Background & Summary

Biodiversity crisis is a major ongoing problem occurring at the global scale, spanning all types of environments1,2,3. Such crisis is directly or indirectly enhanced by multiple human activities that determine, or favour, the conditions that pose species at risk of extinction4. Among them, environmental changes, and the introduction of new species (either competitors or pathogens) have substantial impacts. In the first case, the relatively fast environmental changes, for example caused by climate change or by pollutants, can quickly create conditions that overcome species physiological limits, preventing their potential gradual adaptation to the novel environmental conditions and driving populations to a dramatic end5,6,7,8. In the second case, native species have to deal with newly introduced species that may be stronger competitors9,10 or that can overcome immunological barriers of individuals11,12. All these processes often result in the progressive disappearance of local biodiversity.

Amphibians are the most threatened vertebrates worldwide11,13. These vertebrates are generally characterized by a series of specific traits that synergistically enhance their sensitivity to human impacts. For example, many amphibians are characterized by a biphasic life cycle, and their life history stages require both aquatic and terrestrial environments. That means that disturbance in just one of the two environments can provoke significant effects that can lead to the local extinction of the species. Further, amphibians completely rely on the environment for their physiological homeostasis, for example to maintain the body temperature within their optimal range or to balance the water loss14,15,16. Considering that most of amphibians are able to exploit a narrow range of environmental conditions17,18, even a small change in such conditions can be unbearable. Furthermore, amphibians are often characterized by a low dispersal ability19, a characteristic that hampers species to track their preferred conditions, thus exacerbating the impacts of environmental changes.

Long-term monitoring is a necessary prerequisite to set effective protection of amphibians and prevent their declines and extinctions20. Repeated surveys are indeed crucial for early identification of population decline and the emergence of potential threats21, allowing timely actions and mitigations of related harmful effects22,23. With this dataset we provide data from repeated surveys performed on the strictly protected Speleomantes cave salamanders, the only plethodontid species present in Europe24,25. There are eight allopatric species of Speleomantes, endemic to specific areas distributed in mainland Italy, in the Republic of San Marino, in south-eastern France and on the island of Sardinia24,26. Populations of natural hybrids (between S. ambrosii and S. bianchii) are also known from a small contact zone located in Apuan Alps (Northern Tuscany)27. Their overall conservation status is alarming: one species is Near Threatened, one Vulnerable, four Endangered and two Critically Endangered (www.iucnredlist.org; last access on 05 February 2024). A mandatory monitoring activity of protected species has been imposed by the European Union (Article 17 of the Habitats Directive; https://nature-art17.eionet.europa.eu/article17/) with the aim of constantly updating their conservation status; however, very limited data is currently available for the genus Speleomantes. In addition to providing a large amount of data covering a four-year period, this dataset offers the possibility of being combined with those previously published28,29,30, allowing the monitoring period for some populations to be extended to seven years. The availability of standardized data from multiple conspecific populations can also increase the possibility of detecting intraspecific variation as key traits. This can contribute to the conservation of not just species, but also of populations and intraspecific diversity31,32,33

Methods

The monitoring activity involved multiple surveys on 66 populations of Speleomantes salamanders in the period 2021–2024 (Tables 1 and 2). The surveyed populations were subterranean or fully epigean. In the first case, the subterranean environments (both natural and artificial) were surveyed entirely or up to the point where the exploration required the use of speleological equipment; in surface environments, specific plots delimited the study area. The surveys were carried out producing an average sampling effort of approximately 27 m2/minute34. Speleomantes were opportunistically searched and captured within the sampling area: these were the active individuals and those spotted hiding under stones or inside crevices. All captured individuals from a population were temporarily placed in disinfected (using bleach) fauna boxes until processing. A portable photo studio has been set in the field which allows high quality images to be obtained from which it is possible to extrapolate biometric data35,36. This portable studio is designed to obtain high-quality images by controlling the position of the camera and lights (i.e. flash), allowing individuals to be shot perpendicularly from above and producing comparable images useful for individual recognition through the use of the dorsal pattern37. This also allows a high standardization of the coloration of salamanders35. During each survey, we shot a first image taken on a Pantone Calibrite ColorChecker Passport Photo 2 (for brevity, hereafter Pantone); this allowed us to obtain the white reference for image calibration, and the length reference to insert within each image a as a unit of measurement. Salamanders were weighed on a digital scale (0.01 g) and visually inspected to record multiple data.

Table 1 Summary of the surveys carried out in the period 2021–2024.
Full size table

An initial inspection was carried out to assess the presence of ectoparasites, i.e. leeches of the genus Batracobdella, which are known to parasitise the Sardinian Speleomantes24,38 (Fig. 1). When leeches were present, they were counted and weighed separately. Subsequently, a check was carried out for any malformation (e.g., forked tail, foot shape39,40;) (Fig. 2). In some of the monitored populations, individuals were previously marked using either the Visual Implant Elastomers41 and Visual Implant Alpha tags42, we used an UV light to assess the presence of these tags. Tags were placed on arms, flanks and the base of the tail (for further information on tag implantation refer to41,42). The individuals’ head was then checked to assess the presence of the “mental gland”, a male sexual character located under the lower jaw24; all salamanders found with this character were considered adult males. Considering the exclusivity of males in showing distinctive morphological characters, adult females and juveniles are generally distinguished based on body size24. Speleomantes can lose and regrow their tail to avoid predation43; this makes total length (TL) unreliable for making this distinction. We then used the snout-vent length (SVL, in mm) to size individuals and distinguish between juveniles and adult females44. The threshold was set based on the average minimum size observed for gravid females and mature males45,46: for the three continental species (S. strinatii, S. ambrosii and S. italicus), hybrids and for S. genei it was set at 50 mm, while for the other “larger” Sardinian species (S. flavus, S. supramontis, S. imperialis and S. sarrabusensis) to 55 mm. This distinction was only possible in the laboratory after estimating the SVL of individuals from the images (see below).

Fig. 1
figure 1

Example of the high-quality images composing the dataset. In this picture an individual of Speleomantes flavus is parasitized by a leech of the genus Batracobdella. Scale bar = 10 mm.

Full size image
Fig. 2
figure 2

Some examples of malformations observed in Speleomantes during our study.

Full size image

Camera RAW files (.CR2) were opened and processed with Adobe Photoshop. For each population we first opened the Pantone photo with Adobe Camera Raw and set white calibration using the White Balance Tool function; the resulting profile was used to white balance all photos taken during a single session. We then set the scale to 10 mm using the Pantone reference measurement; the same amount of pixels was used for the entire population thanks to the fixed position of the camera35. The images were cropped and cleaned by adding a white background. After adding the reference scale they were transformed into JPEG, allowing to reduce the weight of the file (~2–3 MB) without compromising its quality28 (Fig. 1). We then used Fiji47 to estimate individuals SVL36. After setting the image scale (straight line) via the scale bar, we used the segmented line to draw a line from the snout to the tip of the tail, following the centre of the body; this allowed to measure the individual’s total length. Another segmented line was drawn from the tip of the snout to the extremity of the body, that is the point that corresponds to the opening of the cloaca36; this allowed us to estimate SVL. The differences between TL and SVL were used to obtain tail length.

Data Records

The dataset presented here (Speleomantes photographic dataset 2021–2024, available on figshare48) consists of 3,836 high-quality images of individuals from 66 Speleomantes populations, including the eight known species distributed in mainland Italy (including the Republic of San Marino) and Sardinia24, as well as hybrid populations occurring in the north (i.e., natural populations) and south (i.e., introduced populations) of Tuscany27,49. Detailed information on Speleomantes populations (coded following28) is shown in Tables 1 and 2.

Table 2 Summary of the surveys carried out in the period 2021–2024.
Full size table

Each image comes with additional information about the individual’s biometric data (e.g., size, weight) and condition (e.g., presence of parasites, malformations) (Information on Speleomantes individuals48). The .CSV supporting dataset is composed as shown in Table 3.

Table 3 Description of the accompanying dataset “Information on Speleomantes individuals”.
Full size table

Technical Validation

Surveys were performed during a single day and by operators adopting the same sampling effort. The standardisation and quality of images provided here is guaranteed by the adoption of the best protocol described for these species28,29. The snout-vent and total length were estimated through analysis of the high-quality images provided here; this methodology allows reliable estimates with very small measurement error (~2 mm) even if measures are taken by a non-expert operator36. Potential errors due to measurement or transcription were double checked by plotting the data (SVL vs TL and TL vs weight) (Fig. 3). Anomalous data were compared to paper field data, and dubious measurements were repeated. Considering the SVL threshold used to distinguish juveniles (50 and 55 mm for normal and giant species, respectively), all individuals with SVL lower than the respective threshold were considered juveniles.

Fig. 3
figure 3

Linear regression plots used to evaluate the presence of erroneous data. (A) Correlation between SVL and tail length; (B) correlation between TL and weight.

Full size image

Usage Notes

This dataset has broad application. In addition to being an important tool for estimating and monitoring the abundance of populations over time, and for studying demographic parameters (e.g., age distribution, survival, growth rate) as specified above, we list some other important uses here. The conditions of individuals can be evaluated from different points of view: in addition to verifying the presence of ectoparasites and malformations, using the ratio between total length (TL) and weight we can evaluate the body condition of each individual50. This fitness-related trait allows inter- and intrapopulation comparisons in space and time51,52, making it possible to evaluate factors that negatively influence individual condition53. In this circumstance, the use of TL in estimating individual body condition is more appropriate, since these salamanders usually accumulate fat in the tail54. The images can be used for geometric morphometry55, for coloration analyses47,56,57, and can also be used to test the reliability of different software for the automatic photographic identification of individuals58,59,60.