The research focused on a site in the Roztocký háj–Tiché údolí Nature Reserve, along the Únětický stream on the Prague–Central Bohemia border. Like all Bohemian populations of the species, the lizards there are bound to what ecologists call the river phenomenon: deeply incised river valleys whose south-facing slopes trap warmth and allow thermophilous organisms to persist far beyond the climatic conditions of the surrounding landscape. Because these valley populations cannot easily form metapopulations with one another, each survives as a relict, and their habitats — often rocky steppes and heath fragments — differ dramatically from the surrounding landscape matrix. Molecular work has shown that the Bohemian lizards carry noticeably lower genetic diversity, heterozygosity and allele richness than their South Moravian relatives connected to the species’ core range, making every remaining population genetically precious.

To understand what the lizards actually select at the microhabitat scale, the researchers conducted an unusually rigorous field campaign. Between 2011 and 2014, a single observer walked a 1.9-kilometre linear transect through the population’s 4.2-hectare home range — two abandoned quarries and a rocky, sun-exposed slope — a total of 60 times, spread across 119 site visits during the entire activity season. Every lizard seen was logged with GPS, its sex and age category recorded, and its original position noted if the observer’s approach flushed it. The positional data were remarkably precise: post-processing with geodetic software achieved accuracy within 50 centimetres for 87.6 percent of the records and within one metre for another 9 percent.

The result was 403 presence records across four years, averaging 6.7 lizards per transect walk, with peak detections in May and June. For comparison, the team generated 200 random points within the same habitat polygon using QGIS, and the shortest distances between real and random points — calculated with the Haversine formula — averaged just 6.9 metres. That tight spatial overlap made the discrimination task demanding, yet the lizards’ choices turned out to be anything but random.

The analytical core of the study combined principal component analysis and discriminant function analysis. Twenty-four environmental variables, visually assessed within radii of 0.5 and 2.5 metres of each point — the 2.5-metre figure chosen because it approximates the maximum distance an adult can run without stopping — were first screened for spatial correlation using the Mantel test. Only the percentage of exposed soil exceeded the correlation threshold and was removed, leaving nine variables. The resulting model was highly significant, with Wilks’ Lambda of 0.598, and classified 82.5 percent of random points and 79.9 percent of presence points correctly — a strong performance given how close the two sets of points lay to each other.

The variables that mattered tell a clear ecological story. Within half a metre of the observation point, the key positive features were elevated rock, scree-related elements and the presence of shelters; within 2.5 metres, the count of shelters and deep shelters usable for overwintering or egg-laying became decisive. At the opposite pole, the strongest negative predictors were tall grass and high vegetation coverage. The principal component analysis condensed this into interpretable vectors: one dominated by elevated rocks and shelters, reflecting thermoregulation and antipredator needs, since a raised position offers basking and a vantage point while nearby cover enables rapid escape; a second consisting of bushes and thorny shrubs, again oriented along the direction of recorded presence.

The shrub finding was striking in its consistency: 95 percent of recorded observation points had at least one shrub or shelter within 0.5 metres, and 99 percent had at least one within 2.5 metres. Lizards disturbed by the observer were frequently seen darting first into a nearby shrub and only retreating underground if pursued further. Intriguingly, the lizards showed no preference between thorny and non-thorny shrubs, and the discriminant analyses found no meaningful differences in microhabitat use between adult males and females, nor a convincing ontogenetic shift between juveniles and adults — a contrast with patterns reported from other populations of the species.

The negative association with tall vegetation has an equally practical explanation. Dense grass and high herbs shade the ground surface, lowering temperatures the lizards need for basking, and simultaneously block their field of view without providing genuine protection from predators — a dangerous combination. Scree occupied a more nuanced position: its vector neither tracked lizard presence nor pointed away from it. Scree warms quickly in the sun but surface temperatures measured on site with an infrared thermometer can reach 60 degrees Celsius in direct sunlight, making it unusable during hot periods, and the gaps between stones serve as cover mainly for juveniles and subadults rather than larger adults. With the frequency of hot days rising in the region, the authors note that scree-dependent habitat use could shift in the future.

What elevates this study beyond local natural history is its explicit bridge between scales and its direct management prescriptions. Earlier state-wide modelling across the whole Czech Republic had identified the macrohabitat envelope of the species — annual precipitation up to 600 millimetres, slope inclination of 5 to 25 degrees, mean temperature of the warmest quarter up to 20 degrees Celsius, and winter precipitation above 150 millimetres — and the new microhabitat findings fit comfortably within that framework. The authors recommend that managers of similar isolated populations maintain potential shelters no more than about five metres apart, keep the landscape mosaic heterogeneous, retain shrubby cover against aerial predation, and above all prevent the overgrowth of habitat by tall vegetation, ideally through a combination of grazing and cutting.

The site itself offers encouraging evidence that such care works. Conservation management at Tiché údolí has run since 1998, and by 2013 both population density and the area used by the lizards had risen significantly compared with the late 1990s, reaching levels comparable to populations in similar habitats elsewhere. Yet the study’s authors caution that isolation still poses a high long-term risk, since small, isolated populations are considered the most vulnerable to habitat erosion under climate change. As multivariate methods of this kind remain rare in reptile habitat research despite their routine use in botany and invertebrate ecology, this Czech valley offers both a template and a warning: save the rocks, the shrubs and the open sun-warmed patches, and a blue-throated relict of a warmer Europe may yet hold on at the very edge of its range.

Subject of Research: Microhabitat preferences of isolated European green lizard populations

Article Title: Microhabitat preferences in the European green lizard ( Lacerta viridis ): implications for conservation management of isolated populations

Article References: Chmelař, J., Frynta, D., Rudolfová, V., Fischer, D., & Rehák, I. (2024). Microhabitat preferences in the European green lizard (Lacerta viridis): implications for conservation management of isolated populations. Herpetozoa, 37, 287-294. https://doi.org/10.3897/herpetozoa.37.e120806

Image Credits: AI Generated

DOI: 10.3897/herpetozoa.37.e120806

Keywords: Lacerta viridis, European green lizard, microhabitat, conservation management, reptile ecology, Bohemia, river phenomenon, discriminant function analysis, habitat fragmentation, thermoregulation, rocky steppes, isolated populations