the area occupied by the typical C. lineata.The two species compared have never been found together in the same locality (they are probably vicarious; see fig.19).The distribution area of C. almeriensis is mainly inside the limits of the biogeographic sector called "murciano-almeriense", which is known in faunistic works for its importance as a centre of endemic species.The Penibetic mountain range  19) and only specimens with two pair of stylets were selected, as it has been recently suggested that the variety pilifera (Lucas, 1840) with three pair of stylets is a different species (Molero-Baltanás, 1995).
As a result of a standardized principal component analysis (fig.20), two components, 1 and 2, account for more than the 90% of variability.All the variables selected show a significant weight to explain this variability and therefore all of them are included in the discriminant analysis.
Taking into account the 600 data obtained from the total of 100 specimens (25 per predefined group), the following discriminant analyses were carried out: (1) comparison between males of the two predefined species; (2) comparison between females of the two predefined species; (3) and ( 4) Comparison between males and females within each one of the two species; and (5) comparison between the four groups The results are shown in table 1 and in fig.21.This multivariate analysis forms two groups of specimens that exactly fit with the two predefined species, as can be seen in the biplot of the discriminant functions 1 and 2 (fig.21), and in the classification variables (see material and methods).The analysis finds significant differences between species, but these differences are not found between sexes of a given species.
The results of the Mann-Whitney U-tests to compare the medians of the four groups are shown in table 2. They are highly significant when different predefined species were compared, and not significant when both sexes within a species were compared (except for one variable in C. lineata, which could have a significant value for sexual dimorphism).future but for the time being the findings in this work appear to be sufficient to maintain C. almeriensis as a good species, and subsequently to demonstrate the validity of the double combs of thoracic sternites as an anatomic characteristic with taxonomic importance in the genus Ctenolepisma.
Comparing C. almeriensis n. sp. with other Ctenolepisma species

Fig. 19 .
Fig. 19.Map showing the localities of Spain where the typical form of Ctenolepisma lineata and C. almeriensis n. sp. were collected.

Fig. 20 .
Fig. 20.Biplot of component principal analysis.Names of the variables are given in Material and methods.

Fig. 21 .
Fig. 21.Biplot of the discriminant analysis: SPSEX groups are established on the basis of sex and predefined species Ctenolepisma lineata and C. almeriensis.

Table 1 .
Classification table of the discriminant analysis: A. Actual SPSEX; S. Group size.Predicted SPSEX groups are the same as in fig.21.

Table 2 .
Z values calculated by two-sample Mann-Whitney rank sum tests (U-tests): Sp1.Ctenolepisma lineata predefined groups; Sp2. C. almeriensis n. sp.predefined groups; M. Males; F. Females; * Significant evidence for different median values between the two groups compared (P value < 0.05).All the specimens of C. lineata were from Spain (localities nearest to the area where C. almeriensis is found, see map in fig.