Rome, with approximately 3 million residents, holds the third position amongst all European Union cities, following Berlin (4 million people) and Madrid (3.2 million people). Urban Rome is typically defined as the region within the motorway ring known as the Grande Raccordo Anulare (GRA), which encompasses an area of approximately 360 km² (Fattorini 2011a, Di Pietro et al. 2021). The Appia Antica Regional Park (41°50’00” N – 12°33’00” E) is a protected green space covering roughly 4,580 hectares, extending from Rome's city centre to the surrounding rural zones of the Alban Hills. This protected area includes archaeological sites along the historical Appian Way, from Rome's centre to the 10^th Mile, such as the Villa of the Quintilii and the Tombs of Via Latina. The Park also includes the Caffarella Park and the Park of the Aqueducts.

The Appia Antica Regional Park has a highly diversified landscape, with a mosaic of cultivated fields and natural and semi-natural areas. Vegetation is mainly represented by Mediterranean maquis, including species such as Pistacia lentiscus L., Rhamnus alaternus L. and Euonymus europaeus L. Due to the millenary human presence, ruderal species such as Sonchus asper (L.) Hill, Pteridium aquilinum (L.) Kuhn and Cymbalaria muralis G.Gaertn., B.Mey. & Scherb., as well as cultivated species, such as Olea europaea L. and Prunus amygdalus Batsch, are widely distributed. The area also incorporates fragments of wet meadows and ponds, along with watercourses with associated vegetation (Populus nigra L., Salix alba L., and Ulmus minor Mill.). Further information on the vegetation of the study area can be found in Buccomino and Stanisci (2000), Ceschin et al. (2006) and Iamonico (2022).

Sampling was conducted in nine sites, distributed along a transect of about 8 km, at increasing distances from the city (Fig. 1): the closest site was at about 5.5 km from the city centre and was surrounded mainly by built-up areas (minimum distance from buildings: 300 m); the most peripherical site was at about 12 km from the city centre, near the GRA, i.e. at the borders of the city. A minimum distance of 500 m separated sampling sites.

Figure 1.  

Study area. a: Location of Rome Municipality (red circle) within the Italian territory. Latium Regioni is in black. b: Location of the study area (in green) within the Rome Municipality (red line). GRA is the Grande Raccordo Anulare, a motorway conventionally used to define the urban area of Rome. S.L. = Sealing Layer. c: Land use categories occurring in the study area. Numbers indicate location of sampling sites, numbered according to the urban-rural gradient.

Caffarella Park (Site 1, 19 traps): This site is highly heterogeneous, including a small forest area, a pond, areas with sparse vegetation, shrub vegetation, grasslands, riparian vegetation with sparse willow and poplar trees, remnants of old constructions.

San Sebastiano (Site 2, 7 traps): This site corresponds to the archaeological area of catacombs of San Sebastiano, an underground cemetery located along the Via Appia Antica. Traps were placed in the area between the building of the Basilica of San Sebastiano and Via delle Sette Chiese. Vegetation is mainly artificial, including trees and herbs, which are frequently mowed and cut.

Tor Marancia (Site 3, 7 traps): This site is located near a densely populated area with numerous archaeological remains. In 2002, the City Council of Rome and later the Council of Rome decided to establish the Tor Marancia Park as a part of the Appia Antica Park. Vegetation in this site includes a woodland and cultivated fields.

Acqua Santa (Site 4, 6 traps): This site is mainly occupied by open vegetation and wooded edges. The site is partially crossed by Via dell'Acqua Santa, a road with no motorised traffic. Traps were placed close to trees to minimise potential disturbance.

Farnesiana (Site 5, 5 trap): Vegetation in this site was mostly herbaceous, with a predominance of gramineous plants (cultivated fields); a few trees are present in small groups. Ruins of old buildings are also present.

Cava Fiorucci (Site 6, 7 traps): This site (located a few dozen metres south of State Road 7 - via Appia Nuova) has an irregular surface, with depressions and ditches due to past excavation activities. Vegetation includes sporadic bushes and trees; a small gravel road crosses the quarry.

Casal Verbeni (Site 7, 5 traps): This site is characterised by the presence of a few old buildings in the centre of a cultivated area. Trees have been planted between and around the buildings for ornamental purposes and to visually isolate the area from the outside.

Torre Selce (Site 8, 5 traps): This site is adjacent to a section of the original Via Appia and is characterised by the ruins of a tower that stands on a large tumulus from the 1^st century. This site includes a small grassy area surrounding the tower, shrubby vegetation and a wooded patch.

Appia Antica 300 (Site 9, 3 traps): Traps were placed near the via Appia Antica, where there are bushes and trees, between the edge of an adjacent cultivated field and the road. The vegetation cover was characterised by a predominance of herbaceous species.

Pitfall traps consisted of clear plastic cups (diameter: 9.5 cm, depth: 15 cm) sunk in the ground with the cup-lip level with the soil surface and covered by sloped stones to limit the rainwater influx and capture of non-target taxa. Covering stones were elevated 5-7 cm above the ground using smaller stones at their corners. Each trap was filled with 250 ml of beer with salt, with a drop of unscented detergent to break the surface tension. The number of pitfall traps varied amongst sites, from a minimum of three traps to a maximum of 19 traps, depending on site habitat heterogeneity. In total, 64 pitfal traps were placed in the study area. Moreover, the number of recovered traps per site varied amongst sampling sessions because of trap damage and trap loss. Traps within sites were separated by at least 10 m from each other.

Sampling was conducted in four autumn sampling periods (from 18 October to 6 December 2013) and four spring sampling periods (from 7 May to 24 June 2014). Each sampling period lasted about ten days. We used this temporal distribution of sampling periods at the turn of two consecutive years because Mediterranean spring climatic conditions and, hence, biotic responses recorded in a given spring-summer are strongly influenced by those of the previous autumn-winter (Fanfani et al. 2014). For each trapping period, traps were active for a minimum of seven days to a maximum of twelve. Variation in trapping duration and time between trapping sessions was due to unstable weather conditions. Upon collection, trap content was rinsed with water and stored in vials with 70% ethanol.

Spiders were identified by the first author on morphological basis under A Zeiss DiscoveryV.12 stereomicroscope using multiple taxonomic keys (Roberts 1987, Trotta 2005, Nentwig et al. 2024). When necessary, female genitals were dissected, observed with an Olympus BX51 microscope and compared with images in Oger (2024). Nomenclature follows the World Spider Catalog (2024). Juveniles were not assigned to a species and a few specimens that were not identifiable due to bad conditions were not considered. All material is preserved in 70% alcohol in the Department of Science of the University of Roma Tre.

To describe the taxonomic composition of the ground spider fauna, we calculated the proportion of species in each family and compared these proportions with those that can be obtained at progressively larger scales, i.e. at the province (Rome Province), regional (Latium) and national level (the whole Italian territory), using the data reported in Pantini and Isaia (2019), updated with Trotta (2024), Decae (2024) and the new records given in the current paper. This led to the following values of number of recorded species: Rome Province 322 species, Latium Region 431 species, Italy 1716 species. It is important to take into account that these comparisons are biased by the specific sampling procedure used in our study. We collected only ground-dwelling species, whereas estimates at province, regional and national scale also consider species with different (e.g. arboreal) ecology. Thus, our values are underestimates of the total spider communities. Due to the non-linearity of the species-area relationship (SAR), it is not correct to divide the number of species by the size of the area to compare areas with different size (Fattorini 2021). Assuming that the SAR is best modelled by the power function, S = cA^z, where S represents species richness, A the area, c is the number of species per unit area and z is a measure of the rate of change in the slope with increasing area (Lomolino 2001, Fattorini et al. 2017), we calculated the c-parameter (c = S/A^z) as a measure of species richness standardised by area (Brummitt and Lughadha 2003, Ovadia 2003). To obtain realistic estimates of relative diversity, it is, however, important to use appropriate values of z. For example, for hotspot identification, Ovadia (2003) and Brummitt and Lughadha (2003) used a priori z-values such as 0.14 (Brummitt and Lughadha 2003), 0.18 (Ovadia 2003) and 0.25 (Brummitt and Lughadha 2003). Due tothe lack of empirical information on spider SARs in Italy, we conducted analyses using alternatively z = 0.14, z = 0.18 and z = 0.25, as they are representative of the full range of z-values commonly found in mainland and island systems (see Fattorini (2021)). As area values, we used A = 46 km² for the Appia Antica Regional Park, A = 5,363 km² for Rome Province, A = 17,232 km² for Latium Region and A = 302,073 km² for whole Italian territory.

To study possible variations in taxonomic composition at the assemblage level, we calculated the proportion of species in each family in the nine sampled sites. To express the contribution of each family in terms of abundance, we calculated the proportion of individuals belonging to each family both for the entire fauna of the study area and the nine sampled sites separately.

To describe the biogeographical composition of the Italian spider fauna, species were assigned to chorotypes, i.e. groups of species with similar distributions (Fattorini 2015, Fattorini 2016, Fattorini 2017a, Gatto and Cohn-Haft 2021). Chorotypes are established by an inductive and recursive process in which species distributions are mapped, their contours are compared and species with similar ranges are classified with the same group, i.e. they form a given chorotype (Fattorini 2015, Fattorini 2016, Gatto and Cohn-Haft 2021). After a chorotype is defined by the overlap of multiple species distributions, any other species showing a similar distribution can be assigned to that chorotype. Species that are classified under the same chorotype can belong to completely unrelated taxa or ecological groups. Thus, chorotypes can be compared to the concept of biota of Morrone (2014). As “abstractions” used to express recurrent species distributions, chorotypes are also roughly similar to the “generalised tracks” of Croizat (1958), although, in the case of chorotypes, no shared history is implied. Chorotypes are widely used both to shortly indicate species distributions and to make hypotheses about the origin of plant and animal assemblages (Fattorini 2015, Fattorini 2016, Gatto and Cohn-Haft 2021). As species with similar distributions should also have similar macroecological needs (Olivero et al. 2011), the analysis of the chorotype composition of local species assemblages can be used to draw inferences about which ecological and historical factors have shaped such assemblages.

We extracted species chorotypes from the checklist of Pantini and Isaia (2018), who followed the nomenclature proposed by Vigna Taglianti et al. (1999). For the species not included in Pantini and Isaia (2018) (e.g. new records for the Italian fauna reported here for the first time and other species not included in the checklist) and species for which the known distribution has changed, we examined individual distributions and assigned to each species the respective chorotype following the nomenclature proposed by Vigna Taglianti et al. (1999). For species endemic and subendemic to Italy, Pantini and Isaia (2018) used codes proposed by Vigna Taglianti et al. (1999) for recurrent patterns of species with restricted distributions. We categorised all endemic species as Endemic (END) and assigned subendemic species to the respective main chorotype as follows: Tyrrhenian subendemics were assigned to the West Mediterranean (WME) chorotype; Central-Apenninic, Apennino-Dinaric and Alpino-Apenninic subendemics were assigned to the S-European (SEU) chorotype. Therefore, the word ‘endemic’ is used to indicate the exclusive occurrence of a species in a defined geographical area (the Italian territory in our case) and does not necessarily imply narrow distributions (Fattorini 2017b).

The distribution of each species in the checklist was obtained from World Spider Catalog (2024).