Eco-geographical variability of a chigger species Neotrombicula delijani Kudryashova, 1977 (Acari, Trombiculidae)

Entomological Review, Vol. 78, No. 2, 1998. pp. 190-196. Translated from Entdmoiogicheskoe Obvzrenie. Vol. 77. No. 1, 1998. pp. 246-255. Original Russian Text Copyright © 1998 by Stekolnikov. English Translation Copyright © 1998 by MAI1K HayKa/lnterperiodica Publishing (Russia).

Eco-Geographical Variability of a Chigger Species Neotrombicula delijani Kudryashova, 1977 (Acari, Trombiculidae) A. A. Stekolnikov Zoological Institute, Russian Academy of Sciences, St. Petersburg, Russia

Received April 16,1997. Abstract—Statistical methods were applied to study the geographical variability of a chigger mite species, Neotrombicula delijani. The scutum size and length of some setae were found to be negatively correlated with temperature and aridity of climate in the sampling sites. Despite great traditional interest toward chigger mites (family Trombiculidae Ewing, 1944), aroused by their medical importance, the systematics of this group is still insufficiently developed. In particular, the intraspecific variability remains completely obscure. Almost every report of trombiculid systematics naturally includes the main statistical characteristics of the so-called "standard" measurements of the material examined: mean, maximum, and minimum values; sometimes also variation ranges, mean errors, etc. However, data concerning the geographical variability are seldom reported; authors usually restrict themselves to merely indicating a variability over a relatively small area, without any attempts to reveal its fundamental aspects (Kardos, 1961; Kepka, 1966). Published data are often proved subsequently to reflect the difference between separate species, rather than the intraspecific variability, as was the case with, e.g., data presented by Richards (1950). In our previous communications, 3 cases of regular intraspecific variability in chigger mites were revealed. A latitudinal cline over an area from W Caucasus to Tuva was revealed for an integrated index obtained by multidimensional scaling for Neotrombicula scrupulosa Kudryashova, 1993 (Stekolnikov, 1995). No gradient of conditions was found that would coincide with the sequence of geographically adjacent samples. The same result was obtained when studying the variability of N. autumnalis (Shaw, 1790) (Stekolnikov, 1997a). Finally, N. talmiensis (Schluger, 1955) displayed a negative correlation between most of measurable characters and the aridity of territory. Specimens from C Kazakhstan and Steppe Crimea were "diminutive" as compared to those collected in

Primorskii Territory, W Caucasus, Carpathians, and other places (Stekolnikov, 1996). The same trend, though to a lesser degree, was observed also in comparing different species belonging to the talmiensis complex: N. talmiensis and N. carpathica Schluger et Vysotskaya, 1970. This communication presents the results obtained in studying the variability of N. delijani Kudryashova, 1977, a species of the autumnalis complex, described from Iran and also found by the author in Volgograd Region, Caucasus, and W Kopet Dagh (Stekolnikov, 1997b). The observed relationship between a number of measurable characters and the climate proved to be similar to that observed in the talmiensis complex. The data obtained, regardless of their essential biological meaning, provide for the first time a valid estimate of the comparative taxonomic value of quantitative characters in the genus Neotrombicula. MATERIALS AND METHODS A detailed description of the material dealt with in this communication was provided in our previous paper (Stekolnikov, 1997b). Characteristics of collection sites are given below (Fig. 1). Leninsk District (Volgograd Region) is located in the semi-desert zone characterized by arid climate and hot summer. It should be noted that, according to labels, the collecting site lies in the Akhtuba floodlands, which may have milder climatic conditions. However, species of the autumnalis complex from arid regions appear to occupy intrazonal habitats in general. This trend, shown by the Transcaucasian material, was reported by Mulyarskaya (1979). According to this

190

ECO-GEOGRAPHICAL VARIABILITY OF A CHIGGER SPECIES

191

Fig. 1. Sampling sites of Neotrombicula delijani.

author, the "steppe" species, including N. autumnalis (some other species of this group, in particular N. delijani, might have been misidentified as N. autumnalis), can be found in semi-deserts but inhabit only intrazonal biotopes, such as forest areas. Sukko settlement is located in the northern part of the Black Sea coast of the Caucasus, characterized by Mediterranean landscape types. The settlement is surrounded by vineyards and oak forests similar to Crimean broad-leaved forests. The climate of this territory is characterized by dry summer and cold rainy winter (Gvozdetskii, 1954).

sort place is characterized by "wholesome mediumalpine climate" (Topographical Map, 1996, p. 60). The exact collection locality is unknown. Kussu settlement is located in the upper course of Urukh River in North Ossetia, above a large settlement of Stur-Digora. The environs belong to the so-calle4 Central Depression: a basin between Glavnyi and Bokovoi Ridges in the middle part of the Great Caucasus. The exact collection locality and its altitude are unknown. Some data of the labels ("W of rockslide") indicate an alpine landscape.

Maikop lies in W Caucasus foothills, about 200 m above sea level. The collecting site is located S of Maikop, at about 300 m above sea level, and is occupied by fields, vegetable gardens, low beech forest, and shrub thickets along Kurdzhips River banks.

The collection site in the environs of Kurush settlement (Akhtyn District, S Dagestan) belongs to the alpine belt. The material was collected on alpine meadows and screes in N foothills of Charyndag Mountain (W of Bazar-Dyuzi, the highest peak of E Caucasus), at about 2600 m above sea level.

Kislovodsk lies in a depression between Pastbishchnyi and Skalistyi Ridges of the Great Caucasus, at 720-1060 m above sea level. This well-known re-

Musaelyan settlement (Gusakyan District, Armenia) is located on Dzhavakhet-Armenian Plateau. The collection site lies on an alpine plateau (more than 2000

ENTOMOLOGICAL REVIEW Vol. 78 No. 2 1998

192

STEKOLNIKOV

Fig. 2. The distance pattern in the space of 15 characters among Neotrombicula delijani specimens, obtained by multidimensional scaling.

Fig. 3. Maximum, minimum, and mean values for different characters of Neotrombicula delijani, urn. Sampling sites: A—Leninsk; B—Sukko; C—Maikop; D—Kislovodsk; E— Kurush; F—Musaelyan; G—Delidzhan.

m above sea level), in a swamped area covered with rather sparse low forbs (N.A. Filippova, pers. comm.).

The collection site lies at about 600-800 m above sea level.

Parkhai settlement lies in S foothills of Syunt and Khasardag Mountains (W Kopet Dagh spurs), NE of Kara-Kala District center of Krasnovodsk (= Balkan) Region in Turkmenia. The surrounding foothills are covered with low subtropical forbs with sparse shrubs.

The type locality of N, delijani lies in Kukhrud Ridge spurs, 16 km N of Delidzhan settlement (Iran, Central District), at 1600 m above sea level. The region belongs to the desert zone and must have a hot and arid climate; however, it includes quite diverse ENTOMOLOGICAL REVIEW Vol. 78 No. 2 1998

ECO-GEOGRAPHICAL VARIABILITY OF A CHIGGER SPECIES

193

biotopcs: steppe on river terraces, rocky slopes, and cultivated land (Neronov et al., 1974). This is the reason why N. delijani is found there, although the species does not inhabit deserts. According to the original description, the mites were collected at a garden edge (Kudryashova, 1977).

TISTICA, similarly to the multidimensional scaling. In our study of the variability in Neotrombicula autumnalis (Stekolnikov, 1997a), the hierarchical cluster analysis was used, to provide a quite satisfactory and easily interpretable result. In the present case, however, the dendrogram obtained by this method proved to furnish less information than the diagram of factor All available N. delijani specimens, whose condivalues did. The factor analysis is based on the astion was good enough for measuring, were used in this sumption that correlations between characters result study. In all, 43 larvae from different locations were from the influence of a small nuniber of immeasurable measured: Leninsk, 5; Sukko, 5; Maikop, 8; Kislovariables, or factors; the essence of the method convodsk, 8; Kussu, 1; Kurush, 10; Musaelyan, 3; Karasists in constructing such factors in a certain way. Kala, 1; and Delidzhan, 2. Only 2 factors are usually considered, which allows Each specimen was initially described using the correlation structure between characters to be 27 standard quantitative characters. Of these charac- visualized as a 2D scatterplot, with characters repreters, 3 were countable (DS, the number of dorsal setae; sented by points whose coordinates correspond to the VS, the number of ventral setae; and NDV, a sum of so-called factor loadings, i.e., measurements of relathese numbers), 1 character was represented by a ratio tions of a certain character with each factor. Extraction (mt, relative distance from mastitarsala to base of tar- of factors was performed using the method of principal sus HI); others were measurements of separate struc- components. The result obtained was transformed by tures, sums of measured values (Ip, sum of leg rotating the factor space coordinate system, so as to lengths), mean values for serial structures (Dm and produce the simplest pattern of relations between the Vm, mean length of dorsal and ventral setae, respec- factors and measurable characters. This procedure was tively), or intervals (D and V). Some characters were accomplished using the van max method (Computer later excluded from analysis, for the same reason as in Biometry, 1990). our previous communications (Stekolnikov, 1997a). All data processing was performed using an IBM The pattern of general similarities between all PC AT-486 computer. specimens was visually represented using the method of multidimensional scaling, which produces a plane RESULTS projection of specimen distribution in the multidimenThe general similarity diagram between samples of sional character space, with the distance pattern predifferent species belonging to the autumnalis complex served. The common Euclidean distance was selected to estimate the similarities in the character space. The (Stekolnikov, 1997b: Fig. 1) demonstrated a very high scaling method utilized by the "STATISTICA for dispersion of N. delijani samples. Analysis of the reWindows 4.3" program was used. The scaling was sults of multidimensional scaling for separate speciperformed with respect to the following 15 characters. mens (Fig. 2) shows that areas occupied by specimens Distances: AW, between antero-lateral setae of scu- from different localities mostly do not overlap. The tum; PW, between postero-lateral setae of scutum; SB, geographically close samples from Sukko and Maikop between bases of sensilla; ASB, from bases of sensilla do overlap, while a specimen from North Ossetia is to anterior margin of scutum; PSB, from bases of sen- clustered marginally with another sample from N Causilla to posterior margin of scutum; P-PL, from pos- casus: namely, from Kislovodsk. tero-lateral setae to posterior margin of scutum; AP, The values of certain characters are very diversely from antero- to postero-lateral setae; Lengths: SD, distributed among the samples (Fig. 3).1 Evident gaps scutum; AM, antero-medial seta of scutum; AL, an- occur between the value ranges for different samples, tero-lateral setae of scutum; PL, postero-lateral setae with samples grouped in various ways on different of scutum; H, humeral setae; Talll, tarsus III; Dm, diagrams; this peculiarity results in that the specimens mean length of dorsal setae; and NDV, the number of from different samples cluster separately in Fig. 2. For idiosomal setae. The relations between characters were estimated by the method of factor analysis, performed using STAENTOMOLOGICAL REVIEW Vol. 78 No. 2 1998

Samples from North Ossetia and W Kopet Dagh, consisting of 1 specimen each, were not included in this analysis.

194

STEKOLNIKOV

Fig. 4. Maximum, minimum, and mean values for characters describing scutum size in Neotrombicula delijani, Jim. Designations as in Fig. 3.

example, samples from Iran and Armenia are similar in mean length of dorsal setae (Dm), contrasting with all other samples. With respect to the length of tarsus III, specimens from Iran differ greatly from those from Volgograd Region, Kislovodsk, Dagestan, and Armenia, while Sukko and Maikop samples occupy an intermediate position; the intervals for Volgograd Region and Kislovodsk do not overlap as well. With respect to the number of idiosomal setae (NDV), the sample from Iran also contrasts with those from Volgograd Region, Dagestan, and Armenia; however, the sample from Kislovodsk has an entirely different position. The mean length of antero-medial seta of scutum in mites from Sukko and Maikop is noticeably smaller; the interval for Maikop overlaps, except that for Sukko, with that for Iran only. Measurable characters describing the width (AW, PW) and length (PSB, SD) of scutum show a peculiar distribution pattern (Fig. 4). In the sequence of collecting sites located along the north general slope of the Great Caucasus (Sukko-Maikop-KislovodskKurush), the character values increase eastward. The sample from Volgograd Region (Leninsk) adjoins this sequence on the left side, while those from Iran (Delidzhan) and Armenia (Musaelyan) are adjacent to it on the right side. The same trend is to some extent evident in characters describing the setae (PL, Dm;

Fig. 3). This trend can be correlated with the natural conditions at collecting sites. As it can be seen from the collecting site characteristics provided above, the sequence Leninsk-SukkoMaikop-Kislovodsk-Kurush-Musaelyan corresponds to a transition from semi-deserts to Mediterranean landscapes, then to the foothill zone of W Caucasus, and, finally, to middle and high mountains. According to published geographical data, the climate in this sequence becomes colder and evidently less arid. The general eastward trend of increasing aridity and continental climatic features typical of the Great Caucasus, is completely overruled by the alpine zonality. On the northern slope and southern Ciscaucasian plains, where Maikop and Kislovodsk are located, "... this trend is weakened and obscured by the climatic changes which occur, owing to the vertical zonality, as the landscape changes from plains to mountain ridges" (Kavkaz, 1966, p. 94). The alpine regions of the Great Caucasus (Kurush) and Transcaucasian Plateau (Musaelyan), starting from an altitude of 2000 m above sea level, are influenced by free atmospheric currents (Gvozdetskii, 1954) and therefore are also not included in the mentioned trend. Climate maps also demonstrate a decrease in aridity and some temperature characteristics in this sequence. For example, the precipitation during the warm period is 175-200 mm ENTOMOLOGICAL REVIEW Vol. 78 No. 2 1998

ECO-GEOGRAPHICAL VARIABILITY OF A CHIGGER SPECIES

195

near Leninsk; 200-350 in Sukko; 400-500 in Maikop; about 500 in Kislovodsk; 400-600 in environs of Kurush; and 600-700 near Musaelyan. The mean July air temperature is 24°C in Leninsk, 23°C in Sukko, 22°C in Maikop, 16-20°C in Kislovodsk, and 12-16°C in Kurush and Musaelyan (Climatic Atlas of the USSR, 1962). Owing to the small scale of the maps, the climate of each locality cannot be characterized precisely, but it is unnecessary in this case. Thus, the scutum size and length of some setae in Neotrombicula delijani display an obvious negative correlation with the temperature and aridity of climate at the collecting sites. Because of the insufficient material examined, the exact relative position of a certain interval among others can be doubted, whereas the obvious general trend, well shown in Fig. 4, is indubitable. This fact agrees well with the diminution trend revealed in mites of the talmiensis complex from more arid habitats (Stekolnikov, 1996). Such an ecology-related diminution may be considered a general feature of the species N. delijani, as compared to N, autumnalis and N. caucasica, if record data for these species from areas with different climate are summarized. The position of the type specimens is entirely inconsistent with the general pattern. The sample from Iran should have been adjacent to that from Volgograd Region, rather than lie between the samples from alpine areas of Dagestan and Armenia. The pattern of character relations, obtained by factor analysis (Fig. 5) shows the changes of most characters to be highly correlated, which results in that "diminution" of mites from some localities can be recognized. Characters describing the width and length of scutum and length of most setae make up a dense group along the axis of Factor 1, which can be unambiguously interpreted as the mite size, at least partly correlated with climatic parameters. Factor 2 allows no meaningful interpretation. DISCUSSION Correlations between size characteristics and climatic parameters have been reported many times for various animal groups. In particular, in poikilothermal animals, the size of adult individuals usually decreases with increasing temperature (Mina and Klevezal, 1976). The same situation occurs in our case, though with respect to larvae rather than adults. This trend is explained by several factors: growth acceleration and ENTOMOLOGICAL REVIEW Vol. 78 No. 2

1998

Fig. 5. Neotrombicula delijani: results of the factor analysis.

earlier beginning of some maturation stages at higher temperatures; selective advantage of larger individuals under more severe climatic conditions; etc. (Mayr, 1963; Mina and Klevezal, 1976). Different factors often act in opposite directions, and their interaction may lead to different results. To choose from the possible explanations, experimental data concerning the influence of different environmental factors on the size of the animal studied are therefore necessary. The only data obtained so far for chigger mites concern the influence of temperature on the development rates and the effect of humidity on the survival rate (Hsu and Chen, 1960; Kulkarni, 1988; etc.; references in Shatrov, 1992). The effect of abiotic factors on the size was not studied. It should be noted that the size of the larvae (on whose characters the entire chigger mite systematics is based) can be determined both by direct action of various factors on the morphogenesis and by egg size; the latter, in its turn, must in some way depend on the size and other features of the female, which are formed under the influence of climatic conditions. The selective size-dependent elimination of larvae also cannot be excluded and may lead to the observed differences of mean character values in different populations. Finally, it is possible that characters in different populations arc, to varied extent, genetically determined; in the case of strict determination, such factors as the history of a population and a landscape must also play some role. As a result, we have to restrict ourselves to establishing the fact of the correlation, the reasons for which remain unknown. In particular, the relative role of temperature and humidity in this trend is obscure.

196

STEKOLNIKOV

The observed states of measurable characters in the sample from Iran can be explained in several ways. The situation may result, for example, from the fact that under strongly arid conditions of Central Iran, the prolongation of development due to insufficient humidity "overrides" its shortening at a higher temperature. The longer development period leads to a greater size reached upon maturation. In this case, larger eggs are laid and the size of the larvae increases correspondingly. It seems premature to consider this problem in detail, since the material of N. delijani from Iran includes only 2 specimens. The distinctly isolated position of characters describing the leg length (Ip, Talll) on the factor analysis diagram (Fig. 5) is interesting. In the talmiensis group, these characters were included in the complex of measurements subject to the ecology-related diminution in samples from Central Kazakhstan and Crimea, as compared to the majority of N. talmiensis (Stekolnikov, 1996). The fact that the leg length characters are excluded from this trend in N. delijani is probably related to some peculiarities of its adaptive strategy. It should be noted that legs of steppe and desert Neotrombicula are generally either much longer or, contrariwise, shorter than legs of the forest zone inhabitants. The same divergence can be easily observed by comparing the seta lengths and scutum sizes in different species. N. tragardhiana (Feider, 1953) is an example of "small" steppe species; and N. heptneri Kudryashova, 1973 represents a "large" species. Thus, a broader investigation of the pattern of character correlations combined with ecological and parasitological data will be necessary to solve the outlined problem. ACKNOWLEDGMENTS This work was supported by the Russian Foundation for Basic Research, project no. 97-04-50094: 'Taxonomic Structure of a Species of Blood-sucking Mites". REFERENCES 1. Gvozdetskii, N.A., Fizicheskaya geografiya Kavkaza, Kurs lektsii (Physical Geography of the Caucasus: Course of Lectures), Moscow: Moscow Univ., 1954, no. 1. 2. Hsu, P.-K. and Chen, H.TT., Observations on the Relationship between the Survival of Unfed Trombicula Larvae (T. akamushi var. deliensis) and the Temperature and Relative Humidity, Ada Microbiol. Sinica, 1960, vol. 8, no. l,pp. 8-16. 3. Kardos, E.H., Taxonomic Studies on the Larval Trombicula (Neotrombicula) nagayoi Complex of Central Korea (Acarina: Trombiculidae), Ann. Ent. Soc. America, 1961, vol. 54, pp. 499-508.

4. Kavkaz (The Caucasus), Gerasimov, I.P., Ed., Moscow: Nauka, 1966. 5. Kepka, O., Trombiculidae (Acari) aus der Turkei: II, Zeitschr. Parasitenk., 1966, vol. 27, nos. 1/4, pp. 43-63. 6. Klimaticheskii atlas SSSR (Climatic Atlas of the USSR), Davitaya, F.F., Chief Ed., Moscow, 1960, vol. 1. 7. Komp'yuternaya biometrika (Computer Biometry), Nosov, V.N., Ed., Moscow: Moscow State Univ., 1990. 8. Kudryashova, N.I., New Species of the Genus Neotrombicula Hirst, 1915 (Acariformes, Trombiculidae) from Iran, Byul Mosk. O-va Ispyt. Prirody., Otd. Biol, 1917, vol. 82, no. 3, pp. 46-59. 9. Kulkarni, S.M., Some Observations on the Life History and Behavior of Three Species of Trombiculids in India, Progress in Acarology, Channa Basavanna, G.P. and Viraktamath, C.A., Eds., New Delhi: Oxford & IBH Publ.Co., 1988, vol. l,pp. 133-138. 10. Mayr, E., Animal Species and Evolution, Cambridge, Mass.: Beknap Press, Harvard University Press, 1963. U. Mina, M.V. and Klevezal, G.A., Rost zhivotnykh (Growth of Animals), Moscow: Nauka, 1976. 12. Mulyarskaya, L.V., Some Peculiarities of the Station Change Rule in Mountain Landscapes Exemplified by Trombiculidae Mites, Zool. Zh., 1979, vol. 58, no. 11, pp. 1653-1658. 13. Neronov, V.M., Farang-Azad, A., and Teslenko, E.B., Distribution and Abundance of Gerbils of the Genera Meriones, Tatera, and Gerbillus in Iran and Problems of Zoogeographical Zonation of this Area, Byul. Mosk. Ova Ispyt. Prirody, Otd. Biol., 1974, vol. 79, no. 5, pp. 41-71. 14. Richards, W.S., The Variation of the British Harvest Mite (Trombiculidae, Acarina), Parasitology, 1950, vol. 40, nos. 1-2, pp. 105-117. 15. Shatrov, A.B., Captive Rearing and Life Cycle of the Chigger Mite Leptotrombidium orientale (Schluger, 1948) (Acariformes, Trombiculidae), Entom. Obozr,% 1992, vol. 71, no. 2, pp. 478^97. 16. Stekolnikov, A.A., Fauna and Systematics of Chigger Mites of minuta Complex, the Genus Neotrombicula (Trombiculidae), Parazitologiya, 1995, vol. 29, no. 4, pp. 250-266. 17. Stekolnikov, A.A., Chigger Mites of talmiensis Complex (Trombiculidae: Neotrombicula) of Russia and Adjacent Regions: Computer-Aided Taxonomic Analysis, Parazitologiya, 1996, vol. 30, no. 5, pp. 377-397. 18. Stekolnikov, A.A., Geographical Variability of the Chigger Neotrombicula autumnalis and Relation of this Species to N. caucasica stat. nov. (Trombiculidae), Parazitologiya, 1997a, vol. 31, no. 5, pp. 397-413. 19. Stekolnikov, A.A., New Data on the Fauna and Systematics of Chigger Mites of autumnalis Complex (Trombiculidae, Neotrombicula), Parazitologiya, 1997b, vol. 31, no. 6, pp. 527-542. 20. Topograficheskaya karta Stavropol'skogo kraya (Topographical Map of Stavropol Territory), Eroshkin, I., Ed., Moscow, 1996. ENTOMOLOGICAL REVIEW Vol. 78 No. 2 1998