List of leaf beetles (Coleoptera: Chrysomelidae) from Lagodekhi reserve with new records for Transcaucasia and Georgia Author Aslan, Ebru Gül Author Mumladze, Levan Author Japoshvili, George text Zootaxa 2017 4277 1 journal volume 32835 10.11646/zootaxa.4277.1.6 a8f97942-a2c4-4964-a8e3-0188bc119220 1175-5326 809095 BFC106DD-B9B5-46A1-9216-96CC255595AB Genus Psylliodes Berthold, 1827 47. Psylliodes cuprea (Koch) Specimens examined. 1 male, H3, 06.10- 16.10.2014 ; 1 female, H1, 15- 25.05.2014 ; 1 female, H3, 04- 14.06.2014 . Distribution. Known from Georgia ( Gruev, 2003 ). Palaearctic. 48. Ps. isatidis Heikertinger ** Specimens examined. 1 male, H2, 15- 25.05.2014 ; 3 males, 1 female, H3, 15- 25.05.2014 . Distribution. Sibero-European. Species inventory. The number of leaf-beetle species known from Lagodekhi protected areas arrived at 48 species, with 14 new records due to our study. In spite of the relatively low number of captured species ( 32 in total), they constitute ca. 11% of all Chrysomelidae known from Georgia ( Zaitsev, 1929 - 1953 ; Shengelia 1951; Kobakhidze, 1956 ; Seperteladze, 1960 - 1983 ; Orlova-Bienkowskaja, 2014 ). Most of the species (53%) collected during this study were represented by singletons (11species) and in pairs (6 species), indicating the spatial rarity of leaf beetles in LNP. The correlation between the number of individuals and the species richness was very strong ( r =0.9, p =0.032) meaning that the actual species number would increase significantly in case of increasing sampling effort. In contrast, asymptotic species richness estimation (calculated for each sampling plot) predicted in average additional 4 (±1sd) species at each elevation and 9 additional species when all the data were pooled together which could be interpreted as a fairly complete inventory ( Table 1 ). The estimated low species richness could be ascribed to a relatively low number of species for each elevation while species turnover rate is high (34% in average) considering the short distance between plots (in average 1.2 km ). Although not similar studies are available for neighbour areas, the sampling of Chrysomelidae is usually based on sweep nets rather than Malaise traps ( Bouzan et al. , 2015 ; Sánchez-Reyes et al. , 2014 , 2016 ). Indeed, the actual species richness could significantly increase in case of sweep netting and beating as the sampling with Malaise traps may not be able to effectively collect beetle species associated with herbaceous and arboreal vegetation ( Aslan et al. , 2012 ). TABLE 1. A summary table describing the sampling localities, habitat description, total abundances for each locality, total and estimated species richness of chrysomelids and plants. The last column indicates the length of the vegetation gradient expressed as counts of 10-day sampling events.

Sampling localities	Elevation	Longitude	Latitude	Habitat
H1	665	41.85248	46.28776	Low-land forest
H2	845	41.85585	46.29273	Mountain forest
H3	1345	41.87146	46.31153	Mountain forest
H4	1850	41.88273	46.32185	Mountain forest
H5	1900	41.88557	46.32413	Subalpine forest
H6	2230	41.89805	46.33387	Subalpine meadow
H7	2559	41.90616	46.33340	Alpine meadow
continued.
Sampling localities	Total abundance	Species richness	Expected richness	Unique Plant species richness
H1	100	13	15	5 51
H2	74	10	16	1 50
H3	111	11	23	3 39
H4	24	9	15	1 34
H5	49	8	12	3 90
H6	5	3	4	2 139
H7	11	4	7	2 119

Patterns of diversity distribution. Leaf beetles, while being a very diverse family, is rather poorly studied from an elevational diversity gradient perspective. Only a few studies are available, mainly from Mexico ( Furth, 2009 ; Sánchez-Reyes et al. , 2016 , 2014 ) and Brazil ( Bouzan et al. , 2015 ). All these studies are reporting a hump shaped diversity patterns along with elevational gradient except Sánchez-Reyes et al. (2016) . In this case, the highest diversity was found at the highest elevation which was at 1080m a.s.l.. This on the other hand indicates that the actual pattern of species diversity may be hump-shaped rather than positive linear if the data from elevations above 1080 m would have been gathered. In contrast to species richness distribution in the above mentioned studies, there is no consistent pattern in the distribution of total abundances with elevations. In our case, both measures of species richness (observed and estimated) as well as individual density are decreasing with increasing elevations in a linear manner ( Table 2 ; Fig. 1 ). GLM with Poisson errors proved to fit data well with elevation explaining a significant amount of variation in diversity distribution of leaf beetles ( Table 2 ). In contrast to elevation the beetle diversity and abundance were inversely related to plant species richness (Spairman’s r =-0.6, p>0.05 and r=- 0.7, p>0.05) ( Fig. 2 ). Such kind of inconsistency between the diversities of plants and terrestrial snails were also observed in LNP where plant species richness (also increased with elevation) was not related to snail species richness at all (showing unimodal pattern) ( Mumladze et al. , 2017 ). However, the distribution of snails as well as plants is rather well explained by a climatic variability hypothesis while there is no obvious sampling bias introduced. Unexpected (although insignificant) relationship between plant and leaf beetle species richness in our case is supposed to be a sampling artifact rather than an anomalous phenomena. Indeed, the increase of plant species richness with increasing elevations is completely due to herbaceous vegetations, beetle diversity of which may not be correctly accounted for Malaise traps. On the other hand, the observed elevational diversity pattern of leaf beetles could also be a biased by the same reason. E.g. the number of plant species does not uniformly increases along with elevation. In the forest area (H1-H4) plant species richness is generally decreasing which is in concordance with beetle diversity ( Table 1 ; Fig. 2 ). In the higher elevations (H5-H7) number of plant species is increasing quickly (from 34 at H4 to 90 at H5) which is due to herbaceous vegetation, and the beetle species richness does not reflect this change. The last two highest elevational plots which represent completely treeless subalpine meadows exposed highest plant species richness and lowest beetle species diversity. These clearly show that the Malaise trap is able to capture leaf beetle communities more or less effectively only in forest habitats. Nevertheless, the beetle species richness in forests is decreasing with elevation, and the estimate of beetle species richness is presumably affected by a strong sampling bias in subalpine to alpine areas. If this assumption will hold then meadows above tree line harbouring a large plant species diversity would also result in a much diverse beetle community. To test these hypotheses, additional data collection is necessary in subalpine to alpine areas of LNP.