Israel Journal of Plant Sciences        Vol.46     1998    pp. 213-221
VEGETATION-ENVIRONMENT RELATIONSHIPS IN THE ULTRAMAFIC AREA
OF MONTE FERRATO, ITALY
Alessandro Chiarucci,* Michele Riccucci, Carlo Celesti, and Vincenzo De Dominicis Dipartimento di Biologia Ambientale, Universita di Siena via P.A. Mattioli 4, Siena 53100, Italy
(Received 18 March 1997 and in revised form 17 November 1997)
ABSTRACT
Relationships between some environmental features and species composition and abundance of grassland and dwarf shrub vegetation were investigated on Monte Ferrato, one of the best known ultramafic (serpentine) sites of Italy. The main aim was to test the importance of the available fraction of soil metals in causing the typical infertility of ultramafic soils. The physical and chemical features of soil were determined for each plot in which species composition and cover were recorded. The plots were classified by cluster analysis and ANOVA was applied to compare the environmental variables of groups of plots. Canonical Correspondence Analysis (CCA) was used to detect the principle factors related to the gradient of species composition within the plant communities. It was found that the grassland and dwarf vegetation of Monte Ferrato is not negatively influenced by soil content of nickel and other metals. Pine canopy cover, which provides additional nutrient input and protects against erosion, was found to be important for evolution of the garigues into grasslands. The evolution of grassland turf induced the retention of higher levels of exchangeable cations, including potentially toxic metals, in the evolved soil.
INTRODUCTION
Because of their physical-chemical features, ultramafic (traditionally referred as "serpentine") areas usually have very distinctive vegetation with respect to adjacent areas. Ultramafic soils are frequently shallow, permeable, and deficient in nutrients, with a high Mg/Ca ratio and relatively high concentrations of potentially toxic metals, such as Ni, Cr, and Co (Proctor and Woodell, 1975; Brooks, 1987; Baker et al., 1992; Roberts and Proctor, 1992). Several ultramafic outcrops can be found in Tuscany, central Italy, and their typical flora and vegetation have been studied (Vergnano Gambi, 1992). The most characteristic vegetation type is a steppe-garigue, described as Armerio-Alyssetum bertolonii, which is characterized by plants endemic to these environments or showing typical serpentino-morphoses and by low ground cover and species richness (Arrigoni et al., 1983; Chiarucci et al., 1995). Several authors explained the typical infertility of the ultramafic soils of Tuscany on the basis of the elevated metal, especially nickel, content of the soil (Sasse, 1979;
Arrigoni et al., 1983; Vergnano Gambi, 1992). Recently, Chiarucci and De Dominicis (1995) and Chiarucci (1996) suggested that the role of metals should be reconsidered with respect to hydrological and nutritional stresses.
The aim of the present paper is to investigate the relationships between the different plant communities and the main soil features, with special reference to the influence of soil metal content on the vegetation.
MATERIALS AND METHODS
SITE DESCRIPTION
Monte Ferrato (lat 43°55'N, long 11 °05'W) is a hilly complex consisting of three hills, located near the town of Prato and constituted by serpentinites and, to a lesser extent, gabbros and basalts (Vergano Gambi, 1992). The elevation of this area is 100 - 440 m a.s.l. Climate is sub-Mediterranean, with an average annual rainfall of * Author to whom correspondence should be addressed. E-mail: chiarucci@unisi.it
© 1998 Laser Pages Publishing Ltd., Jerusalem
214
953 mm and annual temperature of 14.3 °C. July is the driest month, with an average rainfall of 32 mm, and October the wettest, with 132 mm. July is the hottest month, with an average temperature of 24.6 °C, and January the coldest, with 4.7 °C (data from the climatic station of Prato).
Monte Ferrato is one of the most important and well-studied ultramafic sites in Italy. In 1874, during the International Congress of Botany in Florence, an excursion was made to Monte Ferrato (Atti Congresso Inter-nazionale Botanico, 1876). Fiori (1914) and Messeri (1936) have studied its flora and vegetation. In 1934, the Seventh International Phytogeographical Excursion was on Monte Ferrato (Negri et al, 1934). In 1974 there were descriptive papers on bryophytes (Cortini Pedrotti, 1974), flora (Arrigoni, 1974), and vegetation (Corti, 1974) of this site. A third international excursion was made to Monte Ferrato during an Organization for the Phyto-Taxonomic Investigation of the Mediterranean Area (OPTIMA) meeting in Florence (Arrigoni et al., 1979). During this meeting a motion for the safeguarding of this area was passed (Webbia 34: 24-25, 1979). Finally, a phytosociological description of the garigues and grasslands was provided by Arrigoni et al. (1983) and a synthetic study of the garigues of Tuscan ultramafic soils was published by Chiarucci et al. (1995). A revision of the vascular flora is currently underway (C. Ricceri, pers. commun.)
VEGETA TION SAMPLING
The plant communities investigated in the present study were restricted to the grassland and dwarf shrub communities most characteristic of the ultramafic soils of Tuscany. Woody coenoses are, with the exception of artificial pine plantations, very rare and often limited to sites of geological contacts (Messeri, 1936; Pichi Sermolli, 1948).
The vascular plant communities were recorded by means of fifty 2 x 2 m plots. The percentage cover of each species within a plot was carefully estimated. Sample quadrats were placed at random in uniform areas where the vegetation cover was estimated as homogeneous and away from marked discontinuities. The average canopy cover of pines within a radius of 10 m from the plot and number of pine seedlings in each plot were also recorded. Nomenclature is in accordance with Arrigoni (1974) and Chiarucci et al. (1995) for serpentinophytes and with Pignatti (1982) for other species.
SOILS AND ENVIRONMENTAL VARIABLES
Five regularly spaced soil subsamples were collected, at a depth of about 10 cm in each plot. The composite soil samples were air dried and passed
through a 2-mm sieve. A 1:2.5 soil-water suspension was used for a potentiometric measure of pH. Percentage loss-on-ignition (LOI) was measured after heating soil subsamples overnight at 400 °C. Exchangeable cations were extracted by shaking 2-g subsamples of soil, sieved at 125 |nm, for 24 h with 20 ml of 1 M ammonium acetate (adjusted to pH 7) and filtering (Whatman No. 42 paper). Element concentrations were determined by inductively coupled plasma emission spectrometry (Perkin Elmer, Plasma 400).
The following features were also recorded for each plot: aspect, slope (°), total ground cover of vegetation (%), rockiness (%), and stoniness (%). The index of normalized insolation (Bartorelli, 1967) was also calculated for each plot. This index expresses the hours per year equivalent of sunshine normal to a site as a measure of the radiation the site actually receives, given its aspect, slope, and latitude.
STATISTICAL ANALYSIS
The total cover of each species was analyzed. The plots were classified by a homogeneity-optimizing procedure: cluster analysis, using the minimization of sums of squares in new clusters (Anderberg, 1973; Podani, 1994) as the agglomeration algorithm, was applied to a distance matrix constructed on the basis of chord distance, after logarithmic transformation of cover data. Cluster analysis was performed with the program package SYN-TAX 5.0 (Podani, 1993). Soil and environmental data of the groups of plots emerging from cluster analysis were subjected to ANOVA; statistical differences between groups were tested by the Student Newman Keuls test (SNK test) alp < 0.05. To satisfy the normality assumptions of statistical analysis, the variables rockiness, stoniness, LOI, and exchangeable cations were logarithmically transformed.
To detect gradients in species composition and species-environment relationships, a canonical correspondence analysis (CCA) ordination was performed using the program package CANOCO 3.11 (ter Braak, 1988). Cover data were log-transformed and only species with a frequency higher than 2 were included. In CCA, rare species were downweighted, and to test the significance of the eigenvalue corresponding to the first CCA canonical axis, a Monte Carlo permutation test was applied (ter Braak, 1988).
RESULTS
VEGETATION
The main clusters (Fig. 1) divided the plots dominated by Brachypodium rupestre (Group 4) from the others. Then the steppe-garigues (Group 1) were sepa-
Israel Journal of Plant Sciences 46 1998
215
Fig. 1. Dendrogram showing the classification of the 50 vegetation plots.
rated from the xeric grasslands dominated by Bromus erectus (Group 2) or Danthonia alpina (Group 3). The steppe-garigues of Group 1 belong to the Armerio-Alyssetum bertoloni subass. typicum, originally described for Monte Ferrato (Arrigoni et al., 1983), common in all the ultramafic areas of inland Tuscany (Chiarucci et al., 1995), and are characterized by several species endemic to Tuscan serpentine soils (Table 1).
The grasslands dominated by Bromus erectus (Group 2) belong to the alliance Bromion erecti and follow the garigues in the vegetation dynamics of Monte Ferrato ultramafic soils (Arrigoni et al., 1983). They are characterized by an abundance of graminoid species, e.g., B. erectus and Festuca robustifolia, but also by a high frequency of garigue species, e.g., Alyssum bertolonii, Centaurea aplolepa ssp. carueliana, Thymus acicularis var. ophioliticus, and chamaephytes, e.g., Cistus salvifolius, Helichrysum italicum. The grasslands of Group 3 were dominated by Danthonia alpina. B. erectus was also abundant and F. robustifolia was even more abundant than in Group 2. In their study on garigue and grassland vegetation of Monte Ferrato, Arrigoni et al. (1983) did not record plant communities dominated by D. alpina and did not find this species. Messeri (1936) quoted its presence in the plant collection of the Florentine botanist Levier. Danthonia alpina is, presently, very common on Monte Ferrato and dominates this vegetation type.
The grassland communities of Group 4 were dominated by Brachypodium rupestre. Messeri (1936) reported that these grasslands were rare on Monte Ferrato and did not grow under pine cover as did "xeric Festuca grasslands" (which probably correspond to our Bromus
grasslands, in which F. inops and F. robustifolia are abundant). Arrigoni et al. (1983) reported B. rupestre grasslands under a constant but low pine cover: 1-2 of the Braun-Blanquet (1932) scale. B. rupestre may have spread in recent times owing to natural or anthropogenic causes.
SPECIES DIVERSITY AND GROUND COVER
A total of 79 vascular plants were found in the 50 plots; average species number per plot was 16.6. The steppe-garigues, characterized by low ground cover and a relatively low species richness (Pichi Sermolli, 1948; Chiarucci et al., 1995), with respect to similar plant communities on other soils, showed the highest species richness in the present survey (Table 2). Species richness of Groups 2 and 3, which was similar to that of Group 1, was statistically higher than in the Brachypodium rupestre grasslands (Group 4). Ground cover was significantly higher in Group 4 than Group 1; the cover of B. erectus grasslands (Group 2) was statistically lower than that of Brachypodium rupestre grasslands (Group 4), whereas that of Danthonia alpina grasslands (Group 3) was intermediate (Table 2).
Pine seedlings established preferentially in the xeric grasslands, especially those dominated by Danthonia alpina. They did not colonize Brachypodium rupestre grasslands (Table 3), probably being prevented by the dense turf.
SOIL AND PHYSICAL ENVIRONMENT
The main soil features are reported in Table 4. Organic matter content and availability of exchangeable cations increased from garigues to xeric grassland and Brachypodium rupestre grassland vegetation types, whereas pH, rockiness, and stoniness decreased. Chromium, cobalt, and copper, present in high total concentrations in ultramafic soils (Brooks, 1987), were at the lower limits of detection (0.5 jxg g~') in the soil extracts and are not reported; however, an increasing pattern similar to the other exchangeable cations was observed. The index of normalized insolation did not show any significant differences between the four groups. The Brachypodium rupestre grasslands differed in that they were located on steeper slopes and under denser pine cover. The xeric grasslands had intermediate pine cover, whereas garigues were found in places with the lowest pine cover. The magnesiumxalcium ratio, which has been considered an important limiting factor for some serpentine flora (Brooks, 1987; Proctor and Nagy, 1992), was significantly higher under the garrigue communities.
The greater availability of metals in the soil under the more mature and structured grassland vegetation types
Chiarucci et al. / Vegetation-environment relationships, Monte Ferrato
216
Table 1
Species frequency (F) according to a five-point scale (1, <20%; II, 20^10%; III, 40-60%; IV, 60-80%, V, 80-100%) and mean cover (C) in the four groups of plots obtained by cluster analysis. +, mean cover less than 0.01
Group of plots 12 3 4
	F	c	F	c	F	c
Poa bulbosa	II	0.19				
Micropus erectus	11	0.01				
Tuberaria guttata	I	0.01				
Cheilanthes marante	I	0.03				
Jasione montana	I	+				
Cynodon dactylon	I	0.03				
Cerastium ligusticum	V	0.13	II	0.03		
Sedum rupestre	IV	0.78	I	0.02		
Hernaria glabra	IV	0.23	I	+		
Anthyllis vulneraria ssp. rubriflora	III	0.17	I	+		
Aira elegans	III	0.02	II	0.03		
Leopoldia comosa	II	0.01	I	+		
Hypochoeris achyrophorus	II	0.04	I	0.01		
Dianthus sylvestris ssp. longicaulis	II	0.07	I	0.01		
Filago germanica	II	0.06	II	0.05		
Vulpia ciliata	I	0.03	I	+		
Echium vulgare	I	+	I	0.01		
Centaurum pulchellum	I	0.02	I	0.01		
Asterolinun linum-stellatum	I	+	I	+		
Artemisia alba	I	0.72	I	3.57		
Briza sp.	I	0.01	I	0.01		
Stipa etrusca	I	0.17	II	1.30		
Euphorbia exigua	1	+	II	0.08		
Armeria denticulala	IV	0.58			I	0.02
Minuartia laricifolia ssp. ophiolitica	I	0.06			I	+
Brachypodium dystachyum	I	0.16			I	0.01
Einum tenuifolium	I	0.01			I	+
Alyssum bertolonii	IV	0.20	II	0.08	I	0.09
Potentilla hirta	IV	1.12	III	0.32	II	0.16
Timus acicularis var. ophioliticus	V	3.81	V	1.82	III	0.80
Festuca inops	V	7.00	V	3.50	III	0.75
Linum trigynum	V	0.28	V	0.19	IV	0.05
Centaurea aplolepa ssp. carueliana	IV	1.29	IV	1.59	III	0.55
Hieracium pilosella	III	0.15	IV	0.13	III	0.03
Cuscuta epithymum	I	+	II	0.04	I	+
Hieracium piloselloides	III	0.13	IV	0.36	IV	0.19
Trinia glauca	II	0.05	I	0.08	IV	0.73
Sonchus asper			I	+		
Koeleria splendens			I	0.13		
Erica scoparia			I	0.04		
Silene italica					I	0.04
Rubus ulmifolius					I	0.01
Pseudolysimachion barrelieri ssp. nitens					I	0.01
Poligala flavescens					I	0.04
Plantago lanceolata					I	0.01
Lactuca serriola					I	+
Genista pilosa					I	0.23
Aster linosyris					II	0.33
Asperula cynanchica					I	+
Agrostis tenuis					I	+
Allium sphaerocephalon	IV	0.03	IV	0.03	III	0.01
Israel Journal of Plant Sciences 46 1998
217
Table 1 continued
Group of plots	f	1 c	F	2 c	F	3 c		4 "f "~č
Reicardia picroides	iv	0.22	iv	0.25	ii	0.03	i	0.04
Bromus erectus	iii	0.69	v	19.30	v	8.18	i	0.21
Helichrysum italicum	ii	0.42	iv	1.01	iii	0.93	i	0.04
Festuca robustifolia	iv	7.72	iv	8.25	v	22.7	iii	0.81
Centaurea rupestris	ii	0.33	iii	1.48	iv	2.35	iii	1.06
Genista januensis	ii	0.59	iii	0.34	v	0.45	iii	0.07
Cistus salvifolius	ii	0.31	iv	1.90	i	0.39	iii	0.29
Sanguisorba minor ssp. muricata	i	0.17	ii	0.20	iii	0.27	i	0.01
Galium corrudifolium	ii	0.14	iv	1.14	v	0.74	v	0.53
Dantonia alpina	i	+	i	0.46	v	40.1	ii	0.86
Vincetoxicum hirundinaria	i	+	II	0.17	iv	0.09	iii	0.24
Carlina corymbosa	i	0.01	iii	0.32	iii	0.07	iii	0.20
Silene paradoxa	i	0.01	i	0.01			ii	0.09
Stachys recta ssp. serpentinu	i	0.01	i	0.02			i	0.01
Anthericum liliago	i	0.01					i	0.01
Euphorbia nicaensis var. prostrata	i	0.01			iii	0.20	ii	0.07
Brachipodium rupestre	1	0.17			i	0.05	v	81.40
Tanacetum corymbosum			i	0.06			iii	0.30
Inula viscosa			i	+	I	0.01	I	0.14
Picris hieracioides			i	+	I	0.08	I	+
Scorzonera austriaca			i	0.03	iii	0.07	I	0.09
Hippocrepis multisiliquosa			i	+	iii	0.09	II	0.08
Lotus corniculatus					ii	0.10	ii	0.06
Knautia arvensis					i	0.02	iii	0.19
Dorycnium hirsutum					i	0.09	i	0.06
Carex humilis					ii	1.03	iii	0.79
Filipendula vulgaris							i	0.57
Table 2
Species richness and ground cover (mean ± S.E.) in the four groups of plots obtained by cluster analysis. Different letters indicate statistically significant differences
between groups
Group of plots	1	2	3	4
n	18	14	11	7
Species richness	18.8 ±0.9 a	17.0 ±0.8 a	17.3 ± 1.8 a	9.4 ± 1.2 b
Ground cover	31.7 + 4.4 a	50.9 ± 6.6 b	61.8 ±7.4 be	76.4 ± 8.5 c
Table 3
Number of pine seedlings (mean ± S.E. calculated from logarithmically transformed data) in the four groups of plots obtained by cluster analysis. Different letters indicate statistically significant differences between groups
Group of plots	1	2	3	4
n	18	14	11	7
Pinus seedlings	2.8 ab	5.7 a	8.3 a	0.1 b
	(2.0-4.0)	(4.1-7.8)	(5.2-12.8)	(0.4-1.6)
Chiarucci et al. / Vegetation-Environment relationships, Monte Ferrato
218
Table 4
Soil and physical environment variables (mean ± S.E.) in the four groups of plots obtained by cluster analysis. Different letters indicate statistically significant differences between groups
Group of plots	1	2	3	4
n	18	14	11	7
Normalized insolation	2030 ± 57	2242 ± 50	2085 ± 84	2240±193
Slope	13.6+ 1.2 a	13.3+1.3 a	15.5 ±2.0 a	25.4 ±2.4 b
Rockiness	6.6a	2.8b	1.4b	1.1b
	(5.5-7.8)	(2.0-3.8)	(0.9-2.1)	(0.5-2.2)
Stoniness	53.1a	16.1b	1.5c	0.4c
	(47.7-59.0)	(11.6-22.1)	(1.0 -2.2)	(0.2-0.7)
LOI	10.8a	13.5b	18.8c	23.7c
	(10.1-11.5)	(12.6-14.3)	(17.6-19.9)	(22.5-24.9)
pH	6.93 ± 0.04 a	6.68 + 0.05 b	6.55 ± 0.08 b	6.18 ±0.09 c
Pine cover	9.2 ±2.0 a	21.6 ± 3.1 b	24.5 ±2.7 b	42.9 ±4.7 c
Mg/Ca	3.14 + 0.35	2.29 + 0.15	2.69 ±0.21	2.17 ±2.01
Ca	752 a	1077 b	1172 b	1318b
	(706-801)	(1031-1126)	(1104-1244)	(1232-1410)
Mg	2165 a	2408 ab	3048 b	2814 b
	(2037-2301)	(2259-2566)	(2832-3279)	(2631-3009)
K	65.4 a	96.6 b	94.2 b	95.0 b
	(61.2-69.8)	(89.3-104.4)	(85.4-103 .8)	(87.3-103.4)
Na	33.4 a	37.7 a	41.1 a	56.2 b
	(31.0-36.0)	(36.2-39.3)	(38.4-44.0)	(50.6-62.4)
Mn	6.1 a	19.6 b	21.9b	42.1 c
	(5.4-6.9)	(16.6-23.9)	(17.3-27.6)	(29.4-60.3)
Fe	0.4 a	0.5 a	0.6 a	1.3 b
	(0.3-0.5)	(0.4-0.6)	(0.5-0.6)	(1.0-1.6)
Zn	0.8 a	0.9 a	1.0 a	1.9 b
	(0.7 -0.9)	(0.8-1.1)	(0.9-1.1)	(1.5-2.4)
Ni	7.3 a	15.5 b	23.1 c	25.6 c
	(6.7-10.0)	(13.7-19.5)	(20.2-28.3)	(22.6-31.0)
Normalized insolation is expressed in hours per year equivalent of sunshine normal to a site; slope in degrees; rockiness, stoniness, loss on ignition (LOI), and pine cover in percentage; and the concentration of exchangeable cations in ug g-'. Values in parentheses represent S.E. range after log-transformed data.
may be due to the lower pH. Robinson et al. (1996), for example, found that metal solubility in New Zealand serpentines increased exponentially with decreasing pH of the extracting solution.
GRADIENT ANAL YSES
The CCA ordination of plots and environmental variables is shown in Fig. 2. The eigenvalues of axes 1 and 2 are 0.526 and 0.268, respectively. The cumulative percentage of variance explained by the first two axes accounts for 27.4% (18.2% and 9.2%) of species data and 56.5% (37.4% and 19.1%) of species-environment relations; A Monte Carlo permutation test showed that the first axis significantly contributed to the explained variance (p < 0.01). A clear trend from the Armerio-Alyssetum bertolonii garigues to the Bromus erectus grasslands, the Danthonia alpina grasslands, and then to the Brachypodium rupestre grasslands was observed
from the negative to the positive side of the first axis. This axis was positively correlated with slope, pine cover, organic matter (LOI), nickel, manganese, calcium, sodium, and negatively with stoniness and pH (Table 5).
As to the behavior of the different species (Fig. 3), Herniaria glabra, Cerastium ligusticum, Sedum rupestre, Potentilla hirta, Linum trigynum, Festuca inops, and Timus acicularis var. ophioliticus (typical of the garigues and the less evolved grassland stages) are related to high rockiness, stoniness, pH, and Mg/Ca ratio. The most typical grassland species, such as Bromus erectus, Danthonia aplina, and Carex humilis, tend to locate in sites with the highest content of soil exchangeable cations and with elevated organic matter and pine canopy cover. Brachypodium rupestre is related to the highest slopes, also in sites with elevated pine canopy cover.
Israel Journal of Plant Sciences 46 1998
219
Stoniness
Organic matter
Fig. 2. CCA ordination of the environmental variables analyzed and the 50 plots. Asterisks: group 1; circles: group 2; squares: group 3; diamond: group 4.
o i—i
+
Stoniness
Rockiness        Her gl»
Fei ino 4 +„
-1.0
Slope
4Brarup
Dan alp
Organic matter
Fig. 3. CCA ordination of the environmental variables analyzed and the plant species with a fit higher than 25. Her gla: Herniaria glabra; Cer lig: Cerastium ligusticum; Sed rup: Sedum rupestre; Pot hir: Potentilla hirta; Lin tri: Linum trigynum; Fes ino: Festuca inops; Thy oph: Thymus acycularis ssp. ophioliticus; Bro ere: Bromus erectus; Dan alp: Danthonia alpina; Car hum: carex humilis; Bra rup: Brachypodium rupestre.
Chiarucci et al. / Vegetation-environment relationships, Monte Ferrato
220
Table 5
Correlations of environmental variables with the first two axes of CCA ordination
Environmental variable	Axis 1	Axis 2
Slope	0.5527	0.3037
Normalized insolation	0.0999	0.0884
Rockiness	-0.4261	0.3441
Stoniness	-0.7741	0.3619
Pine cover	0.7107	-0.0755
LOl	0.7567	-0.2567
pH	-0.7247	0.1272
Zn	0.4535	0.0696
Ni	0.6242	-0.4666
Fe	0.3788	0.1876
Mn	0.5568	-0.2193
Mg	0.3504	-0.2290
Ca	0.5545	-0.3965
Na	0.5006	0.0335
K	0.2952	-0.3681
Ca/Mg	-0.1800	0.1540
DISCUSSION
The present survey enabled four main types of grassland and dwarf shrub vegetation to be distinguished in Monte Ferrato, one of which, the Danthonia alpina grasslands, has never previously been reported. As found in other recent studies of Tuscan ultramafic soils (Angelone et al., 1991, 1993) and vegetation (Chiarucci and De Dominicis, 1995; Chiarucci, 1998a,b), the influence of soil metal content on plant communities is not as great as other ecological factors, such as water and nutritional stress. Moreover, the chemical features found to be correlated with the vegetation pattern should not be considered the cause, but partially the effect of vegetation dynamics, which strongly influence soil development. In this sense, soil and vegetation processes and dynamics influence each other by positive feedback.
Pine cover was found to be one of the most important factors influencing vegetation. As found in another ultramafic site, pine cover can improve plant diversity and cover, probably by protecting the soil from splash erosion and providing an additional nutrient input (Chiarucci and De Dominicis, 1995; Chiarucci, 1996). This input may be related to pine cover in a direct (needles, branches, and bark fall) and indirect (higher interception of wet and dry deposition from atmosphere) ways. In the Monte Ferrato area, pines were introduced at the beginning of this century and reproduced readily (Messeri, 1936) especially in the xeric grasslands, which are probably one of the most natural vegetation types in this area.
Another important factor influencing the present vegetation of Monte Ferrato may be the increased atmospheric nitrogen input due to pollution from the cities of Florence, Prato, and Pistoia. Increased atmospheric nitrogen input due to pollution has been found to promote dominance of Brachypodium pinnatum and reduce species diversity in oligotrophic vegetation types, such as chalk grasslands (Bobbink and Willems, 1987; Bobbink, 1991; Willems et al., 1993). Brachypodium rupestre, which is vicariant of B. pinnatum in almost all of Italy, was once rare in the Monte Ferrato area (Messeri, 1936) and may have spread in recent times because of eutrophication due to the pine canopy and increased atmospheric nitrogen input.
The conservation of the ultramafic vegetation of Monte Ferrato, with its typical barren appearance and its richness in Tuscan serpentine endemics, is probably endangered by the spread of the pines and by pollution. Both these factors should be monitored and possibly reduced. The reduction of pollution is generally desired but difficult to achieve quickly. However a reduction in the spread of pines is a practical possibility at least in some parts of the area. Without a control, and possibly a reduction of pines, their presence could increase since they colonize xeric grasslands well, and vegetation structure and composition of this site of great ecological interest are likely to be modified.
ACKNOWLEDGMENTS
The authors thank Dr. Bruno Foggi and Dr. Simona Maccherini for help in field sampling, Dr. Ermanno Busoni for information about the index of normalized insolation, and Dr. Claudia Angiolini for help during data elaboration. This work was funded by the National Council of Research of Italy (CNR), contribution no. 96.032111.CT04.
REFERENCES
Anderberg, M.R. 1973. Cluster analysis for applications. Academic Press, New York.
Angelone, M., Vaselli, O., Bini, C, and Coradossi. N. 1993. Pedogeochemical evolution and trace element availability to plants in ophiolitic soils. Sci. Total Environ. 129: 291-309.
Angelone, M., Vaselli, O., Bini, C, Coradossi, N., and Pancani, M.G. 1991. Total and EDTA-extractable element contents in ophiolitic soils from Tuscany (Italy). Z. Pflanzenernaehr. Bodenkd. 154: 217-223.
Arrigoni, P.V. 1974. La flora del Monte Ferrato. Atti Soc. Toscana Sci. Nat. Pisa, Mem., P. V. Ser. B 81: 1-10.
Arrigoni, P.V., Ricceri, C, and Mazzanti, A. 1983. La vegetazione serpentinicola del Monte Ferrato di Prato in Toscana. Centra di Scienze Naturali, Prato, Race. Scr. Bot.
Israel Journal of Plant Sciences 46
1998
221
28 pp.
Arrigoni, P.V., Ricceri, C, and Vergnano Gambi, O. 1979. Excursion to Monte Ferrato (25 May 1977). Webbia 34: 27-31.
Atti Congresso Internazionale Botanico. 1876. Gita a
Monteferrato, 17.5.1874. Regia Societě Toscana di
Orticoltura, Firenze, pp. 357-358. Baker, A.J.M., Proctor, J., and Reeves, R.D., eds. 1992. The
vegetation of ultramafic (serpentine) soils. Proc. 1st Int.
Conf. Serpentine Ecol. Intercept, Andover, UK. Bartorelli, U. 1967. Tavole numeriche dell'assolazione annua.
Annali Accad. Ital. Sci. For. 16: 61-83. Bobbink, R. 1991. Effects of nutrient enrichment in Dutch
chalk grassland. J. Appl. Ecol. 65: 759-791. Bobbink, R. and Willems, J.H. 1987. Increasing dominance of
Brachypodium pinnatum (L.) Beauv. in chalk grasslands: a
threat to a species-rich ecosystem. Biol. Conserv. 40: 301-
314.
Braun-Blanquet, J. 1932. Plant sociology. The study of plant communities. Hafner Publishing, New York.
Brooks, R.R. 1987. Serpentine and its vegetation. A multidisciplinary approach. Croom Helm, London.
Chiarucci. A. 1996. Species diversity in ultramafic plant communities in relation to pine afforestation. J. Veg. Sci. 7: 57-62.
Chiarucci, A. and De Dominicis, V. 1995. Effects of pine
plantations on ultramafic vegetation of central Italy. Isr. J.
Plant Sci. 43: 7-20 Chiarucci, A., Foggi, B., and Selvi, F. 1995. Garigue plant
communities of ultramafic outcrops of Tuscany (Italy).
Webbia 49: 179-192. Chiarucci, A., Maccherini, S., Bonini, I., and De Dominicis,
V. 1998a. Effects of nutrient addition on species diversity
and cover on "serpentine" vegetation. Plant Biosystems
132, in press.
Chiarucci, A., Robinson, B.H., Bonini, I., Petit, D., Brooks,
R.R., and De Dominicis, V. 1998b. Vegetation of Tuscan
ultramafic soils in relation to edaphic and physical factors.
Folia Geobot. 33: 113-132 Corti, R. 1974. Caratteristiche generali della vegetazione del
Monteferrato (Prato). Atti Soc. Toscana Sci. Nat. Pisa
Mem., P. V. Ser. B. 81: 32-38. Cortini Pedrotti, C. 1974. La vegetazione pioniera del Monte
Ferrato (Prato). Atti Soc. Toscana Sci. Nat. Pisa Mem., P.
V. Ser. B. 81: 39^4. Fiori, A. 1914. La flora dei serpentini della Toscana. II.
Confronto fra la Flora del M. Ferrato (serpentino) e quella
della Calvana (calcare alberese). Nuovo Giorn. Bot. Ital.
n.s. 21: 216-240.
Messeri, A. 1936. Ricerche sulla vegetazione dei dintorni di Firenze. IV. La vegetazione delle rocce ofiolitiche del Monte Ferrato (presso Prato). Nuovo Giorn. Bot. Ital. n.s. 43: 277-372.
Negri, G., Tongiorgi. E., Corti, R., Pampanini, R., and Chiarugi, A. 1934. Septieme Excursion Phytogeo-graphique Int. Ital. 1934. Guide Itineraire. Tip. M. Ricci, Firenze, 55 pp.
Pichi Sermolli, R. 1948. Flora e vegetazione delle serpentine e delle altre ofioliti dell'Alta valle del Tevere (Toscana). Webbia 6: 1-380.
Pignatti, S. 1982. Flora d'ltalia. Edagricole, Bologna. Vol. 3.
Podani, J. 1993. SYN-TAX version 5.0. User's guide. Scientia Publishing, Budapest.
Podani, J. 1994. Multivariate data analysis in ecology and systematics. A methodological guide to the SYN-TAX 5.0 package. SPB Academic Publishing, The Hague.
Proctor, J. and Nagy, L. 1992. Ultramafic rocks and their vegetation: an overview. In: Baker, A.J.M., Proctor, I., and Reeves, R.D., eds. The vegetation of ultramafic (serpentine) soils. Proc. 1st Int. Conf. Serpentine Ecol. Intercept, Andover, UK, pp. 469^194.
Proctor, J. and Woodell, S.R.J. 1975. The ecology of serpentine soils. Adv. Ecol. Res. 9: 255-366.
Roberts, B.A. and Proctor, J., eds. 1992. The ecology of areas with serpentinized rocks: a world view. Kluwer, Dordrecht.
Robinson, B.H., Brooks, R.R., Kirkman, J.H., Gregg, P.E.H., and Gremigni, P. 1996. Plant-available elements in soils and their influence on the vegetation over ultramafic "serpentine" rocks in New Zealand. J. R. Soc. N. Z. 26: 455^166.
Sasse, F. 1979. Untersuchungen an Serpentinstandorten in Frankreich, Italien, Oesterreich und der Bundesrepublik Deutschland. I. Bodenanalysen. Flora 168: 379-395.
ter Braak, C.J.F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179.
ter Braak, C.J.F. 1988. CANOCO. A fortran program for canonical community ordination by [partial] [detrended] [canonical] correspondence analysis, principal component analysis and redundancy analysis (Version 2.1). Upgrade to version 3.1. Scientia Publishing, Budapest.
Vergnano Gambi, O. 1992. The distribution and ecology of the vegetation of ultramafic soils of Italy. In: Roberts, B.A. and Proctor J., eds. The ecology of areas with serpentinized rocks: A world view. Kluwer, Dordrecht, pp. 217-247.
Willems, J.H., Peet, R.K., and Bik, L. 1993. Changes in chalk-grassland structure and species richness resulting from selective nutrient additions. J. Veg. Sci. 4: 203-212.
Chiarucci et al. / Vegetation-environment relationships, Monte Ferrato