Plant Diversity in Archaeological Sites and Its Bioindication Values for Nature Conservation: Assessments in the UNESCO Site Etruscan Necropolis of Tarquinia (Italy)

Abstract

In archaeological sites, plants can be a risk for monument conservation. However, in these sites, a refugium for plant biodiversity is often detected, such as in the UNESCO site Etruscan necropolis of “Monterozzi” in Tarquinia, which still holds a Special Protection Area for bats. In this site, we previously evaluated the positive and negative effects of vascular plants on the conservation of the hypogeal tombs. To contribute in assessing the role of archaeological sites in supporting plant diversity and interpreting its bioindication values for nature conservation, we analyse in this relevant place the floristic interest and richness and the plant communities growing on tumuli, trampled, and less disturbed areas. The results revealed the presence of several plants with high naturalistic interest, such as the community’s representative of synanthropic and natural Mediterranean grasslands, which arise both from the present and the past uses of the area. The high naturalistic values of the site are also assessed, considering its remarkable richness of species/area compared with the well-known archaeological sites of Rome. These findings further indicate that plant diversity needs to be considered in planning management activities in archaeological sites to also protect their natural values.

1. Introduction

Plant growth in archaeological areas can cause several problems for the management and conservation of archaeological monuments, leading to biodeterioration [1,2,3,4,5,6,7,8,9]. Biodeterioration can be caused by mechanical stress on stone materials, including root damage to the walls of buried archaeological structures, or by chemical erosion processes due to the release of acid metabolites and chelating agents [6,10,11,12]. Then, it is also necessary to evaluate the characteristics, benefits, or harmful potential of species that may ultimately avoid or mitigate conflict between vegetation and archaeological ruins [13,14,15].

Furthermore, in addition to the important role of maintaining the natural vegetation cover in such areas, previous studies have also shown that archaeological sites can provide refuges for plant and animal species, protecting them from human impact and stresses associated with urban development and thus helping to conserve their levels of biodiversity [16,17,18,19,20,21,22,23,24,25,26,27,28,29]. Indeed, cultural heritage management activity provides some protection for natural heritage compared to other highly disturbed human ecosystems. This allows the growth of vegetation with natural or semi-natural features in these areas, with species and plant communities of conservation interest [28,30,31]. In some instances, it has been observed that the latter could even quadruple in number compared to other human-modified contexts, where they are often threatened [29,32,33,34,35,36]. Furthermore, vegetation cover may positively influence the conservation of monuments, as in the case of several hypogeal structures, via maintaining constant microclimatic conditions and thus limiting deterioration phenomena mainly due to water evaporation and phenomena of salt efflorescence [14,37,38]. Moreover, vegetation can provide interesting tools to define a preventive conservation plan, thereby reducing the weathering effect, protecting the monument and contributing to their conservation against aggressive weathering agents [39,40].

Several investigations on the use of plant species or communities as bioindicators of edaphic and environmental conditions, such as of human and animal activities, have been carried out in recent decades, especially in urban environments, to achieve a deeper knowledge of the natural plant heritage and its protection from the threatening expansion of urban areas [27,41,42,43,44]. In fact, plant species, with their diversity, structure, chorology and autecology, represent valuable tools for analysing ecosystems and understanding changes in the main environmental factors which occur at spatial and temporal levels [45,46,47,48,49,50]. Protecting biodiversity is also essential for a range of goods and services that are of fundamental importance to human well-being and the functioning of ecosystems [48]. Therefore, species loss adversely affects the overall functioning of an ecosystem since the latter is determined by combined effects [51,52]. Consequently, it is essential to focus on how ecosystems can be managed to provide ecosystem services to society and the environment efficiently. Indeed, a great deal of international effort has been directed towards conserving biodiversity, including green areas and infrastructure in anthropised environments, to advance global biodiversity and sustainability goals [53,54,55,56,57,58].

In archaeological areas, the study of plant biodiversity and the ecological dynamics by which it is driven is functional both to the conservation of the species and habitats they contain and to the promotion of sustainable tourism. In fact, management of archaeological areas carried out taking into account the naturalistic peculiarities of these areas allows for planning of the modalities of tourist fruition and of maintenance operations in order to prevent harmful effects on the vegetation [59,60,61,62,63,64,65].

Therefore, to contribute to the naturalistic knowledge of an archaeological area of relevant cultural interest, we selected the Etruscan archaeological site of “Monterozzi” in Tarquinia, which has been included in the UNESCO World Heritage Sites (WHS) since 2004 together with the site of Cerveteri [66]. This site has a particular naturalistic relevance since it contains a SAC (Special Area of Conservation), code IT6010028 (Habitats Directive 92/43/EEC), for the protection of five species of bats and is also included within a broader SPA (Special Protection Area), code IT6030005, for the conservation of various plants, animals, and habitats. Its naturalistic value was preliminarily confirmed by our recent floristic analysis of the area [67], showing that the plant communities host a high floristic richness, with the presence of protected, endangered, or rare species for the region, as in the case of several orchids. We also carried out in the site research on the influence of the vegetation cover on the conservation of underground wall paintings [10,14,68], such as on the identification of herbaceous roots penetrating in the hypogeas [10]. However, no studies have been carried out here on the bioindication role of the plant species and communities growing on the site and the associated environmental drivers.

Then, this research aims to further contribute to the knowledge of plant diversity in archaeological sites as bioindicators of edaphic and environmental factors and to the related naturalistic values. In particular, we will analyse (I) the general floristic characteristics, defining their naturalistic interest; (II) the ecological and syntaxonomical characteristics of vegetation types and their species richness; and finally, (III) bioindication values of vegetation types about the past and present conditions.

2. Materials and Methods

2.1. The Site

The Etruscan “Monterozzi” Necropolis of Tarquinia, which was used from the 7th to the 2nd century BC, holds paintings of extraordinary value, and is “the only significant evidence of pre-Roman classical art existing in the Mediterranean basin” [66]. The site is located in central Italy, around 4 km from the Tyrrhenian Sea and 70 km NW from Rome, at an altitude of around 200 m a.s.l.

The area falls within the Mediterranean bioclimate (characterised by a lower mesomediterranean thermotype and an upper dry ombrotype) [69]. Meteorological data on temperature and rainfall obtained from the public archive database show an average annual rainfall of 648.2 mm, with an average monthly value of 55.0 mm, with July being the driest (20.7 mm) and November the wettest (102.5 mm) (http://www.idrografico.regione.lazio.it/annali/index.htm, accessed on 1 July 2022). Temperatures show a certain variability among the seasons, with an average annual temperature of 15.7 °C, the temperature of the coldest month 9.6 °C, the temperature of the warmest month 24.3 °C, and the average minimum temperature of the coldest month 3.3 °C.

The Mitrakos indexes [70] were used for defining both the intensity and duration of cold (stress from winter cold) and aridity (summer drought stress). Such indexes are defined as follows:

-

Monthly temperature stress (MCS) = 8 (10 − t) where t = average minimum monthly temperature [°C];

-

Winter Cold Stress (WCS) = sum of MCSs of December, January and February;

-

Year Cold Stress (YCS) = sum of MCSs of all months of the year;

-

Monthly Drought Stress (MDS) = 2 (50 − p), where p = monthly rainfall [mm];

-

Summer Drought Stress (SDS) = sum of MDSs of June, July, and August;

-

Year Drought Stress (YDS) = sum MDSs of all months of the year.

Data from [71] showed that the values of Winter Cold Stress in Tarquinia were not high (WCS = 76.5, YCS = 91.5), whereas the values for the Summer Aridity Stress were quite high (SDS = 166.6, YDS = 184.8).

The necropolis rises on a Pliocene plan and includes more than 6000 underground tombs excavated in the rock and covered with soil mounds, which were often destroyed. The geological substrate of the site is mainly composed of clays and marls [72], sometimes with olistostromes in the northern parts and detritic and organogenic limestones in the southern one [73].

Our study area covers approximately 5.7 ha, with a plant cover constituted mainly of grasslands interspersed by small shrub formations in some parts. The grassland plant cover is managed with mowing activities several times during the year.

2.2. Floristic Analysis and Evaluation of the Naturalistic Interest

During spring and autumn 2020–2021, we carried out a broad survey in the area to collect floristic information on species and life forms, observing and picking the plant individual that was further dried and stored as herbarium samples. Plant species were identified using the “Flora d’Italia” [74], which was also helpful for attributing chorological data. We also checked and updated the plant names according to the “The World of Flora Online project” [75] and the Portal to the Flora of Italy version 2021.2 [76].

Then, we assessed the natural value of the species (based on conservation interest, rarity, vulnerability, and endemism), following a methodology proposed by Caneva [13] and successively applied by Cicinelli et al. [77]. In particular, we considered their inclusion in lists protected by local, national, and international laws and their inclusion in international directives (Habitat Directive—92/43/EEC and the CITES Convention on International Trade in Endangered Species of Wild Fauna and Flora), and in the red list of threatened vascular plants in Italy [78], Latium Regional Law on protected spontaneous plants (l.r. 19 Settembre 1974, n. 61). Further information on frequency and relictuality were obtained via Atlases of the vascular flora of Latium [79,80], highlighting the first record in the site and attributing the following categories: RR = extremely rare; MR = very rare; and PC = uncommon [79,80].

2.3. Ecological and Syntaxonomical Characteristics of Vegetation Types and Their Species Richness

During spring 2021, we also carried out 42 vegetation surveys within the study area on tumuli, trampled areas and other zones less subject to visitor influence (Figure 1), using the phytosociological approach of the Zurigo-Montpellier school [81].

According to such methodology, in each plot of 25 m², carried out in a homogeneous zone, we identified the occurring species, assessing their cover values (using the coverage index of plants of the Braun–Blanquet scale: + = 0.5%; 1 = 1–5%; 2 = 5–25%; 3 = 25–50%; 4 = 50–75%; 5 = 75–100%), and the most relevant environmental variables and edaphic factors, such as slope, aspect, altitude and soil features.

To describe the plant composition and the ecological characteristics of the communities, we performed a cluster analysis and an NMDS (Non-Metric Multidimensional Scaling) ordination graph for quantitative data (abundance). Non-metric multidimensional scaling (NMDS) is an unconstrained analysis, which, unlike methods which attempt to maximise the variance or correspondence between objects in an ordination, attempts to represent, as closely as possible, the pairwise dissimilarity between objects in a low-dimensional space. The dissimilarity matrix of the data was calculated using the Bray–Curtis dissimilarity index, which takes into account the abundance and presence or absence of species in each dataset to assess the diversity and compare the composition of species in different environments. A hierarchical cluster analysis was performed on this matrix using the mean agglomeration method UPGMA (unweighted pair group method with arithmetic mean), which uses the unweighted arithmetic averages of the measures of dissimilarity, thus avoiding characterising the dissimilarity by extreme values (minimum and maximum). The Silhouette index [82], which studies the separation distance between clusters by measuring how close each point in one cluster is to points in the neighbouring clusters, was used to decide the optimal number of clusters. A dendrogram of the dissimilarities between samples and between species was then derived and sorted according to distance matrices. An indicator species analysis of the individual clusters was then performed, which identifies the associations between the species or combinations of them and the clusters using the Indval index, which considers the indices of fidelity and specificity [83,84]. A synoptic table was elaborated using the percentage frequency of occurring species among the different clusters. The ordination graph of sampling sites was obtained by the NMDS method. To highlight the ecological variations between the vegetation clusters, we projected environmental variables measured in the field on the NMDS plots.

Furthermore, we carried out a syntaxonomic analysis of the plant communities based on the ecological interpretation, which was suggested by a wide literature and synthesised in the Italian Vegetation Prodrome [85]. We also considered a wider analysis of European vegetation, and finally, the syntaxonomic classification and nomenclature follow [86] and for some particular syntaxa [87].

Finally, we evaluated the species richness per plot in the Monterozzi site to discriminate the richness among the different identified plant communities. We compared this richness and the occurrence of species with conservation interest collected in the site with those obtained from bibliographic data of all Roman archaeological areas [27]. For the latter, we selected only 12 sites, for a total of 63 relieves, comparable in size areas and similar vegetation types to our surveys. Thus, we considered that the species–area relationship is approximately described by a power function of the form S = cAz, which is usually presented as a log10 transformation of variables [88]. We then applied the formula log₁₀ S = log₁₀ C + z log₁₀ A (S is the number of species and A is the area size) and thus performed regression analysis. Following the same method, we performed a residual analysis of species richness across the different vegetation clusters of the Monterozzi Site. Software R (Version 4.2.1) [89] was used for all statistical analyses with the packages: ade4, vegan, gclus, cluster, vegclust and indicspecies.

2.4. The Bioindication Values of the Different Plant Communities

For analysing their bioindication values, we obtained information from the scientific literature on the habitats usually hosting the detected floristic elements [85,86,87]. Furthermore, we also collected the Ellenberg Indicators Values (EIVs; see [90,91]) of the species to derive information on the influence of the associated environmental variables: light (Light), temperature (Temp), moisture (Moist), nitrogen (Nutr), continentality (Cont) and soil pH (React). We calculated the average indicator values of such communities, also considering the cover values of the species, and then we compared them to highlight differences.

3. Results

3.1. Floristic Analysis and Evaluation of Naturalistic Interest

Improving the previous floristic analysis [92], where 110 species have been described, in the present plots, we detected 167 species, included in 39 families and 126 genera. The most represented families were Asteraceae (35 = 19.6%), Fabaceae (24 = 13.4%) and Poaceae (22 = 12.3%), while the most species-rich genera resulted in Trifolium (7) (Fabaceae), Medicago (5) (Fabaceae), Crepis (4) (Asteraceae), Plantago (4) (Asteraceae) and Erodium (4) (Geraniaceae). The most frequent species are listed in

Table 1. List of species with a conservation interest.

	Frequency (Latium)	Endemics	CITES	Latium Reg. Law
* Ajuga iva	PC
* Carduus nutans	PC
Centaurea aspera	MR
Crepis bursifolia L.	PC
Crepis capillaris	RR
* Helianthemum apenninum	PC
* Helianthemum salicifolium	PC
Hyoseris radiata	PC
Linaria purpurea		X		X
* Onosma echioides	PC	X
Ophrys bombyliflora			X
Ophrys tarquinia		X	X
Serapias parviflora	PC		X
* Trifolium spumosum	RR

Frequency is based on [79] (RR = extremely rare; MR = very rare; PC = uncommon). Species marked with an asterisk are those found for the first time at the site; the others had already been mentioned in [67].

, which enhances the highest frequency of Hyoseris radiata, Plantago lanceolata, Seseli tortuosum, Salvia verbenaca and Avena barbata.

From a floristic point of view, Therophytes, which exceed 50%, are the most recurrent species, followed by Hemicryptophytes (34%), and in decreasing order, Geophytes (8%), Chamaephytes (4%) and Phanerophytes (4%). Most species have a Mediterranean chorotype (61%) with a similar contribution of the Euri-Mediterranean (33%) and Steno-Mediterranean (28%) species. Species with a wide distribution and Eurasian species follow next with a frequency of 18% and 16%, respectively. Several endemic, boreal and Atlantic species are also present, with a lower percentage. Erigeron sumatrensis resulted in the only exotic species.

Several occurring species have a particular conservation interest, as shown in Table 1, and are infrequent, rare or very rare occurrences in Latium [73], as in the case of Trifolium spumosum, of which only very few reports exist in this region.

Furthermore, several species that were found to have a high frequency within the site are relatively rare or not very common at the regional level. For example, the species Hyoseris radiata, which has the highest frequency at the site (88%), is uncommon in the region, and Centaurea aspera, which has a high frequency at the site (69%), is reported to occur very rarely in the region. In addition, several other protected species (Figure 2) were observed: endemic species, species protected by regional regulations and several orchid species listed by CITES.

3.2. Ecological and Syntaxonomical Characteristics of Vegetation Types and Their Species Richness

3.2.1. Ecological Statistical Evaluations

The dendrogram of the cluster analysis concerning dissimilarities between relieves in the plots enhanced the presence of different communities (Figure 3). Separating the clusters at the highest level, the resulting main groups were (1) (Cl. A) shrub formations of Mediterranean scrub dominated by Rhamnus alaternus (Figure 4A); (2) (Cl. B1) therophytic grasslands with high cover of the chamaephyte Teucrium capitatum and with frequent occurrence of several geophytes, such as the orchids Serapias parviflora and Ophrys bombyliflora (Figure 4B1); (3) (Cl. B2) grasslands dominated by Asphodelus ramosus (Figure 4B2); (4) (Cl. C) variable herbaceous synanthropic vegetation growing on grave mounds (Figure 4C); (5) (Cl. D) grasslands dominated by Dasypyrum villosum and Avena barbata (Figure 4D); and (6) (Cl. E) grasslands dominated by Malva sylvestris and Hordeum murinum subsp. leporinum (Figure 4E).

The grasslands dominated by Asphodelus ramosus and therophytic grasslands showed a lower separation since some of these sampling areas were grouped together, and some of them were also mixed with the grave mounds. All the samples carried out on the grave mounds fall into a main cluster, even if with a certain further differentiation. In fact, the tombs of Leonesse and Fiore di Loto create a separate cluster from all the other tombs.

The indicator species of the different groups according to the IndVal index, which considers the indices of specificity (A) and fidelity (B), is reported in Table S1. The most indicative species of Cluster E include Hordeum murinum subsp. leporinum, Erodium cicutarium and Calendula arvensis. Concerning Cluster D, among the most indicative species, we found Silene italica, Vicia hybrida, Anchusa azurea and Avena sterilis (Figure 4D). Cluster A is characterised by Convolvulus cantabrica, Urospermum dalechampii and Trifolium stellatum for most of the tombs and, more rarely, Carduus pycnocephalus, Convolvulus althaeoides and Campanula erinus. For Clusters B1 and B2, Lysimachia linum-stellatum, Linum bienne, Teucrium capitatum, Sanguisorba minor, Asperula aristata and Ononis reclinata resulted in the most indicative species, whereas for Cluster C, they were Brachypodium rupestre, Allium roseum and Ornithogalum narbonense.

The ordinations with the arrangement of clusters obtained via the NMDS in an ecological space are shown in Figure 5. Such ordination confirms the results from the cluster analysis concerning the organisation of the groups.

The variability observed along the x-axis seems to follow anthropogenic disturbance, which has the highest values in Cluster D and E and the lowest in Cluster A. The variability explained by the y-axis seems mainly linked to the soil characteristics since the detected shrublands and grasslands (Clusters A–E) grow in deeper soils, and the other therophytic grasslands grow mainly in thin soils with rockiness.

3.2.2. Syntaxonomic Analysis

The syntaxonomic analysis of the vegetation growing at the Monterozzi archaeological site showed a certain variability and richness of the vegetation types present in the area. Such types describe communities or fragments which are representative of different alliances, orders, and classes (

Table 2. Synoptic table showing percentage frequency of occurring species among the different clusters. Clusters: A—R. alaternus communities, B1—Therophytic grasslands communities, B2—A. ramosus communities, C—Grave mounds communities, D—D. villosum communities, E—M. sylvestris communities.

Synoptic Table
Clusters	A	B1	B2	C	D	E
Oleo-Ceratonion siliquae Br.-Bl. ex Guinochet et Drouineau 1944
Phillyrea latifolia L.	III	I
Clematis flammula L.	V	III
Rhamnus alaternus L.	V		II
Asparagus acutifolius L.	V
Rubia peregrina L.	V
Pistacia terebinthus L.	III
Pistacio lentisci-Rhamnetalia alaterni Rivas-Mart. 1975 and Quercetea ilicis Br.-Bl. ex A. Bolòs et O. de Bolòs in A. Bolòs y Vayreda 1950
Fraxinus ornus L.	I
Chamaerops humilis L.	I
Brachypodium rupestre (Host) Roem. & Schult.	IV	II
Spartium junceum L.	V	I
Hordeetum leporini Br.-Bl. (1931) 1936 malvetosum sylvestris Braun-Blanquet et al., 1952
Hordeum murinum subsp. leporinum (Link) Arcang.						V
Malva sylvestris L.				III	IV	V
Echium plantagineum L.		I	II	IV	IV	I
Crepis bursifolia L.	I			I	I	II
Hordeion leporini Braun-Blanquet (1931) 1947
Erodium ciconium (L.) L’Hér.				I	V	III
Capsella bursa-pastoris Medik.			II		I	IV
Anthemis arvensis L.	I	V	V	V	III	IV
Medicago polymorpha L.				II	IV	IV
Plantago lagopus L.	I	IV	III	II		V
Diplotaxis tenuifolia (L.) DC.		I			II	I
Geranium molle L.			II	I	V	IV
Veronica polita Fr.		III	III		I	I
Galactites tomentosus Moench						I
Vulpio ligusticae-Dasypyretum villosi Fanelli 1998 and Securigero securidacae-Dasypyrion villosi Cano-Ortiz, Biondi et Cano in Cano-Ortiz et al. ex Di Pietro in Di Pietro et al., 2015
Dasypyrum villosum (L.) P. Candargy		I	III	III	V
Avena barbata Pott ex Link	I	III	V	V	V	V
Reichardia picroides (L.) Roth	II	V	V	V	IV	II
Avena sterilis L.			II	II	IV
Foeniculum vulgare Mill.			II	III	I	II
Lotus ornithopodioides L.	I	III		IV	III
Festuca ligustica Bertol.		III		I	I	II
Brometalia rubenti-tectorum (Rivas Godayet Rivas-Mart. 1973) Rivas-Mart. et Izco 1977 and Chenopodietea Br.-Bl. in Br.-Bl. et al., 1952
Crepis sancta (L.) Babc.		I		II	V	V
Bromus madritensis L.			II	I	III	I
Convolvulus arvensis L.					IV	II
Reseda lutea L.	I			II		I
Papaver rhoeas L.				II	I
Poa annua L.	I			I		IV
Scandix pecten-veneris L.	I			II		I
Senecio vulgaris L.						II
Tordylium apulum L.	IV	II	V		II	II
Raphanus raphanistrum L.					II
Vicia sativa L.				II	IV
Sonchus oleraceus L.			II	II	II
Carduus pycnocephalus L.	III	I
Ajuga chamaepitys (L.) Schreb.			II
Nigella damascena L.		I
Rostraria cristata (L.) Tzvelev		IV		II
Calendula arvensis M.Bieb.		I	II	III	III	V
Cerastium ligusticum Viv.		III	II		II	I
Sherardia arvensis L.		IV	III	V	IV	V
Trifolium campestre Schreb.	I	V	IV	III	I
Euphorbia helioscopia L.	II	III	V	V	IV	II
Hypochoeridion Achyrophori  Biondi et Guerra 2008
Hypochaeris achyrophorus L.		IV	II	III
Medicago truncatula Gaertn.		IV	IV	V	IV	II
Ononis reclinata L.		V	V	IV
Linum strictum L.		II
Sixalix atropurpurea (L.) Greuter & Burdet	I	IV	III	II
Catapodium rigidum (L.) C.E.Hubb.		IV	V			I
Helianthemum salicifolium (L.) Mill.		IV	III			II
Medicago minima (L.) L.		I	IV	I	II
Campanula erinus L.		V		I
Aegilops geniculata Roth		II	V	IV	IV
Ptilostemono stellati-Vulpietalia ciliatae  Mucina ined. and Stipo-Trachynietea distachyae S. Brullo in S. Brullo et al., 2001
Dactylis glomerata L. subsp. hispanica (Roth) Nyman	II	V	V	V	V	IV
Festuca ambigua Le Gall		IV	V	III	I
Trifolium scabrum L.		IV	V	IV	I
Trifolium stellatum L.		IV	V	V	IV
Briza maxima L.	I	II	II	I
Crepis neglecta L.	II	IV	III	I
Lagurus ovatus L.			II	I
Hippocrepis biflora Spreng.		I	II	I
Onobrychis caput-galli (L.) Lam.		V	II	IV	I
Plantago bellardii All.		II		I
Romulea columnae Sebast. & Mauri		I	IV	I
Parentucellia latifolia (L.) Caruel		III	III
Scorzoneretalia villosae  Kovač ević 1959 and Festuco-Brometea Br.-Bl. et Tx. ex Soó 1947
Onosma echioides (L.) L.		III	II	I
Anthyllis vulneraria L. subsp. rubriflora (DC.) Arcang.		III	II	I
Teucrium capitatum L.		IV	III	I
Satureja montana L.		II
Helianthemum apenninum Mill.		II	III
Ophrys bombyliflora Link	I	IV	III	I
Ophrys tarquinia P. Delforge		I
Centaurea aspera L.	III	III	V	IV	IV	IV
Serapias parviflora Parl.	I	IV	II	I
Convolvulus cantabrica L.		V	IV	V		II
Petrorhagia saxifraga Link		I	II	IV	II	II
Thesium humifusum DC.		I
Sanguisorba minor Scop.	I	V	IV	II
Bromus erectus Huds.		I	II	II	I	II
Galium lucidum All.		III
Artemisietea vulgaris  Lohmeyer et al. in Tx. ex von Rochow1951
Plantago lanceolata L.		V	V	V	III	II
Convolvulus althaeoides L.				II		I
Echium vulgare L.	I					I
Picris hieracioides Sm.	I			II		I
Poa bulbosa L.		II		I		II
Medicago sativa L.	I	I	II	III	IV	V
Crepis capillaris (L.) Wallr.		II	II		I
Linum bienne Mill.		IV	III
Salvia verbenaca L.	I	V	V	V	II	IV
Leontodon tuberosus L.	IV	IV	III	III	I	I
Tragopogon porrifolius L.			II	III
Asphodelo ramosi-Feruletum communis  Biondi et al., 2016
Asphodelus microcarpus Rchb.	III	III	V	I
Ferula communis L.			V	I
Asphodelo ramosi-Ferulion communis  Biondi et al., 2016 and Asphodeletalia ramosi Biondi et al., 2016
Anemone hortensis L.	IV	I	IV	I	II
Carlina corymbosa L.		I
Charybdido pancratii-Asphodeletea ramosi  Biondi et al., 2016
Hyoseris radiata L.	IV	V	V	V	III	IV
Galium lucidum subsp. corrudifolium (Vill.) Bonnier	II	II
Leopoldia comosa (L.) Parl.		I	III	I
Scrophulario-Helichrysetalia S. Brullo 1984 and Drypidetea spinosae Quézel 1964
Chondrilla juncea L.				I
Linaria purpurea (L.) Mill.	I	I	II	II
Micromeria graeca (L.) Benth. ex Rchb. subsp. tenuifolia (Ten.) Nyman	I			I
Urospermum dalechampii (L.) Scop.		V	IV	V	IV	II
Companions
Astragalus hamosus L.		I	III	II	III	IV
Seseli tortuosum L.	IV	V	V	V	II	II
Silene italica (L.) Pers.		I			V	II
Vicia hybrida L.		I		I	V	I
Erodium cicutarium (L.) L’Hér.		I	III	II		IV
Lysimachia linum-stellatum L.		V	V	I		I
Verbascum sinuatum L.		I		II	I	III
Plantago media L.	I					II
Cynodon dactylon (L.) Pers.		I	II	II
Cerastium glomeratum Thuill.		II	IV	II	I	IV
Valeriana eriocarpa (Desv.) Christenh. & Byng		III	III	II		I
Erodium malacoides (L.) L’Hér.						II
Erodium moschatum (L.) L’Hér.						II
Theligonum cynocrambe L.						II
Anchusa azurea Mill.					III
Anisantha diandra (Roth) Tutin ex Tzvelev				I	I
Trifolium spumosum L.					I
Orobanche ramosa L.	I	I	II		I
Ajuga iva (L.) Schreb.		III	III	II
Allium roseum L.	I	I		II
Anisantha rubens (L.) Nevski		I		I
Bellardia trixago (L.) All.		I
Blackstonia perfoliata (L.) Huds.		IV	II	I
Carduus nutans L.		III	II	II
Centaurium erythraea Rafn		IV
Clinopodium nepeta (L.) Kuntze		I	II	III
Crupina vulgaris Pers. ex Cass.			II
Filago germanica (L.) Huds.		III
Galium murale All.			II
Hedypnois rhagadioloides (L.) F.W.Schmidt			II
Lysimachia arvensis (L.) U.Manns & Anderb.		V	III	II
Phleum hirsutum Honck.			II	I
Poa trivialis L.			II
Polygala monspeliaca L.	I	II	II
Silene pendula L.		I
Sinapis alba L.	I		II
Stachys ocymastrum Briq.			III
Trifolium resupinatum L.		I	IV
Hypericum perfoliatum L.		I		I
Lathyrus cicera Hauman			II	II
Medicago orbicularis (L.) Bartal.			II	I
Borago officinalis L.		I			I
Bellis perennis L.			II		I	I
Sonchus bulbosus (L.) N. Kilian & Greuter	III			I
Ornithogalum narbonense L.	I			I
Pallenis spinosa (L.) Cass.	I			I
Lolium perenne L.	I			I
Anthyllis circinnata (L.) D.D.Sokoloff	I
Arabis hirsuta (L.) Scop.	I
Cirsium vulgare (Savi) Ten.	I			I
Helichrysum italicum (Roth) G.Don	I
Euonymus europaeus L.	I
Sonchus asper (L.) Hill	I			II
Fumaria capreolata L.	I
Lactuca serriola L.	I			I
Cuscuta epithymum L.	I			II

,

Table 3. Syntaxonomic scheme.

1. ANTHROPOGENIC VEGETATION

1.1 MEDITERRANEAN SEGETAL AND RUDERAL VEGETATION OF MAN-MADE HABITATS

CHENOPODIETEA Br.-Bl. in Br.-Bl. et al., 1952

Brometalia rubenti-tectorum Rivas-Mart. et Izco 1977

Hordeion murini Br.-Bl. in Br.-Bl. et al., 1936

Hordeetum leporini Br.-Bl. (1931) 1936

malvetosum sylvestris Braun-Blanquet et al., 1952

Securigero securidacae-Dasypyrion villosi Cano-Ortiz, Biondi et Cano in Cano-                  Ortiz et al. ex Di Pietro in Di Pietro et al., 2015

Vulpio ligusticae-Dasypyretum villosi Fanelli 1998

1.2 MEDITERRANEAN AND SUB-MEDITERRANEAN VEGETATION OF PERENNIAL GRASSLANDS IN POST ABANDON OF AGRO-PASTORAL ACTIVITY

CHARYBDIDO PANCRATII-ASPHODELETEA RAMOSI Biondi et al., 2016

Asphodeletalia ramosi Biondi et al., 2016

Asphodelo ramosi-Ferulion communis Biondi et al., 2016

Asphodelo ramosi-Feruletum communis Biondi et al., 2016

1.3 PERENNIAL (SUB)XEROPHILOUS RUDERAL VEGETATION OF THE TEMPERATE AND SUBMEDITERRANEAN REGIONS

ARTEMISIETEA VULGARIS Lohmeyer et al. in Tx. ex von Rochow1951

2. NATURAL GRASSLANDS

2.1 MEDITERRANEAN CALCIPHILOUS ANNUAL AND EPHEMEROID SWARDS AND GRASSLANDS

STIPO-TRACHYNIETEA DISTACHYAE S. Brullo in S. Brullo et al., 2001

Ptilostemono stellati-Vulpietalia ciliatae Mucina ined.

Hypochoeridion achyrophori Biondi et Guerra 2008

2.2 SUBMEDITERRANEAN DRY GRASSLAND AND STEPPE VEGETATION

FESTUCO-BROMETEA Br.-Bl. et Tx. ex Soó 1947

Scorzoneretalia villosae Kovač ević 1959

3. SUBMEDITERRANEAN VEGETATION OF SCREE AND ROCKY HABITATS

DRYPIDETEA SPINOSAE Quézel 1964

Scrophulario-Helichrysetalia S. Brullo 1984

4. MEDITERRANEAN FORESTS AND SCRUB

QUERCETEA ILICIS Br.-Bl. ex A. Bolòs et O. de Bolòsin A. Bolòs y Vayreda 1950

Pistacio lentisci-Rhamnetalia alaterni Rivas-Mart. 1975

Oleo-Ceratonion siliquae Br.-Bl. ex Guinochet et Drouineau 1944

and Tables S2–S4 as synthetic descriptions) [78,79].

3.2.3. Floristic Richness Assessment of the Plant Communities

The graph of floristic richness in the different plant communities (Figure 6) shows a gradient from the therophytic grasslands with the dominance of Teucrium capitatum (B1) with the highest species richness, followed by grasslands dominated by Asphodelus ramosus (B2). The latter are followed by tumulis’ vegetation (C) and grasslands dominated by Dasypyrum villosum and Avena barbata (D). After that, we find the grasslands dominated by Malva sylvestris and Hordeum murinum subsp. leporini (E), and last are the shrub formations of Mediterranean scrub dominated by Rhamnus alaternus (A).

Species richness and occurrence of species with conservation interest in the Monterozzi site surveys resulted much higher than in the surveys of several Roman archaeological areas with similar vegetation types and sizes and the therophytic grasslands with the dominance of Teucrium capitatum show the highest richness (62 species), while the shrublands showed the lower richness (20 species) (Figure 7a). In fact, in the occurring range of the logarithm scale, the increase of one point triplicates the values of plant diversity.

Regarding the occurrence of species with conservation interest (Figure 7b), again, the Monterozzi surveys show much higher values than the regression line. All surveys of Monterozzi have at least one species with conservation interest, while many surveys of Roman archaeological areas have none. Therophytic grasslands of Monterozzi, with the dominance of T. capitatum, also contain the highest number of species with conservation interest, such as several orchids.

3.3. The Bioindication Values of the Different Plant Communities

The graph of Ellenberg Indicators Values (Figure 8) showed that, in general, the plant communities describe conditions of high solar radiation (LIGHT 7–9) and values of transition among xeric Mediterranean-mountain and typical xeric Mediterranean ones (TEMP 7–9). It is confirmed by the continentality value (CONT 5), and the xeric conditions of the site were also indicated by the low hydric values (MOIST 3). The average soil preference for pH values (REACT 5–7) and for nutrients (NUTR 3–5) highly vary. The latter two indicators show greater variability. Specifically, regarding reaction, therophytic grasslands (cluster B1) and grave mounds communities (cluster C) show higher pH values, while A. ramosus and D. villosum communities lower. Regarding nutrition, on the other hand, the more synanthropic grasslands (Clusters B2, C, D and E) show higher values than the more natural therophytic grasslands (Cluster B1).

The ecological characteristics of the dominant species in the vegetation allow using them as bioindicators of the present environmental conditions or also those occurring in the past.

Grasslands dominated by M. sylvestris and H. murinum subsp. leporinum (Cluster E) are indicators of disturbed land subject to trampling, with a certain accumulation of debris [66,86]. Furthermore, when M. sylvestris is dominant (as in our case), it indicates a deeper soil, rich in nitrates and subject to frequent trampling [74,93].

Grasslands dominated by D. villosum and A. barbata (Cluster D) are typical grasslands of the so-called “Campagna Romana” [94,95,96], and they indicate a Mediterranean xerophilous herbaceous vegetation.

Grasslands dominated by A. ramosus (Cluster B2) are related to grazing and fires, and such Mediterranean geophyte shows a relevant resprouting capacity from the rhizome. This species is favoured by heavy grazing and recurrent fires [74,97]; in fact, it flowers massively in burned areas [98].

Therophytic grasslands hosting high cover of T. capitatum and with frequent occurrence of several orchids such as S. parviflora and O. bombyliflora (Cluster B1) refers to vegetation in hot and arid Mediterranean environments, on stony, embryonic, or decapitated soils [74,85].

The Mediterranean scrub dominated by R. alaternus (Cluster A) refers to the scrubland’s vegetation of basal and hilly belts, with moderate resilience in fire-disturbed contexts.

4. Discussion

The floristic and vegetational assessment of the Monterozzi site revealed high plant biodiversity, with still remaining naturalness and interesting features of relictuality.

4.1. Floristic Analysis and Evaluation of the Naturalistic Interest

The dominance of therophytes over hemicryptophytes confirmed the character of the present bioclimate of the site, in particular of the lower mesomediterranean thermotype, and the effects of the past and current anthropogenic disturbances [99]. The relevant spread of geophytes (8%) can be justified by the presence of rather evolved soils and the presence of some habitats less affected by human disturbance [100], especially in the north-eastern area of the site, farthest from the entrance and bordering Mediterranean scrub vegetation. In addition, there is a higher percentage (60%) of bulbous geophytes, which are usually more abundant in grassland areas or synanthropic environments, than rhizomatous geophytes, which are associated with wetter areas. Instead, the considerable occurrence of chamaephytes (8%) can be explained by the presence of thin soils and rock emergences in the north-central part of the site due to the remnant fragments of coastal boulder and travertine beds, which were more widely distributed in the past over all the surrounding areas [80].

The marked prevalence of Mediterranean and sub-Mediterranean components and Eurasian ones is due to the bioclimatic conditions, and the high proportion of widely distributed species is due to the anthropic disturbance in the area [101].

Several species of special conservation interest were found within the site. First and foremost, T. spumosum, a thermo-xerophilous species mainly found in the arid and uncultivated soils of coastal areas with a very fragmented and declining distribution in Italy, especially in the central regions [102]. In Latium, there are very few findings around Rome and near Gaeta and Tarquinia [103]; in the latter two cases, however, only bibliographic or field data and not herbarium samples are available.

4.2. Ecological and Syntaxonomical Characteristics of Vegetation Types and Their Species Richness

Despite our study not considering the influence of seasonal variations influence, the syntaxonomic analysis showed a certain heterogeneity of the vegetation types, containing small relictual fragments of steppic zones and pastoral grasslands mixed with the expected Mediterranean vegetation types.

These remnant fragments record some natural vegetation types as therophytic grasslands growing on travertine or coastal macco. Indeed, in the Tarquinia area, there is a marine terrace at a relatively low altitude, dating from the Upper Pliocene-Lower Pleistocene, which was home to heterogeneous and highly biodiverse vegetation, of which only a few fragments remain today [80]. The flora found in these environments are very similar to that of the travertines, with many steppic species adapted to higher elevation environments, such as some chamaephytes present in site H. apenninum, H. salicifolium, T. capitatum and O. echioides, which are relicts of vegetation that managed to be preserved following the last glaciation. In a large portion of the site, therophytic grasslands with T. capitatum are bioindicators of thin soils and superficial bedrocks. They are mixed in a patchy distribution with A. ramosus grasslands, which are a consequence of the past pastoral activities showing a long-time permanence. After that, due to the resilience of several therophytic species of calcareous soils, especially in the areas with higher outcrops of coastal macco and stony soils, therophytic vegetation with T. capitatum can become dominant again. This would also explain the similarity in species composition between clusters B1 and B2. As observed by several authors, when grazing is managed in a non-intensive way, it can favour greater biodiversity, and the effects of such activities can remain for a long time (in this case, for several decades) [104,105,106]. Furthermore, the higher richness of clusters B1 and B2, shown in Figure 6, can be related to microhabitats created by the variability of bedrocks and erosion phenomena, which favours a great amount of therophitic species requiring lower nutrient and water availability [107,108]. The separation between the mounds of the Leonesse and the Fiore di Loto tombs (Cluster C1) with respect to the other tombs (Cluster C2) is mainly due to the artificialisation of the vegetation cover to which these two tombs have recently been subject.

In general, the resulting species richness is very relevant not only when considering the archaeological areas in the Roman context (the double) [32] but also other grasslands in a wider Euro-Mediterranean context [109]. Furthermore, the occurrence of species with conservation interest found within the single plots seems mainly due to the high naturalness of this area and its high value in terms of the conservation status of floristic and vegetation elements. These results are in line with a preliminary study of flora on this site [67] and other data from other similar situations, like those collected in the near necropolis “La Banditaccia”, located a few kilometres away in Cerveteri, and included in the same UNESCO World Heritage Site [22,92].

Based on the floristic results, it is recommended that the site management would support the conservation of the species of conservation interest occurring in the site.

4.3. The Bioindication Values of the Different Plant Communities

Although the study did not explore the historical human activities and historical land use practices in the site, since we found such information not easy to retrieve, the bioindication value can replace this lack of providing important information.

On the other hand, grasslands dominated by M. sylvestris and H. murinum subsp. Leporinum, according to [32,33], are common in the Roman archaeological areas, where they grow along the margins of paths and on soils where pedogenesis is embryonic and anthropogenic disturbance is high. These communities are bioindicators of herbaceous formations that grow on abandoned lands where pedogenesis is embryonic and anthropic disturbance has a certain incidence [43].

This vegetation type has often been found in spatial contact with grasslands dominated by D. villosum and A. barbata, similar to other archaeological areas [17,33], where, if trampling disturbance decreases, grasslands dominated by M. sylvestris and H. murinum subsp. leporini tend to evolve towards D. villosum and A. barbata grasslands. Bibliographic data report that it develops in rather poor soil often reworked and made up of heterogeneous materials [33]. Thus, this vegetation type is typical of abandoned fields and uncultivated areas growing, so are witness that several zones in the area were cultivated in the past. Such as the therophytic grasslands witnesses of arid soils abandoned after (non-intensive) cultivation or extensive grazing [74,85].

The shrub formations dominated by R. alaternus, located furthest from the graves and the visitor’s path, indicate a higher naturality. These communities are, in fact, intermediate stages of vegetation series culminating in mature sclerophyllous evergreen forest stages linked to slight natural or anthropogenic disturbance [76]. This vegetation represents an intermediate stage of the vegetation series culminating in mature forest stages of Q. ilex forests, affected by natural and anthropogenic disturbances [85].

The analysis of the Ellenberg indicator values (Figure 8) properly describes the ecological pattern of the site, and usually, the values do not vary considerably among the different communities. Instead, the greater variability of soil reaction and nutrient values better explain some differences between the communities. In particular, the unpredicted lower nutrients requirement of the detected A. ramosus communities (Cluster B2), compared to the other synanthropic grasslands (Clusters C, D and E), seems due to the high bearing in such communities of species belonging to the therophitic grasslands (Cluster B1), which have a significantly lower nutrient requirement.

4.4. Management and Conservation Strategies

Valorisation of the presence of this natural heritage on the site is important, as well as making the tourists aware of it. To achieve this aim, it could be worthwhile to install signage that informs visitors about the naturalistic value of the site and communities with conservation interests they may encounter in order to encourage visitors not to damage them. Furthermore, it can be recommended the delimitation of areas where such species are mostly found, highlighting their presence and their value. In this way, the limited access of visitors in these areas would not cause an issue but rather a stop-off point where the visitors can understand both the beauty of cultural and natural heritage of the site.

In the management plan and in the conservation strategies, it should be important to find a balance between the protection of monuments and the conservation of the natural environment. In this respect, it is important to have a coherent management plan with limited weeding activities in areas that contain the most valuable natural elements and to schedule them after their processes of dissemination do not obstruct their reproduction. Moreover, the weeding activities must be selective to the plant species that in the concrete have a biodeteriogen activity against the building heritage.

5. Conclusions

This work enhancing the natural heritage in a cultural place showed how a detailed analysis of the ecological characteristics of the plant species and of their bioindication values gives useful information for reconstructing past and present activities and for supporting a sustainable management plan. The latter should consider the safeguarding of natural richness in the planning of weeding and maintenance activities and in the modalities in which sites are made accessible to visitors. As the first analysis on this site to consider changes in vegetation and use bioindication values, we encourage for future research. Furthermore, we underline the importance in protecting and promoting the sustainability and conservation of natural and cultural heritages.