Vie et milieu - Life and environment, 2013, 63 (3/4): 121-126

Breeding birds along a vegetation gradient in a Mediterranean landscape: contrasting patterns in spatial diversity

L. Zangari 1, M. Ferraguti 1,4, C. Battisti 3, L. Luiselli 2, M. A. Bologna 1 1 Dipartimento di Scienze, Università degli studi Roma Tre, viale G. Marconi, 446, 00146 Rome, Italy 2 Centre of Environmental Studies Demetra s.r.l., via Olona, 7, 00198 Rome, Italy 3 Servizio”Ambiente” (“aree protette-parchi regionali”), Provincia di Roma, via Tiburtina, 691, 00159 Rome, Italy 4 Present address: Department of Wetland Ecology, Estación Biológica de Doñana (EBD-CSIC), Seville, Spain Corresponding author: [email protected]

ΑLFA-DIVERSITY Abstract. – Diversity may be evaluated by adopting different hierarchical approaches. In ΒETA-DIVERSITY GAMMA-DIVERSITY this study, we analyzed the breeding bird communities of five habitats spatially distributed along HABITAT HETEROGENEITY a gradient of landscape vegetation in central Italy. We highlighted their structural differences at SPATIAL GRADIENT three hierarchical diversity metrics (α-, β- and γ-diversity). The beech forest showed the highest SPECIES TURNOVER averaged values in terms of both abundance and species richness (α–diversity), as well as in CENTRAL ITALY terms of Shannon index, while chestnut and oak woods showed the lowest values. Along the vegetation gradient we observed an inverse pattern of spatial diversity from the Mediterranean sclerophilic wood to the temperate and mesophilic beech forest. The beech forest was very rich at the point count scale (higher α-diversity) but had the lower value both in β- and γ-diversity; in contrast, the sclerophilic wood had a low α-diversity (point-scale) and the higher β- and γ-diversity values. The beech forest showed a low β-diversity (low habitat heterogeneity at the vegetation patch level), but due to the higher tree diameter and the related niche availability, it showed a high α-diversity (at point-scale). The sclerophilic wood, which was more simply structured, had a low α-diversity at point-scale, but a high β-diversity at the vegetation patch level, due to a high heterogeneity. This study demonstrates that patterns of diversity may be complex at different hierarchic levels. If conservation and landscape planning strategies are pro- moted at landscape scale, objectives should be carefully specified in terms of component of spe- cies diversity (α-, β-, γ-level).

INTRODUCTION Bird studies at the level of community allow to obtain information about its structure, functioning and level of Ecological diversity may be calculated in follow- disturbance and stress (Wiens 1989). Moreover, this line ing different hierarchical approaches (Whittaker 1960, of research is useful to obtain evidence on the conserva- Magurran 2004, Magurran & McGill 2011). Since spatial tion status of specific habitat types, utilizing specific indi- diversity is a concept with multiple components, Whit- ces and indicators at community and species level (Gibbs taker (1960) makes a distinction between α (or within- & Wenny 1993, Canterbury et al. 2000, Pereira & Cooper habitat)-diversity and β (or between-habitat)-diversity, 2006, Bani et al. 2009). where the former corresponds to the species richness in In Mediterranean ecosystems, a large number of stud- a defined spatial unit (e.g., a sampling site) and the latter ies on breeding bird communities of different habitats reflects the biotic changes or species replacements among have been carried out, but research in extensive mosaic sites due to recent ecological (e.g., habitat heterogeneity landscapes along altitudinal or vegetational gradients due to anthropogenic or natural disturbances) or histori- is still rare (Manzi & Perna 1992, Calvario & Sarrocco cal causes (Magurran 2004). The concept of β-diversity 1997, Vuerich et al. 2006). has been applied along environmental gradients, e.g., The aim of this study is to analyze the breeding bird vegetational or altitudinal (Wiens 1989, Magurran 2004). communities of five habitats distributed along an altitu- The concept of γ-diversity corresponds to a higher spatial dinal and a vegetational gradient in Tyrrhenian central level (e.g., at patch/landscape or regional scale; Whittaker Italy, highlighting their structural differences and simi- 1972; see the criticism in Gray 2000 and Magurran 2004). larities in their assemblage diversity at the level of both Is widely recognized as the hierarchical patterns in spatial classical community parameters and at the level of three diversity are often contrasting among them (Whittaker hierarchical spatial metrics (i.e., α-, β-, γ-diversity). Our 1972, Wiens 1989): this fact has large implication in the hypothesis is that the patterns of spatial bird diversity in selection of goals and objectives in wildlife management these habitat types might be different (and contrasting and conservation strategies (Magurran 2004, Magurran & among them) depending on the structural and spatial dif- McGill 2011, Dornelas et al. 2011). ferences in vegetation. 122 L. ZANGARI, M. FERRAGUTI, C. BATTISTI, L. LUISELLI, M. A. BOLOGNA

MATERIAL AND METHODS the presence of Q. ilex, Pistacia lentiscus, Myrtus communis, Erica arborea and Crataegus monogyna. Study area: The study area is a forest mosaic located in the (c) Chestnut wood at 400-500 m a.s.l. (locality: “Quattro bot- northern slope of the Mts (, Central tini”), dominated by Castanea sativa, and with the presence of Italy; Allumiere municipality), ranging from about 200 to 650 m Quercus ilex, Q. petraea, Q. cerris, Arbutus unedo, Ilex aquifo- a.s.l. The area refers to the bioclimatic transitional Mediterra- lium, Fagus sylvatica, and Fraxinus ornus. nean region (Blasi 1994) and is geologically characterized by (d) Mixed wood at 450-550 m a.s.l. (locality: “Sbroccati”), acid volcanic substrates and flyschoid sedimentary rocks (Devo- with dominance of Quercus cerris, Q. petraea, Castanea sativa, to & Lombardi 1977, Dowgiallo 1995; Fig. 1). Carpinus betulus, and with the presence of Q. ilex, Fraxinus We studied the assemblage of breeding birds occurring in the ornus, Ostrya carpinifolia, Acer opalus, Arbutus unedo, Ruscus following five habitats, which are distributed gradually on the aculeatus, Ulmus minor and Sorbus torminalis. This ecosystem hills’ slope along a vegetational and altitudinal gradient from is listed among the “Natura 2000” habitats of European inter- lower to higher (details in Contoli et al. 1980, Fanelli et al. est (“Castanea sativa forests” ; code 9260; 92/43/EU “Habitat” 2007, Di Pietro et al. 2010): Directive; Calvario et al. 2008) (a) Sclerophilic mixed wood at 200-300 m a.s.l. (locality: (e) Beech wood, at 500-600 m a.s.l. (locality: “Il Faggeto”), “Ripa Majale”), dominated by Quercus ilex, Q. suber, and with a habitat of very high conservation value for the unusually low the presence of Arbutus unedo, Phyllirea latifolia, Pistacia len- altitude and the coexistence of mesophilic (e.g. Fagus sylvatica, tiscus, Spartium junceum and Erica arborea. This vegetation Quercus petraea, Q. cerris, Carpinus betulus, Ostrya carpini- type is characterized by a high spatial heterogeneity at a fine folia, Acer monspessulanum, Acer pseudoplatanus, Ilex aqui- scale due to local historical disturbances. folium, Castanea sativa, Ruscus aculeatus) and Mediterranean (b) Termophilic mixed wood at 250-350 m a.s.l. (“Bosco di species (e.g. Quercus ilex, Erica arborea, Spartium junceum, Torcimina”), characterized by a prevalence of Quercus cerris, Crataegus monogyna). This ecosystem is listed among the Q. pubescens, Acer monspessulanum, A. campestre, and with “Natura 2000” habitats of European interest (“Apennine beech

Fig. 1. – Map of the study area. Point counts: S1-Sn: for SLC; T1-Tn: for TER; C1-Cn: for CHE; F1-Fn for BEE; Q1-Qn for OAK. See Methods and Table I for acronyms and further details.

Vie Milieu, 2013, 63 (3/4) PATTERNS IN SPATIAL DIVERSITY OF BREEDING BIRDS 123 forests with Taxus and Ilex; code 9210; 92/43/EU “Habitat” outside the fixed radius and time (S*); (iii) the mean number Directive; Scoppola & Caporali 1998, Lorenzetti et al. 2007, of species (Sm), representing the ratio between the sum of the Calvario et al. 2008, Romeo et al. 2010). species recorded in all point counts and that of point counts: this value corresponds to an α-diversity mean index (see Magurran

Protocol: For the characterization of the structure of breed- 2004); (iv) the β-diversity index (βW), as the ratio between S (or ing bird communities in the five forest habitats, during the 2011 γ-diversity index calculated at level of vegetation patch level) spring period (March-June), we used the point count method and Sm (the α-diversity mean index: Whittaker 1960, Whit- (Bibby et al. 2000, Bibby 2004), with fixed radius (100 m) and taker 1972): this value may be an indicator of the intra-habitat fixed time (5 min), recording the observed and/or heard spe- type heterogeneity; (v) the Shannon-Wiener diversity index (H’) cies within 40 sampling points (eight points in each habitat (Shannon & Weaver 1963), calculated as H’ = –Σ fri ln fri (where type; “point count method”; Fig. 1). After the field study from fri.is the ratio between the number of individual of species i and the general data-set, we selected only the breeding bird species the total number of individuals). locally occurring, comparing data to the Regional Breeding Bird For all statistical analyses, we used the SPSS 11.0 and Eco- Atlas (Brunelli et al. 2011). Sim700 softwares for Windows. We used non-parametric statis- The spatial localization of the sampling points was carried tical tests (Kruskal-Wallis ANOVA followed by post hoc Mann- out by GPS, according to the WGS84 projection, to assure a Whitney pairwise comparisons with Bonferroni correction to minimum distance of at least 200 m between each other to allow test for differences among habitat types relating to total Abm independent data and to avoid pseudoreplication. The elevation and Sm). To test the correlation between Abm and Sm and two of each point (m a.s.l.) was recorded. We carried out two ses- environmental parameters (altitude and mean tree diameter), we sions in each point: one in March-April (early spring; aimed to used a Spearman rank correlation test. These two environmental check for sedentary non migrant species), the second in May- parameters were not collinear (r = 0.015; p = 0.998). June (late spring; aimed to check for long-distance migratory species; Blondel et al. 1981). To control for possible observer effects, samplings were RESULTS performed by two or three observers contemporarily (MF, LZ and usually MAB or CB). Sampling sessions were realized in We recorded 568 individuals belonging to 29 breeding the early morning between 7.00 and 11.30 a.m. (solar time) to bird species (Table I). Turdus merula, Sylvia atricapilla, increase the sampling efficiency, given the behavioural charac- Cyanistes caeruleus, Parus major were the dominant spe- teristics of the breeding species (Bibby et al. 2000). All sam- cies in all habitat types (Table I). The beech forest had the pling points were located inside the core of each habitat type (at highest values in mean abundance, while chestnut and oak least 300 m from the boundary), to avoid an edge effect on bird woods showed the lowest values. Mean abundance values counting. Because of the point count method is focused only were significantly different among habitats (H = 19.03; on ‘common species’ (i.e., species locally more abundant, dif- p < 0.05, Kruskal Wallis test), and the beech forest values fuse and easily detectable; Bibby et al. 2000), we did not obtain significantly differed from all other habitats (p < 0.01). data on rare and less detectable species. Finally, since all habitat The beech forest also had the highest value in mean types were characterized by tree and shrub vegetation and all breeding bird species had a territorial behaviour (i.e., not gre- species richness (an α-diversity metric), while chestnut garious), we assumed a comparable bird song detectability, at and oak woods showed the lowest values. In addition, least for ‘common species’. mean species richness was significantly different among To assess the effect of physiognomic and structural vegeta- habitat types (H = 16.45; p < 0.01; Kruskal-Wallis test), tion factors at both the level of whole breeding bird communi- the value of beech forest being significantly higher in ty and single species, we measured in the surrounding area of comparison to sclerophilic, chestnut and oak woods each point, for an extent of radius equal to 50 m, the mean tree (p < 0.01). diameter, calculated at breast height on 10 randomly selected The Shannon diversity index revealed a pattern similar plants (utilized as predictor of structural complexity at point to the α-diversity metric, that was the highest value being scale). In each habitat type we calculated for each bird species: observed in the beech forest, and the lowest values in (i) the mean abundance Index (Abm), as the ratio between the both chestnut and oak woods (Table II). The sclerophilic th total number of individuals sampled of the i species (Ni) and and chestnut woods had the highest values in β-diversity the number of points (eight); (ii) the relative frequency (fri), index (> 2); the beech forest the lowest (Table II). i.e., the ratio between the number of recorded individuals of Comparing single point counts, both the mean tree each species and the total number of individuals in the whole diameter and the altitude were positively correlated communities. At the community level, we also calculated: (i) (rs = 0.379; p = 0.016, n = 40). Both the number of spe- the number of species recorded by the point count method (S): cies and their total abundance were not correlated to this value corresponds to a γ-diversity index (sensu Whittaker either altitude (respectively, rs = 0.238 and rs = 0.135, all 1972); (ii) the total species richness, represented by the whole p > 0.05) or mean tree diameter (respectively, rs = 0.309 number of recorded species, including the species sampled and rs = 0.197, all p > 0.05).

Vie Milieu, 2013, 63 (3/4) 124 L. ZANGARI, M. FERRAGUTI, C. BATTISTI, L. LUISELLI, M. A. BOLOGNA

Table I. – Detected bird species and relative abundance in the five forest habitats. ABm = mean abundance (Ni /total number of point counts); Ni = number of sampled individuals; fri= relative frequency. In bold, the dominant species (fri > 0.05). SCL: Sclerophilic mixed wood; TER: Termophilic mixed wood; CER: Chestnut wood; OAK: Oak mixed wood; BEE: Beech wood. SCL TER CHE OAK BEE

Species ABm fri Ni Abm fri Ni Abm fri Ni Abm fri Ni Abm fri Ni Columba palumbus 0.63 0.04 5 0.13 0.01 1 0.38 0.03 3 Streptopelia turtur 0.38 0.03 3 Upupa epops 0.13 0.01 1 Picus viridis 0.63 0.04 5 0.5 0.04 4 0.5 0.05 4 0.75 0.04 6 Dendrocopos major 0.13 0.01 1 0.38 0.03 3 0.5 0.03 4 Lullula arborea 0.13 0.01 2 Troglodytes troglodytes 0.63 0.04 5 0.63 0.04 5 0.63 0.06 5 0.25 0.02 2 1.25 0.07 10 Erithacus rubecula 0.63 0.04 5 1.25 0.09 10 2 0.18 16 2.38 0.22 19 2.13 0.12 17 Luscinia megarhynchos 0.5 0.03 4 2.25 0.15 18 0.13 0.01 1 Turdus merula 1.25 0.08 10 1.5 0.10 12 0.88 0.08 7 0.88 0.08 7 1.25 0.07 10 Sylvia atricapilla 1.13 0.07 9 2 0.14 16 1.63 0.14 13 1.13 0.10 9 1 0.06 8 Sylvia undata 0.13 0.01 1 Sylvia cantillans 0.75 0.05 6 0.75 0.05 6 Sylvia melanocephala 0.75 0.05 6 Phylloscopus sibilatrix 0.13 0.01 1 Phylloscopus collybita 0.38 0.02 3 0.63 0.04 5 0.25 0.01 2 Regulus ignicapilla 0.13 0.01 1 0.38 0.03 3 0.63 0.03 5 Aegithalos caudatus 0.13 0.01 1 0.13 0.01 1 0.25 0.02 2 0.88 0.05 7 Cyanistes caeruleus 1.25 0.08 10 0.75 0.05 6 1.5 0.13 12 1.63 0.15 13 1.5 0.08 12 Parus major 1.88 0.12 15 1.38 0.09 11 1.75 0.16 14 1.25 0.11 10 1.63 0.09 13 Poecile palustris 0.38 0.02 3 Sitta europaea 0.38 0.03 3 0.25 0.02 2 2.13 0.12 17 Certhia brachydactyla 0.38 0.03 3 0.63 0.03 5 Oriolus oriolus 0.13 0.01 1 Garrulus glandarius 0.63 0.04 5 1.25 0.09 10 0.63 0.06 5 0.88 0.08 7 2.25 0.13 18 Corvus monedula 4.13 0.25 33 0.25 0.02 2 0.13 0.01 1 Corvus cornix 0.25 0.02 2 0.13 0.01 1 0.13 0.01 1 0.13 0.01 1 Fringilla coelebs 1 0.06 8 0.5 0.03 4 0.5 0.04 4 0.13 0.01 1 0.63 0.03 5 Serinus serinus 0.13 0.01 1

Table II. – Breeding bird community parameters in the five forest habitat types. Tot Abm: total mean abundance; S: number of species inside the point count method (a γ-diversity index); S*: total species richness (i.e. number of species totally sampled, also recorded outside the fixed radium and fixed time); Sm: mean number of species (a value of α-diversity); βW: β-diversity index; H’: Shannon- Wiener diversity index. See Methods for further details. See Table I for habitat acronyms Forest habitat types Parameters SCL (n = 130) TER (n = 116) CHE (n = 90) OAK (n = 88) BEE (n = 144) Tot Abm 16.25 (12.22) 14.50 (1.51) 11.25 (2.25) 11.00 (2.39) 18.00 (3.89) S (S*) 19 (28) 18 (26) 16 (19) 15 (19) 18 (22) Sm (α) 8.13 (1.31) 9.75 (1.49) 7.50 (1.51) 7.88 (1.81) 10.88 (1.55)

βW 2.34 1.85 2.13 1.90 1.66 H’ 2.54 2.56 2.39 2.41 2.66

DISCUSSION between altitude and number of species or abundance, dif- ferently from other sites (e.g. Myers & Gill 1988). Birds Probably because of the relatively narrow altitudinal are especially sensitive to habitat heterogeneity induced range analyzed, we did not observe a direct correlation by natural and anthropogenic disturbances (Wiens 1989,

Vie Milieu, 2013, 63 (3/4) PATTERNS IN SPATIAL DIVERSITY OF BREEDING BIRDS 125

Tews et al. 2004). Moreover, community assemblages simply vertically structured at point scale but with a high are strongly related to horizontal and vertical vegetation heterogeneity at the vegetation patch level, typical of the structure (MacArthur & MacArthur 1961, Wiens 1989, habitat types influenced by human activities in Mediter- Malavasi et al. 2009, Battisti & Fanelli 2011). From the ranean landscapes (e.g. Blondel & Aronson 1999). These analysis of our set of breeding bird communities along authors highlighted that these ecosystems were heavily a vegetational-altitudinal gradient, the mesophilic beech disturbed by anthropogenic activities (e.g. fires, grazing, forest is markedly different compared to other habitat cork cultivations, etc.) in the last millennia (i.e. from the types in terms of both species abundance and richness. pre-Roman period). These human-induced processes con- Beech forest hosts significantly more individuals and tributed to produce patchy mosaics at different succes- species than other wood types. Availability of niches sional stages, with open areas alternated to shrubby and and resources are known as the best predictors of species edge habitats, and with isolated trees, increasing the hori- richness and individual abundance in comparable assem- zontal heterogeneity at vegetation patch scale (see also blages (e.g. at the same latitude; Wiens 1989). Therefore Battisti & Fanelli 2011). our data highlight that, along the studied vegetational gra- These evidences suggest that patterns in diversity may dient, the beech forest has a higher species richness and be complex when we consider different hierarchic levels abundance. In birds, these parameters are known to be with different implications. Moreover our data highlight strongly affected by resource and niche availability, espe- that, when conservation and landscape planning strate- cially in terms of vegetation structure and spatial hetero- gies are promoted at landscape scale, objectives should be geneity (Wiens 1989, Ferraguti et al. 2012). carefully specified in terms of which component of spe- One structural variable distinguishing beech for- est from other woodland types is the significantly wider cies diversity (α-, β-, γ-level) constitutes the focus of the diameter of trees. In birds, the tree diameter is an impor- actions. A misleading of concepts related to this topic may tant variable affecting niche and resource availability (e.g. have implications in the interpretation of true patterns and higher plant and dead tree biomass, presence of cavities processes determining biological diversity (see Battisti & and xilophagous and saproxilic invertebrates), as well as Contoli 2011). other environmental parameters (e.g. plant species rich- Acknowledgments – Three anonymous reviewers have ness and vegetation structure), pointed out in the litera- largely improved a first draft of the manuscript. M Ferraguti is ture as important predictors (e.g., MacArthur & MacAr- supported by the FPU Program. thur 1961, Cody 1983). The beech forest shows the low- est value of β-diversity, an index of within-habitat species turnover, indirectly linked to the habitat heterogeneity REFERENCES (Tews et al. 2004). Comparing the γ-diversity (the total number of species) and the β-diversity indexes in each Bani L, Massimino D, Orioli V, Bottoni L, Massa R 2009. habitat, clearly results in higher values in the sclerophilic Assessment of population trends of common breeding birds wood and lower values in the beech forest. in Lombardy, northern Italy, 1992-2007. Ethol Ecol Evol 21: Along our vegetational gradient we also observed 27-44. an interesting inverse pattern from sclerophilic Medi- Battisti C, Contoli L 2011. Diversity indices as ‘magic’ tools in landscape planning: a cautionary note on their uncritical use. terranean forest to the more temperate and mesophilic Landsc Res 36: 111-117. beech forests in terms of spatial diversity (i.e. α-, β- and Battisti C, Fanelli G 2011. Does human-induced heterogeneity γ-diversity). Beech forest is very rich at the point count differently affect diversity in vascular plants and breeding scale (higher α-diversity) but shows the lowest value of birds? Evidences from three Mediterranean forest patches. β-diversity; on the contrary, the sclerophylic wood has a Rend Fis Acc Lincei 22: 25-30. low α-diversity (point-scale) and the highest β-diversity Bibby CJ 2004. Bird diversity survey methods. In Sutherland value. The β-diversity is an expression of the rate of spe- WJ, Newton I, Green RE (eds), Bird Ecology and Conserva- cies turnover among point counts. This rate may be con- tion: A Handbook of Techniques. Oxford University Press, Oxford: 1-11. sidered as an assemblage response to the habitat hetero- geneity at the scale of patch vegetation scale (i.e. the scale Bibby CJ, Burgess ND, Hill DA, Mustoe SH 2000. Bird census techniques. Academic Press, London. of habitat type studied; Koleff et al. 2003). This pattern Blasi C 1994. Fitoclimatologia del . Carta del Fitoclima may indirectly suggest a different habitat heterogeneity del Lazio. Università La Sapienza, Regione Lazio, Roma. of the extreme habitat types along our vegetation gradi- Blondel J, Aronson J 1999. Biology and wildlife of the Mediter- ent, with the beech forest being highly structured (high ranean region. Oxford University Press, Oxford. α-diversity value at point-scale, i.e. high number of mean Blondel J, Ferry C, Frochot B 1970. La méthode des Indices number of species due to higher niche availability) but Ponctuels d’Abondance (I.P.A.) ou des relevés d’avifaune with low habitat heterogeneity at the vegetation patch par stations d’écoute. Alauda 38: 55-71. level (low species turnover expressed by the β-diversity Blondel J, Ferry C, Frochot B 1981. Point counts with unlimited value). On the contrary, the sclerophilic wood is more distance. Stud Avian Biol 6: 414-420.

Vie Milieu, 2013, 63 (3/4) 126 L. ZANGARI, M. FERRAGUTI, C. BATTISTI, L. LUISELLI, M. A. BOLOGNA

Brunelli M, Sarrocco S, Corbi F, Sorace A, Boano A, De Felici Magurran A 2004. Measuring biological diversity. Blackwell S, Guerrieri G, Meschini A, Roma S 2011. Nuovo Atlante Publishing, Malden, MA. degli Uccelli Nidificanti nel Lazio. Edizioni ARP (Agenzia Magurran A, McGill BJ 2011. Biological diversity. Frontiers in Regionale Parchi), Roma. measurements and assessments. Oxford University Press, Calvario E, Sarrocco S 1997. Censimenti delle comunità orniti- Oxford. che degli ambienti a vegetazione erbacea del Lazio (Italia Malavasi R, Battisti C, Carpaneto GM 2009. Seasonal bird centrale) [in Italian]. Alula 4: 87-92. assemblages in a Mediterranean patchy wetland: corroborat- Canterbury GE, Martin TE, Petit DR, Petit LJ, Bradford DF ing the intermediate disturbance hypothesis. Pol J Ecol 57: 2000. Bird communities and habitat as ecological indicators 171-179. of forest condition in regional monitoring. Conserv Biol 14: Manzi A, Perna P 1992. Influenza della vegetazione sulla comu- 544-558. nità di uccelli nidificanti nei pascoli secondari in un’area Cody ML 1983. Continental diversity patterns and convergent dell’Appennino centrale. Alula 1: 90-95. evolution in bird communities. In Kruger FG, Mitchell DT, Margalef R 1958. Information theory in Ecology. Gen Syst 3: Jarvis JUM (eds), Mediterranean-Type Ecosystem. Springer- 36-71. Verlag Berlin: 357-402. Myers AA, Giller PS (Eds) (1988). Analytical Biogeography: an Devoto G, Lombardi G 1997. Le formazioni sedimentarie ed integrated approach to the study of animal and plant distribu- eruttive del settore tolfetano-cerite-manziate (Lazio nord- tions. Chapman and Hall, London and New York. occidentale). Ricerche ecologiche, floristiche e faunistiche nel comprensorio Tolfetano-Cerite-Manziate. Quad Accad Pereira HM, Cooper HD 2006. Towards the global monitoring Naz Lincei 227: 5-31. of biodiversity change. Trends Ecol Evol 21: 123-129. Di Pietro R, Azzella MM, Facioni L 2010. The forest vegetation Romeo P, Azzella M, Facioni L 2010. The Forest Vegetation of of the Tolfa-Ceriti mountains (Northern , Central the Tolfa-Ceriti Mountains (Northern Latium - Central Italy). Italy). Hacquetia 9: 91-150. Hacquetia 9: 91-150. Dornelas M, Soykan CU, Ugland KI 2011. Biodiversity and dis- Shannon CE, Weaver W 1963. The Mathematical Theory of turbance. In Magurran A, McGill BJ (eds), Biological diver- Communication. University of Illinois. Urbana, IL. sity. Frontiers in measurements and assessments. Oxford Soulé ME, Orians GH 2001. Conservation biology research: Its University press, Oxford, New York: 237-251. challenges and contexts. In Soulé ME, Orians GH (eds), Dowgiallo G 1995. Aspetti pedologici dei Monti della Tolfa. Conservation biology. Research Priorities for the Next Geo-Archeologia 1: 27-47. Decade. Society for Conservation Biology, Island Press, Washington DC: 271-285. Fahrig L, Merriam G 1994. Conservation of fragmented popula- tions. Conserv Biol 8: 50-59. Tews J, Brose U, Grimm V, Tielbörger K, Wichmann MC, Schwager M, Jeltsch F 2004. Animal species diversity driven Fanelli G, Bertarelli M, Caroselli V, Cazzagon P, D’Angeli D, by habitat heterogeneity/diversity: the importance of key- De Corso S, De Sanctis M, Gioia P, Serafini Sauli A, Testi A, stone structures. J Biogeogr 31: 79-92. Pignatti S 2007. Carta della vegetazione della Provincia di Roma. Provincia di Roma, Roma. Turcek FJ 1956. Zur Frage der Dominanze in Vogelpopulatio- nen. Waldhygiene 8: 249-257. Ferraguti M, Battisti C, Luiselli L., Zangari L., Bologna MA 2012. Breeding bird assemblages in a Mediterranean mature Vuerich V, Bologna MA, Battisti C 2006. Comunità ornitiche beech forest evidence of an intra-seasonal stability. Rend Fis nidificanti in tre tipologie vegetazionali lungo un gradiente Acc Lincei 24: 1-5. altitudinale nei (Lazio, Appennini centrali) (AVES). Aldrovandia, Riv Mus civ Zool, Roma 2: 49-53. Fazzini P, Gelmini R, Mantovani MP, Pellegrini M 1972. Geolo- gia dei Monti della Tolfa (Lazio Settentrionale Province di Whittaker RH 1960. Vegetation of the Siskiyou Mountains, Ore- Viterbo e di Roma). Mem Soc Geol Ital 21: 65-144. gon and California. Ecol Monogr 30: 279-338. Gibbs JP, Wenny DG 1993. Song output as a population estima- Whittaker RH 1972. Evolution and measurement of species tor: effect of male pairing status. J Field Ornith 64: 316-322. diversity. Taxon 21: 213-251. Gray JS 2000. The measurement of marine species diversity, Wiens JA 1989. The ecology of bird communities. Vols 1, 2. with an application of the benthic fauna of the Norwegian Cambridge studies in ecology, Cambridge University Press, continental shelf. J Exper Mar Biol Ecol 250: 23-49. Cambridge, UK. Koleff P, Lennon JJ, Gaston KJ 2003. Are there latitudinal gra- Wilcove DS, McLellan CH, Dobson AP 1986. Habitat fragmen- dients in species turnover? Glob Ecol Biogeogr 12: 483-498. tation in the temperate zones. In Soulé ME (ed.), Conserva- tion Biology. Sinauer Associates Inc., Sunderland, MA: 237- Lloyd M, Ghelardi RJ 1964. A table for calculating the “Equita- 256. bility” component of species diversity. J Anim Ecol 33: 217- 225. Wilcox BA, Murphy DD 1985. Conservation strategy: the effects of fragmentation on extinction. Am Nat 125: 879-887. Lorenzetti E, Vuerich V, Ukmar E, Battisti C 2007. La comunità ornitica nidificante nella faggeta depressa del SIC “Boschi Mesofili di Allumiere” (Lazio, Italia centrale). Alula 14: 73-77. Received on July 8, 2013 MacArthur RH, MacArthur JW 1961. On bird species diversity. Accepted on February 20, 2014 Ecology 42: 994-998. Associate Editor: J Boissier

Vie Milieu, 2013, 63 (3/4)