Ecology of latvian reptiles: environmental factors and abundances of terrestrial reptiles

Дата канвертавання18.04.2016
Памер252.69 Kb.
  1   2   3

University of Latvia

Andris Čeirāns




Summary of the Promotion Paper

for promotion to the degree of Doctor of Biology

Sub-branch of Biology - Zoology

Riga 2005

Importance of the study

Numerous factors can affect the distribution and abundance of reptiles, such as climate, topography, habitats, and various anthropogenic impacts (Gaston 2003). The influence of some of these factors is relatively well studied, while the connection with others has received less attention from scientists. Thus, the importance of regional climate is largely overlooked in the lowland temperate-zone Europe, where its effect is far less obvious than in the mountain ranges or at the species latitude distribution limits (Gaston 2003). Habitats of European reptiles have been described in detail at least for some species (e.g. Lacerta agilis), although this information (especially from warmer climate zones) is not always applicable to Latvia. The number of studies in temperate Europe on reptile ecology is still insufficient due to their low overall density and very uneven distribution. Most of the studies are limited to only a few sites, and (or) only to 1-2 species in each case. The understanding which climate and habitat characteristics are important for reptile abundance is especially valuable in reptile conservation efforts such as the establishing and management of protected areas.

Investigations on the connection with forest habitats are also essential. Unlike the Nordic Countries, where the intensification of forestry (e.g. using mechanization in forest planting and harvesting, the use of fertilizers, monoculture planting) resulted in considerable loss of biodiversity, the forests of East Europe have remained less affected (Stanners & Bourdeau 1994). Therefore, reptile studies in East European forests are important from several aspects. The first is to gain an understanding of how widely reptiles exploit forest habitats - the habitat that once dominated throughout most of the range for many native species. The other stems from the species conservation aspect -after the collapse of the Soviet Union and establishment of a market economy in the former socialistic countries, threats to biodiversity are increasing in this area, and reptiles could also be affected.

Although first information about reptiles in the territory of modern-day Latvia was published by German authors already in the 18th and 19th centuries (e.g. Fischer 1791, Drumpelman & Friebe 1806), data on species ecology in Latvia is still scanty. Probably the most comprehensive information about reptiles in Latvia was published before World War II in a popular booklet (Silins, Lamsters 1934). The later data are either publications in popular journals or books (e.g. Spuris et al. 1974) or the annotated species lists for particular areas (e.g. Barsevskis et al. 2002). Data regarding a rare reptile species in Latvia -the Smooth Snake (Coronella austriaca), are very fragmentary, and deal mostly with its distribution (Silins & Lamsters 1934; Spuris et al. 1974; Lipsbergs et al. 1990), and parts of these data are doubtful. Therefore, the critical revision of distribution data, species habitat descriptions, and the evaluation of the population status in Latvia are necessary.

Aims and tasks of the study

The aim of the present study was to find out environmental factors that determine spatial distribution and abundances of native lizard and snake species, from large-scale (such as climate and broad habitat groups) to small-scale factors (such as vegetation composition and structure). The specific tasks of the work were the following:

To collect the reptile abundance data representative for all the territory

of Latvia;

To find out relationships between reptile abundances and specific

climate factors and habitats; • To describe the composition and structure of vegetation on reptile sites

and to clarify important vegetation factors;

To study the relative abundances of reptiles in forest habitats, and to

clarify forest type and forest stand preferences;

To verify distribution data of a rare species - the Smooth Snake

(Coronella austriacd), to provide description of its habitats, and to

evaluate the status of the currently known populations.

Scientific novelty and practical usefulness of the study

  1. The clarifying of the important climate factors for native reptiles, and the present survey may be representative for the whole region of the sub-boreal Eastern Europe.

  2. The clarifying of the essential habitats for reptiles in Latvia at different spatial scales, from broad habitat type to microhabitat.

  3. The determination of the general pattern of spatial distribution of several species (Anguis fragilis, Zootoca vivipara, Natrix natrix) among major groups of sub-boreal wooded habitats, among tree stands with various composition and age, and the evaluation of the species niche overlap among forest types.

  4. The evaluation of the effect of several anthropogenic factors (proportions of agricultural lands and urban areas, human population density) on native reptiles.

  5. The evaluation of the population status of a rare reptile species in Latvia - the Smooth Snake (Coronella austriacd).

  6. The acquisition of the information important in conservation efforts, such as establishing and management of protected areas.

Approbation of the study

The results of the dissertation were partly presented as oral presentations in two international meetings:

  1. Biodiversity and Conservation of Boreal Nature, Kuhmo, Finland, 2000.

  1. 11th Ordinary General Meeting of Societas Europaea Herpetologica, Zalec, Slovenia, 2001.


The dissertation is based on the following publications and manuscripts:

  1. Čeirāns A. 2000. The smooth snake (Coronella austriaca Laur.) in Latvia: distribution, habitats, and conservation. - Proceedings of the Latvian Academy of Sciences, section B 54 (3): 85-90.

  1. Čeirāns A. 2002. On the importance of tree stand composition and age in forest habitats of Anguis fragilis, Zootoca vivipara, and Natrix matrix. - Herpetozoa 15 (1/2): 63-74.

  1. Čeirāns A. 2002. Reptiles and amphibians of the Gauja National Park, Latvia. - Biota 3 (1-2): 17-25.

  1. Čeirāns A. 2003. Reptiles and anurans of the Kemeri National Park, Latvia. - Biodiversity and conservation of boreal nature: proceedings of the 10 year anniversary symposium of the nature Reserve Friendship (Heikkila R. & Lindholm T. (eds.)): the Finnish Environment 485, Kainuu Regional Environment Centre: 182-186.

  1. Čeirāns A. 2004. Reptiles in sub-boreal forests of Eastern Europe: patterns of forest type preferences and habitat use in Anguis fragilis, Zootoca vivipara, and Natrix natrix. - Herpetozoa 17 (1/2): 65-74.

  1. Čeirāns A. In press. Reptile abundance in temperate-zone Europe: effect of regional climate and habitat factors in Latvia. - Russian Journal of Herpetology.

  1. Čeirāns A. Vegetation structure and composition in terrestrial reptile habitats in Latvia. - Submitted manuscript.

Structure of the dissertation

The dissertation is written in English, it consists of an introduction (Chapter 1), followed by four chapters each written as separate surveys with their own introduction, methods, results and discussion parts, and the main conclusions. Chapters 2 and 4 are manuscripts in press and submitted respectively (6th and 7th in the publication list). Chapter 3 is based on two publications (2nd and 5th in the publication list), and Chapter 5 - based on a separate paper (1st in the publication list).


1. Material and methods

All the data were collected in three separate surveys. Data on microhabitats (Chapter 4), as well as on the influence of large-scale factors - climate and broad habitat types (Chapter 2) were collected simultaneously in a survey that covered the whole territory of Latvia (Survey 1, bellow). Data on reptile forest habitats -their stand characteristics and forest type preferences (Chapter 3) were collected in another survey, conducted in two National Parks located in the middle part of Latvia (Survey 2, below). And, finally, the Smooth Snake (Coronella austriaca) was investigated in the third survey (Chapter 5). All statistics were realized on SPSS for Windows Version 11.5 (2002)® and Statgraphics Plus® software.

Survey 1: Microhabitats, habitats, and climate

Sampled sites were selected at random from stratified plots. Twenty 25x25 km plots from the Baltic Coordinate System were chosen, and three to five 5x5 km squares (total of 92) were randomly selected from each 25x25 km plot. Geo-botanical zoning was employed because it encompasses many factors, such as soil, geology, geomorphology, and climate, not just vegetation characteristics. Such an approach should provide representative sampling for the whole territory of Latvia. The number of selected squares in each region was roughly correlated with its area (Figure 1). Reptiles were counted on transects that crossed 5x5 km squares through their central parts using the road, path and forest cutting networks.

Censuses were carried out by the same observer, once on each transect, in the field seasons (May-September) of 1999-2003. As the activity and observed frequency of reptiles vary during a season or between years (e.g. Jablokov 1976, Kosov 1983, Glandt 1995), transects in the same 25x25 km plot were surveyed in different months and years to reduce the impact of this factor on abundance estimates. Censuses were carried out over 5-9 hours under predominantly dry and warm (19-25°C) weather conditions. The total length of transects was 689.30 km (average ± standard deviation - 7.46 ± 2.42 km per 5x5 km square).

Vegetation was described on a circular plot with the centre at a point where each reptile specimen was first spotted. The total number of plots was: 27 for the Sand Lizard (Lacerta agilis), 136 -for the Common Lizard (Zootoca vivipara), 57 - for the Slow Worm (Anguis fragilis), 28 - for the Grass Snake (Natrix natrix), and 32 - for the Adder (Vipera berus). The radius was 1.5 m for moss layer and herbs, 5.0 m for shrubs, and 10.0 m for trees. A modified Braun-Blanquet method was used. Vegetation data was described as coverage, which was evaluated visually. To reduce possible estimation error, the coverage was estimated as belonging

to the coverage classes coded as a whole number from 0 to 5 (see table 6 in chapter 4). It was described separately for different height classes, taxons and ecological groups. The latter were selected arbitrarily, on the basis of literature sources (Pētersone & Birkmane 1980, Fitter at al. 1984, Fitter et al. 1996) and author's personal experience. Easily identifiable and frequent taxons (trees, undershrubs, some herbs) were treated at the species level. Microhabitat use among reptile species was examined using Discriminant Function Analysis (DFA).
Figure 1. Location of sampled 25x25 km plots and sampled 5x5 km squares in geobotanical

regions of Latvia (see Survey 1 in methods)

Altitudes and total coverage of landscapes (forest, open landscapes, mires, urban areas) within a circle of 2.5 km radius around each reptile observation were acquired from topographic maps (1996-1997) with a scale of 1 : 50 000. In further analysis, mean values of parameters above for all observations in the given 5x5 km square were used. The number of squares with the species data was: 12 for Lacerta agilis, 89 for Zootoca vivipara, 32 for Anguis fragilis, 15 for Natrix natrix, and 18 - for Vipera berus. For a random sample, the same parameters were determined around centers of all surveyed 5x5 km squares (n = 92). Differences between the random sample and reptile sites were evaluated by Mann-Whitney W tests.

Relative abundance data for Lacerta agilis, Anguis fragilis, Natrix natrix, Vipera berus were expressed as the proportion of occupied 5x5 km squares in each 25x25 km plot. In most common species -Z. vivipara, average density data in 25x25 km plots were used (average ± standard deviataion

-0.51 ± 0.30 records per km). Records of juveniles were omitted to reduce seasonal differences. Variation in density between years was not statistically significant for this species (ANOVA, p>0.1).

Stepwise Multiple Regression Analysis was performed to assess relation­ships between relative abundance of reptiles, and the following predictors: climate factors, average human population density (from Turlajs 1998), and the proportion of different habitats along transects. Only models Statistically significiant in both forward and backward selection were chosen. Climate variables were acquired from published maps (Temnikova 1958; Kavacs 1998), and reduced into climate factors through the Principal Component Analysis (PCA) with varimax rotation. Data on habitat percentages along transects were acquired from field descriptions. Habitats were classified as eight principal types: dry coniferous forest, wet coniferous forest, dry deciduous forest, wet deciduous forest, forest edges, mires, pastures and fallow lands, agrolandscapes. Habitat data were square-root (x+0.5) transformed.

Because reptile abundance could also be affected by the location of transect (in the intact habitat or on the road verges with various characteristics), such data were also included in my original analyses. However, as none of these factors was significant, I left them out of the analyses printed here.
Survey 2: Forest habitats

Data were collected on transects in the field seasons (later April to early September) of 1994-1997 for the Ķemeri National Park (KNP), and 1998-2000 for the Gauja National Park (GNP). Transects (198.1 km in KNP, 103.7 km - in GNP) were evenly distributed and covered the whole territory in both study areas. They were located in forest habitats, mostly in sites with potentially highest reptile density (forest ride verges, cuttings etc). Observations on forest fringes and sides of large roads were excluded from analysis as not representing a forest environment. Censuses were carried out once on each transect, each observation of a reptile was mapped at a scale of 1: 50 000, and a brief description of the site was made. Descriptions were later compared with forest management plans and stand descriptions from the database of the State Forest Service, to ensure coherence between both the find and the database. Observations of juveniles were omitted.

Forest types along transect, height, age, and composition of reptile and random stands were determined from the forest database. Differences in age and composition between the tree stands inhabited by reptiles and the random stands (200 for each of the study areas) were assessed by Mann-Whitney W test. Linear regression was used to test the relationship between age and proportion of various tree species in the various stands. Polynomial regression analysis

was used to test the significance of the relationship between reptile observation dates and stand composition. The forest type of the stands was not specified in these cases. The following values and indexes were calculated for forest type preference analysis: deviation from the expected proportion of records in each forest type (the difference between the proportion of records and the proportion of transect divided by the proportion of transect in the given type), the percentage overlap and the Hurlbert's niche overlap indexes among species (Krebs 1989).

Survey 3: A study on the Smooth Snake (Coronella austriaca)

Data on the distribution of C. austriaca in Latvia were obtained from literature (Ecke 1927; Luta 1973, Lipsbergs et al. 1990 etc.), and from personal communications with zoologists and naturalists. The plausibility of records after the 1980s was assessed by discussions with persons who made these observations. Field observations and descriptions of habitats were made by the author in 1994-1997 in six locations in the Ķemeri National Park, and in one location in the Slitere National Park. Every location was visited 2-4 times, excepting one location, which was visited 18 times. Every location was thoroughly searched for 1-3 hours (mostly in morning), depending on the size of site. Special attention was paid to potential hiding places. Field descriptions of sites were made (exposure, degree of vegetation coverage, dominant plant species). Some additional data on feeding, litter size and size of newborns were collected in captivity in the Riga Zoo, from specimens caught in the wild in the same year.

2. Effect of regional climate and habitat type on reptiles in Latvia

Five of the seven native species of reptiles were observed on transects in this survey. Lacerta agilis was found on 13 % of visited 5x5 km squares, Natrix matrix on 16 %, Vipera berus on 20 %, Anguis fragilis on 35 %, and Zootoca vivipara on 96 %. In Natrix natrix, observed altitudes differed significantly from the random sample (p<0.0001, Figure 2). This species was found only at elevations below 50 m a.s.l.

The landscapes inhabited by Anguis fragilis had higher forest coverage (range (%), mean ± SE, median: 36-95, 69±3, 70) and lower open landscape coverage (4-64, 26 ± 3, 26), than random sites (respectively: 1-97, 54 ± 2, 57; and 0-96,41 ± 2, 39). These differences were significant at p<0.01. Vipera berus inhabited areas with higher coverage of mires (0-40, 8 ± 3, 2) than sites in the random sample (0-69, 4 ± 1, 0.05). The difference was significant at p=0.03. There were no differences in urban coverage between landscapes inhabited by reptiles and a random sample (p>0.1).

Figure 2 Altitudes of reptile observations and random sample (mean values marked with cross)

Box-and-Whisker Plot

PCA grouped climate variables in three components that accounted for 83 % of the total variance. PCA 1 accounted for 47 % of the explained variance and had positive loadings from variables characterizing mild and short winters. PCA 2 (25 % of the variance) had positive loadings from variables characterizing high rates of precipitation and PCA 3 (12 %) - from variables characterizing long and hot summers (Table 1).

Table 1

Weight of climate parameters in PCA after varimax rotation


Component 1

Component 2

Component 3

Precipitation in warm (IV-X) season




Precipitation in cold (XI-III) season




Annual precipitation




Days with snow cover




Percentage of winters with unsteady snow cover




Air temperature in January




Air temperature in July




Frost-free period on ground




Frost-free period on grass




Frost-free period in air




Period with mean temp >10 °C




Period with mean temp > 5 °C




Sum of active temperatures




Annual number of cloudy days




Bold - parameters with weight > 0,55, used in data reduction

In the multiple regression analysis, reptile abundance was predicted only by climate and habitat factors. The transect characteristics and the rural population density were not important. The abundance of Lacerta agilis was predicted by a combination of two factors: the hot summer climate (PCA 3), and the proportion

of dry coniferous forest (R2adj = 48%, DW = 2.20, TPCA3 = 2.50, TDryCon = 3.25, p=0.0014). Zootoca vivipara was most abundant in areas with a large proportion of wet coniferous forests and cool summers (R2adj = 36%, DW = 2.46, TPCA3 =-1.99, 20, TWetCon = 2.43, P=0.0086). Abundance of A. fragilis was negatively related to the proportion of agricultural landscapes (R2adj = 33%, p=0.0049, Figure 3). And the abundances of the two snake species were both predicted by the same climate factor (PCA 1), but it had opposite effects. Natrix natrix was more abundant in areas with relatively mild winters (R2adj = 26%, p=0.013, Figure 4) and Vipera berus - in areas with relatively cold winters (R2adj = 20%, p=0.028, Figure 5).
Figure 3 Relationship between the relative abundance of A.fragilis and the proportion

of agricultural landscapes (p=0.0049)
Component+Residual Plot for Afragilis

Figure 4

Relationship between the relative abundance of N. natrix and PCA 1

(mild winters; p=0.013)
Component+Residual Plot for Nnatrix

Figure 5

Relationship between the relative abundance of V.berus and PCA 1

(mild winters; p=0.028)
Component+Residual Plot for Vberus

Due to the given abundance pattern, the latter species may face threats from global warming. There has been a distinct climate-warming trend in Europe during the 20th century, with a mean increase in annual temperatures of about 0.8 °C. As the temperature increase has been particularly evident during the winter period (IPCC 2001), it could have an adverse effect on Vipera berus. This species prefers to hibernate in collective dens on slopes with southern exposure (Viitanen 1967, Prestt 1971), where the snow cover is less stable. Reduced snow cover in winter could cause a drastic increase in winter mortality of snakes due to freezing (Shine & Mason 2004). Thaws with following frosts would less likely happen in the areas with harsher climate, which possibly explains the observed abundance pattern for V. berus in Latvia.

Anguis fragilis was the only species of reptile in which agricultural development adversely affected abundance. However, this species is still the second most abundant reptile in Latvia and it is unlikely that agriculture poses a threat to its persistence due to the low density of the rural population in Latvia (<15 inhabitants per km2) and the declined coverage of agricultural lands (Berkgaute et al. 1999).
  1   2   3

База данных защищена авторским правом © 2016
звярнуцца да адміністрацыі

    Галоўная старонка