Is vascular plant species diversity a predictor of bryophyte species diversity in Mediterranean forests?

Is vascular plant species diversity a predictor

of bryophyte species diversity in Mediterranean

forests?

ALESSANDRO CHIARUCCI*, FRANCESCA D’AURIA andILARIA BONINIDipartimento di Scienze Ambientali ‘‘G. Sarfatti’’, Universita Di Siena, Via P.A. Mattioli 4, Siena

53100, Italy; *Author for correspondence (e-mail: [email protected]; fax: +39-0577-232896)

Received 2 May 2005; accepted in revised form 5 January 2006

Key words: Bryophytes, Conservation, Reservation sets, Reserve site selection, Taxon congruence,

Taxon surrogacy, Vascular plants, Woody plants

Abstract. This study aimed to (i) investigate the congruence among the species composition and

diversity of bryophytes and vascular plants in forests; (ii) test if site prioritization for conservation

aims by the maximization of the pooled number of vascular plant species is effective to maximize

the pooled number of bryophyte species. The study was performed in six forests in Tuscany, Italy.

Four-hundred and twenty vascular plant species (61 of which were woody) and 128 bryophyte

species were recorded in 109 plots. Despite the good predictive value of the compositional patterns

of both woody plants and total vascular with respect to the compositional pattern of bryophytes,

the species richness of the latter was only marginally related to the species richness of the former

two. Bryophyte rare species were not spatially related to rare plant species and neither coincided

with the sites of highest plant species richness. The species accumulation curves of bryophytes

behaved differently with respect to those of woody plants or total vascular plants. Reserve selection

analysis based on the maximization of the pooled species richness of either woody plants or total

vascular plants were not effective in maximizing the pooled species richness of bryophytes. This

study indicates that species diversity of vascular plants is not likely to be a good indicator of the

bryophyte species diversity in Mediterranean forests.

Introduction

Forest ecosystems in the Mediterranean area are floristically extremely diver-sified. Tuscany, central Italy, hosts a wide variety of forest ecosystems, fromthe coastal Mediterranean ‘‘macchia’’, dominated by sclerophyllous evergreenspecies, to the high-mountain forests (up to 1600–1800 m) dominated by beech(Bernetti 1987). Forests cover more than half of the region’s surface, an areagreater than one million hectares. About 120,000 ha of forest belongs to theRegional Administration of Tuscany and other 100,000 ha to other publicadministrations. In Tuscany, forests have always been considered a funda-mental resource, especially for timber production and soil protection. How-ever, consideration of forest ecosystems as environmental and landscaperesources has developed in recent years. Tuscany was one of the first regions ofItaly to establish a monitoring program of forests, within the ‘‘Scheme for the

Biodiversity and Conservation (2007) 16:525–545 � Springer 2006

DOI 10.1007/s10531-006-6733-1

protection of forests against atmospheric pollution’’ (EU Regulation 3528/86and followings). Originally the monitoring network aimed to collect informa-tion exclusively on tree health, but carbon stock and biodiversity were laterincluded in the program (Bartolozzi et al. 2002). Presently, these forests areundergoing continuous monitoring, in which biodiversity is a fundamental partof the program. During the first stage, biodiversity monitoring was based onsix intensive permanent plots (Chiarucci et al. 2001), but it was then decided tocollect floristic information representative of larger areas, i.e. whole forests(Chiarucci and Bonini 2005).In the Mediterranean area, bryophytes are one of the least investigated

components of forest ecosystems and almost no quantitative survey exists forTuscany. Although many papers have investigated the species composition ofspecific sites and/or habitats, to date there have not been any attempts toidentify quantitative patterns of bryophyte diversity in forests. One method ofestimating species diversity of less studied groups is by assuming ‘‘taxon sur-rogacy’’ (Ryti 1992; Garson et al. 2002), i.e. estimating the species richness of agiven taxon based on the species richness of another taxon. Taxon surrogacy isbased on the assumption of cross-taxon congruence in spatial patterns ofspecies richness (Prendergast et al. 1993; Howard et al. 1998). Taxon surrogacyis interesting from both the evolutionary and ecological points of view and itmight be used to assess spatial and temporal patterns of biodiversity for tax-onomic groups which are difficult and or expensive to sample (Williams 1999;Virolainen et al. 2000).Vascular plants are an attractive group to use as surrogates for estimating the

diversity of other groups as they constitute the bulk of the primary producers,reflect environmental conditions, provide physical habitat for other organisms,and are relatively easy to sample (Ryti 1992; Pharo et al. 2000). Vascular plantshave been reported as good surrogates for identifying reserve areas for inverte-brates (e.g. Panzer and Schwartz 1998) and birds (e.g. Tardif and DesGranges1998). Only a few papers have discussed the use of vascular plants as surrogatetaxon to evaluate the species diversity of bryophytes at different spatial scales inAustralian ecosystems (Pharo et al. 1999; Pharo et al. 2000).This study aimed to: (i) investigate the congruence among the species com-

position and diversity of bryophytes and vascular plants in forest ecosystems atdifferent scales; (ii) test if site prioritization based on maximizing the poolednumber of vascular plant species is effective in maximizing the pooled numberof bryophyte species.

Methods

Sampling sites

Six forest estates (hereafter ‘‘forests’’) owned and managed by the regionaladministration of Tuscany were chosen as study areas (Figure 1). Each forest is

526

formed by contiguous or non-contiguous woodlands located in neighbouringareas and managed as a single unit. These forests cover a total of 37,240 ha,and range in size from 2098 to 10,311 ha (see Table 1 for summary); they arelocated from the lowlands of the coastline to the Apennine mountains and hostdifferent plant communities, ranging from the evergreen Mediterranean forestsdominated by Quercus ilex, along the coastline and the lower hillsides, to theFagus sylvatica and Abies alba forests of mountain sites. Conifer plantationsare present in all the forests.A stratified random sample of sites was selected from the sites used for the

Forest Inventory of Tuscany (IFT). The number of sampling sites was selectedin accordance with an existing network established for the monitoring of treecrown conditions (111 sites), with the number proportional to the true forestsurface of each area, i.e. excluding open patches therein. The IFT was based onpoints placed on a 400 · 400 m regular grid used for photo interpretation offorest types; a subset of them was used for measurement of tree compositionand structure. Two of the selected sites were not sampled since they were justburned.

Madonna delle Querce

Colline Livornesi

Macchia dellaMagona

Foreste Casentinesi

Farma Merse

Bandite diScarlino

Foreste Pistoiesi

Belagaio

Montioni

#Siena

##

Pisa

Firenze

0 50 km

Figure 1. The locations of the six investigated forests owned and managed by the regional

administration of Tuscany.

527

Table

1.

Locationandbasicfeaturesofthesixinvestigatedforests.

Forest

Elevation

(weightedmean

andrange)

Surface

(ha)

Stem

density

a

(stem

ha�1)

Basalareab

(m2ha�1)

Dominanttree

speciesc

BanditediFollonica,Macchia

della

Magona,Montoni(BF)

180

(24–416)

9865

271.6

(224.8–328.0)

1.3±

0.3

Fraxinusornus,Quercusilex,Quercuspubescens,

Arbutusunedo,Quercuscerris.

CollineLivornesi(C

L)

204

(40–594)

2098

120.8

(76.4–191.1)

1.4±

0.3

Quercusilex,Pinushalepensis,Pinuspinea

Farm

a-M

erse

eBelagaio

(FM)

351

(154–572)

7294

74.0

(50.4–108.7)

0.8±

0.3

Fraxinusornus,Arbutusunedo,

Pinuspinaster,Quercuscerris,Castanea

sativa

Madonnadelle

Querce

(MQ)

544

(205–1077)

2205

105.9

(72.6–154.4)

1.1±

0.3

Quercuspubescens,Quercuscerris,Quercusilex

ForesteCasentinesi(FC)

1091

(585–1656)

5468

37.8

(29.4–48.6)

1.7±

0.3

Fagussylvatica,Abiesalba,Quercuscerris,Pinusnigra

ForestePistoiesi(FP)

1188

(631–1891)

10311

63.2

(49.1–81.3)

1.8±

0.3

Fagussylvatica,Abiesalba

aThevalues

reported

are

means±

90%

confidence

intervalscalculatedafter

log-transform

ationto

fitnorm

ality.

bThevalues

reported

are

means±

90%

confidence

intervals.

cDefined

asthose

specieswithafrequency

‡25%

oftheplots

andatleast

5%

oftotalbasalarea,in

order

ofdecreasingfrequency.

528

Data collection

In each site, once located with a high precision GPS, a 20 · 20 m plot wasdelimited. The plot was divided into four 10 · 10 m quadrants. The followingdata were collected for vascular plants: (i) DBH of plants for individuals withDBH ‡ 3 cm and (ii) total list of species within the quadrant/plot.In each plot, bryophytes were sampled in 24 bryoplots whose positions were

determined using four trees with DBH ‡ 12 cm, selected by a stratified randomdesign within each plot (one in each quadrant). On each tree, two bryoplotswere positioned on the trunk at 130 cm of height and two at the trunk base.For each position (130 cm and trunk base), one bryoplot was positioned on thenorthern face and one on the southern face of the trunk. Two additionalbryoplots were located on the ground at 50 cm from the trunk base, one on thenorthern side and one on the southern site. These ground bryoplots were usedto assess the bryophyte component on the ground as influenced by the treepresence. Each bryoplot was 10 · 20 cm. All the mosses and liverworts presentwithin each bryoplot were identified at the species level directly in the field orlater in the laboratory. Species richness at the plot scale is here defined as thepooled number of bryophyte species recorded in the 24 bryoplots within eachplot.In order to prepare the lists of species, the nomenclature of all specimens was

standardized according to Pignatti (1982) for vascular plants, Aleffi andSchumacker (1995) for liverworts and Cortini Pedrotti (2001) for mosses.In the present paper, species richness at the plot scale is discussed in relation

to three groups of plants: (i) woody plants, (ii) total vascular plants and (iii)bryophytes (liverworts and mosses).

Data analysis

Compositional patternsThe main compositional patterns of the three groups were extracted usingDCA (Detrended Correspondence Analysis). The use of the compositionalpatterns of woody plants and total vascular plants, as predictors of the com-positional pattern of bryophytes was tested by linear, logarithmic and qua-dratic regressions of the plot scores for the 109 plots along the first DCA axisusing bryophyte species data with respect to the correspondent plot scoresobtained by using species data on woody plants and total vascular plants, anddeemed significant if p £ 0.05. The Mantel test was also used to test the nullhypothesis of no relationship between the dissimilarity matrix of bryophyteswith respect to those of woody plants and total vascular plants.

Species Richness and rarityThe total number of species recorded in each plot was used for each groupas indicator of species richness at the local scale. For each plot, rarity was

529

analysed by two different indicators, each calculated for woody plants, totalvascular plants and bryophytes: (i) the number of unique species (i.e. occurringin only that site, Colwell and Coddington 1994); (ii) a Rarity Index, calculatedaccording to the formula (1):

R ¼P

j Ijk N� Fj

� �=N

Sk

ð1Þ

where Ijk is the incidence of species j in site k, Fj is total number of sitescontaining species j, N is the total number of sites and Sk is the number ofspecies in the site k (Palmer et al. 2002). This index scales between zero (noinfrequent species) and one (only infrequent species). Linear, logarithmic andquadratic regression were used to test if species richness and rarity indicators ofwoody plants and total vascular plants could be used to predict the sameindicators of bryophytes, and deemed significant if p £ 0.05.

Rarefaction curvesWe analysed how the species rarefaction curve calculated for the three groupswere congruent within the same forest. Plot-based rarefaction curves werecalculated as the mean of 10,000 accumulation curves obtained with differentrandom sequence of plots (Sanders 1968; Gotelli and Colwell 2001), separatelyfor woody plants, total vascular plants and bryophytes. The ratio between thespecies rarefaction curve obtained for bryophytes and those calculated forwoody plants and vascular plants was then calculated. If bryophytes had thesame rarefaction patterns of woody plant species or vascular plant species,these ratios were expected to be horizontal.

Reserve selection analysis and taxon surrogacyWe used an Integer Linear Programming (ILP) approach, as described byRodrigues et al. (2000), to find the combinations of sites that maximized thenumber of species. A ‘reservation set’ was defined as the combination of siteswhich maximized the pooled number of species for that number of sites. Theprioritization of sites to maximize the pooled species richness using richness-based algorithms (Csuti et al. 1997; Polasky et al. 2001) may be done usingoptimization algorithms, which consider all the possible combinations for anynumber of sites, or heuristic (‘greedy’) algorithms. Although greedy algorithmsare reported to provide near-optimal solutions (Csuti et al. 1997; Moore et al.2003), optimization algorithms based on Integer Linear Programming (ILP),may find better solutions (Pressey et al. 1993; Church et al. 1996; Rodrigueset al. 2000; Rodrigues and Gaston 2002). The software C-Plex (ILOG 1999) wasused to find, for any given number of sites, all the possible solutions resulting inthe same maximum pooled number of species (multiple optima) for the threegroups. These combinations were considered as group-specific reservation sets.We then used the reservation sets optimized for woody plant species and

total vascular plant species to calculate the pooled number of bryophytes

530

species ‘captured’ by them. The performance in capturing bryophyte speciesrichness of these reservation sets was tested against the pooled number ofbryophyte species obtained in 10,000 random combinations of plots (Sanders1968; Gotelli and Colwell 2001) and those obtained by bryophyte-specific re-serve sets. Performance was measured as:

P% ¼X Nlv

10000

� �� �

ð2Þ

Where: P% = Performance index (as %) of a combination of n plots;Nlv = number of random combinations of n plots resulting in a poolednumber of bryophyte species lower than that obtained with the taxon-surro-gacy approach. Species richness values falling in the upper decile (i.e. P% ‡ 0.9)were considered to be significantly higher than the random or bryophyte-spe-cific reserve sets.

Results

Compositional patterns

The compositional pattern of bryophytes species was highly congruent withthose of woody plants and total vascular plants, as demonstrated by the rel-atively high predictive value of the first axes of the DCA ordination obtainedwith woody plants and total vascular plants with respect to those obtained withbryophytes (Figure 2a, b). The similarity in species composition of bryophyteswith respect to woody plants and total vascular plants was confirmed by theMantel test (p<0.001 in both cases).

Species richness and rarity

The pooled species richness of the 109 plots resulted in a total of 61 woodyplant species, 420 vascular plant species and 128 bryophyte species. Speciesrichness per plot was highly variable: 1–14 species for woody plants; 3–69species for vascular plants and 0–22 species for bryophytes (note that in con-trast to the woody and total species richness, bryophyte species richness wasnot the total species richness for the 400 m2 plot but the pooled species richnessobtained from the 24 bryoplots).Woody and vascular plant species richness showed a limited capability of

predicting bryophyte species richness at the plot scale. In fact, bryophytespecies richness within each plot was significantly related to woody plantspecies richness and vascular plant species richness (Figure 3a, b), but theamount of explained variance was relatively low in both cases.Neither the number of singletons nor the rarity index of bryophytes were

significantly related (p £ 0.05) to number of singletons or the rarity index of

531

woody plant species and total vascular plant species, indicating that the siteshosting the rare species of bryophytes were not the same in which the rarespecies of woody plants or total vascular plants were recorded. Similarly, thenumber of singletons and the rarity index of bryophytes were not related(p £ 0.05) to the species richness of woody plants or total vascular plantspecies.

y = 0.4419x + 40.243

R2 = 0.750;P < 0.001

0

50

100

150

200

250

300

350

400

450

0 100 200 300 400 500 600 700

Vascular Plants - DCA 1

Bry

ophy

tes

- D

CA

1y = 0.4234x + 53.004

R2 = 0.716; P < 0.001

0

50

100

150

200

250

300

350

400

450

0 200 400 600Woody Plants - DCA 1

Bry

ophy

tes

- D

CA

1

(a)

(b)

Figure 2. Relationship between the scores of the 96 plots (13 plots did not have any bryophyte

species) along the first DCA axis obtained with the species composition of bryophytes with respect

to the scores of the first DCA axis obtained with the species composition of woody plants (a) and

total vascular plants (b).

532

Rarefaction curves

The species rarefaction curves for the three groups did not behave consistentlyacross the six forests, indicating that the gradients of species accumulation ofthe three groups were dissimilar in the six forests (Figure 4). The ratio betweenthe rarefaction curve calculated for bryophytes to that calculated for woodyplants (Figure 5a) showed that in five out of six forests the obtained curveswere almost horizontal, suggesting a similar pattern of species accumulationfor bryophytes and woody plant species within the same forest. However, thisratio ranged from about four to less than one across different forests. In theFM forest, the bryophyte species rarefaction curve was much steeper than thatof woody plants and the corresponding ratio was not as horizontal as those

y = 0.6246x + 5.2712

R2 = 0.1187; p <0.001

0

5

10

15

20

25

30

0 5 10 15Woody plant species richness

Bry

ophy

te s

peci

es r

ichn

ess

y = 0.1321x + 4.6472

R2 = 0.115; p < 0.001

0

5

10

15

20

25

30

0 20 40 60 80Vascular plant species richness

Bry

ophy

te s

peci

es r

ichn

ess

(a)

(b)

Figure 3. Relationship between the bryophyte species richness and woody plant species richness

(a) and vascular plant species richness (b).

533

calculated for the other forests. The ratio between the rarefaction curve cal-culated for bryophytes and that calculated for all vascular plants (Figure 5b)showed almost horizontal curves in all the six forests suggesting similar gra-dients of species accumulation of bryophytes and all vascular plants, within the

0

5

10

15

20

25

30

0 10 20 30 40

Woo

dy P

lant

s

0

50

100

150

200

0 10 20 30 40

Poo

led

Num

ber

of S

peci

esV

ascu

lar

Pla

nts

0

10

20

30

40

50

60

70

0 10 20 30 40

Pooled Number of Plots

Bry

ophy

tes

(a)

(b)

(c)

Figure 4. Rarefaction curves calculated separately for each of the six forests for (a) woody plant

species, (b) vascular plant species and (c) bryophytes. The different symbols represent the six

forests: filled quadrats, Bandite di Follonica, Macchia della Magona, Montoni (BF), open quad-

rats, Colline Livornesi (CL); filled circles, Farma-Merse e Belagaio (FM), open circles, Madonna

delle Querce (MQ); filled triangles, Foreste Casentinesi (FC), open triangles, Foreste Pistoiesi (FP).

534

same forest. This ratio ranged from about 0.2 to less than 0.6 across differentforests, being thus rather homogeneous.

Reserve selection analysis and taxon surrogacy

Optimizing for maximum woody plant species richness gave many differentpossible combinations of plots (reservation sets) even when selecting only a fewplots (Table 2). To obtain a reservation set that included all species 26 plots(with more than 100 possible combinations) were needed (Figure 6a). Opti-mizing for maximum total vascular plant species richness resulted in fewercombinations (Table 2) but needed of 59 plots (6 different combinations) toobtain a reservation set including all species (Figure 6b). The reservation setsoptimized to maximize bryophyte species richness resulted in intermediate

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

0 5 10 15 20 25 30 35

Bry

ophy

te S

/ W

oody

S

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0 5 10 15 20 25 30 35Pooled Number of Plots

Bry

ophy

te S

/ V

ascu

lar

S

(a)

(b)

Figure 5. Ratios of the bryophyte to the woody plant rarefaction curves (a) and bryophyte to the

total vascular plant curves (b). The different symbols represent the six forests: filled quadrats,

Bandite di Follonica, Macchia della Magona, Montoni (BF), open quadrats, Colline Livornesi

(CL); filled circles, Farma-Merse e Belagaio (FM), open circles, Madonna delle Querce (MQ); filled

triangles, Foreste Casentinesi (FC), open triangles, Foreste Pistoiesi (FP).

535

Table

2.

Number

ofpossible

solutionsforthereservationsets

basedontheoptimizationofthemaxim

um

pooledspeciesrichnessofwoodyplants,total

vascularplants

andbryophytes(only

combinationsforstepsof3plots

are

reported

forsimplicity).

Number

of

PooledPlots

36

912

15

18

21

24

27

30

33

36

39

42

45

48

51

54

57

Woodyplants

211

1>100

>100

>100

>100

>100

Vascularplants

11

13

123

225

1>100

>100

>100

4>100

>100

>100

>100

>100

>100

Bryophytes

42

12

12

19

20

43

>100

>100

>100

>100

3

536

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70 80 90 100 110

Number of selected plots

Pool

ed n

umbe

r of

spe

cies

0

50

100

150

200

250

300

350

400

450

0 10 20 30 40 50 60 70 80 90 100 110Number of selected plots

Pool

ed n

umbe

r of

spe

cies

0

20

40

60

80

100

120

140

0 10 20 30 40 50 60 70 80 90 100 110

Number of selected plots

Pool

ed n

umbe

r of

spe

cies

(a)

(b)

(c)

Figure 6. Number of species obtained with specifically optimized reserve sets (filled circles) for

woody plant species (a), vascular plant species (b) and bryophytes (c) compared with their

respective rarefaction curves (open circles±1 SD). For bryophytes the number of species obtained

with the reserve sets optimized for woody plant species (bold line) or total vascular plant species

(thin line) is also given.

537

numbers of possible combinations (Table 2) and needed of 36 plots (2 differentcombinations) to obtain a set of sites including all species (Figure 6c).The pooled bryophyte species richness for reservation sets optimized for

woody plant species or vascular plant species richness were lower than thosespecifically optimized for bryophyte species richness, and only slightly higherthan the mean plus one standard deviation of the reservation set based onrandom combinations of sites (Figure 6c). The reservation sets optimized forwoody plants resulted in extremely variable values of bryophyte pooled speciesrichness. With respect to the specifically optimized reservation sets, a propor-tion from 42.1% to 59.3%, for the lower values, and from 53.9% to 96.6%, forthe higher values, of bryophyte species was found in these reservation sets(Figure 7a). Most of the higher values showed a performance index ‡0.9 butnone of the lower values showed a performance index ‡0.9. This indicates that,for any number of selected plots, some of the reserve sets based on woody plantspecies richness were also effective for optimizing bryophyte species richnesswhile others were not. The reservation sets obtained by optimizing total vas-cular plant species richness performed more homogeneously in terms of pooledbryophyte species richness (Figure 7b); with respect to the specifically opti-mized reservation sets, a proportion from 36.4% to 84.4% (for both the lowerand the higher values) of bryophyte species was found in these reservation sets(Figure 7b). Also in this case, only some of the reservation sets optimized forall vascular plants obtained a performance index ‡0.9 in optimizing bryophytespecies. This indicates that the reservation sets optimized for total vascularplants do not always optimize bryophyte species.

Discussion

Compositional patterns

Our analyses showed that the main compositional pattern of bryophytes washighly congruent with those of woody plants and total vascular plants, sug-gesting that changes in species composition of bryophytes reflect those ofwoody plants and total vascular plants. This suggests that patterns in woodyplant or total vascular plant species composition may be used to infer the maincompositional patterns of bryophytes, at least at the plot scale within Medi-terranean forest ecosystems. Nekola and White (1999) found that, in NorthAmerican spruce-fir forests, the rate of similarity decay with increasing dis-tance was 1.5–1.9 times higher for vascular plants than for bryophytes, indi-cating that the species composition of vascular plants changed more rapidlythan that of bryophytes along environmental gradients. Pharo et al. (1999)found significant correlations between the patterns of species turnover ofbryophytes and different groupings of vascular plants (fern, overstory species,understory species, all vascular plants), though none of the correlations wereparticularly strong. In addition, Saetersdal et al. (2004) found that changes in

538

species composition of vascular plants were reflected in comparable degrees ofchange in the species composition of bryophytes as well as of other taxonomicgroups, including lichens, spiders, carabids, staphylinids, snails and polyporefungi. On the other hand, Negi and Gadgil (2002) reported a reduced con-cordance between the species turnover patterns of vascular plants and bryo-phytes in different habitat types in India.

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20 25

Number of Plots

% o

f m

ax #

of

bryo

phyt

e sp

ecie

s

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60

Number of Plots

% o

f m

ax #

of

bryo

phyt

e sp

ecie

s

(a)

(b)

Figure 7. Proportion of bryophyte species richness (expressed as % of the optimized maximum)

resulting in the reserve sets obtained by maximizing the number of woody plant species (a) and the

number of vascular plant species (b). In each graph, the upper and lower curves denote respectively

the maximum and the minimum proportion of species obtained with different combinations of n

sites. The asterisks denote that the obtained values of bryophyte species richness have a perfor-

mance index higher than 0.9.

539

It appears to be almost a general rule that species composition of bryophytesoften changes in a more or less parallel manner with respect to that of vascularplants. However, this pattern and its magnitude are likely to be ecosystem andscale dependent and it is not possible to see a unique model of cross-taxon co-variation in species composition with the data currently available.

Species richness and rarity

Negi and Gadgil (2002) found that in different habitat types in the Chamolidistrict of Uttaranchal, Indian Garhwal Himalaya, the species richness ofwoody plants was significantly related to that of mosses, though not to liver-worts. Ingerpuu et al. (2001) found a good correlation between bryophyte andvascular plant species richness in boreo-nemoral moist forests and mires, at theregional scale and the ten-stand scale and marginally non-significant correla-tions between them at the 1 ha stand scale. In a range of forest types in thecostal lowlands of eastern Australia, Pharo et al. (1999) found that bryophytespecies richness was well correlated with that of ferns and with that of over-story species, but not at all with total plant species richness. Saetersdal et al.(2004) found that bryophyte species richness was highly correlated with vas-cular plant species richness in 0.25 ha stands sampled in forest ecosystems ofNorway. In our sample, the bryophyte species richness was found to be onlyweakly correlated with the woody plant or total vascular plant species richness,indicating that the species richness of these taxa cannot be used as a goodsurrogate for bryophyte species richness at the local, i.e. plot, scale. Bryophyterarity did not show any apparent relation to the patterns of rarity or speciesrichness of either woody plants or total vascular plants, confirming that vas-cular plant rarity and bryophyte species richness and rarity are not related atall, as suggested by the above cited studies. In addition any cross-taxon cor-relation between these taxa is likely to be dependent on the spatial scale,complicating the whole picture.

Rarefaction curves

Gotelli and Colwell (2001) observed that the category–subcategory taxonomicratios, frequently used in biogeography, suffer from sample-size dependenceand thus should be analyzed by using ratios of the respective rarefactioncurves. We adopted a similar approach – the ratio between the rarefactioncurves of two different taxa from the same sites – to analyze the cross-taxoncongruence in relation to the sampling intensity between bryophytes andwoody plants or total vascular plants. As far as we know this is the firstattempt to investigate this aspect of community ecology and conservationbiology by using such an approach. The six forests investigated differed widelyin this ratio, as well as in their rarefaction curves. This indicates that different

540

gradients of species accumulation operates in the six forests for the threegroups considered and it is extremely difficult to use vascular plants – or woodyplants – to estimate larger scale diversity of bryophytes. However, given thelack of examples of this type of analyses, further studies may reveal interestingpatterns in the species rarefaction processes of cross-taxon congruence. Thisreplication is particularly important given that the use of the slope of speciesaccumulation curves as an index of b-diversity remains controversial (Scheiner2003; Gray et al. 2004). However, it is evident that the greater the slope of theaccumulation curve the higher the species dissimilarity among sampling sites is(Ricotta et al. 2002). The comparison of species accumulation curves or theratio between taxa can reveal analogies or differences in this particular aspectof species diversity.

Reserve selection analysis and taxon surrogacy

Ingerpuu et al. (2001) concluded their study on vascular plants and bryophytesof boreo-nemoral forests and mires, stating that ‘the species richness betweenbryophytes and vascular plants is positively correlated on larger scales, andconservation of communities rich in species of one group maintains also thespecies richness of the other’. This is the logical basis for most conservationprogrammes using one (or a few) taxonomic groups as indicators for theselection of nature reserves.Our results indicate that the selection of reserve sets of sites by maximizing

vascular plant species richness can only be partially effective in maximizingbryophyte species richness. When quantitatively checking the amount ofbryophyte species captured in reserve networks based on vascular plants theresults may be contradictory. Pharo et al. (2000) found that a network of sitesthat captured 90% of vascular plants also captured 87% of lichen species butonly 75% of bryophytes species. In our survey, we found that the averageproportion of bryophyte species captured in reservation sets based on woodyplant species richness varied between 50.7% for the worst performing solutionsand 84.9% for the best performing ones. The corresponding values obtained byreservation sets based on vascular plant species were 69.1% and 72.3%.However, especially for the reservation sets based on woody plants, the rangeof bryophyte species richness captured by different combination of plots wasextremely high.Virolainen et al. (2000) found that, for small patches within old-growth

forests in Finland and Sweden, complementary networks based on vascularplants efficiently captured a high proportion of species richness of other tax-onomic groups, but bryophytes were not analysed. Saetersdal et al. (2004)extended the results of Virolainen et al. (2000) to forest ecosystems of Norwayand also included in their survey bryophytes; in their analysis of the cumulativepercentage of species of the different species groups captured as a function ofcumulative percentage area from a complementary selection of sites based on

541

vascular plant species, the proportion of bryophyte species was constantlyhigher with respect to that of other taxonomic groups. Our work indicatedthat, although a relatively high proportion of bryophyte species, compared torandom expectations, was captured by reserve sets based on vascular plantspecies, this proportion was seldom significantly different from that of randomcombinations of sites.Designing efficient networks of nature reserves based on maximizing the

pooled species richness of a target taxon is computationally possible withpresent-day computers (Rodrigues and Gaston 2002; Rodrigues et al. 2004).This approach is becoming popular and it is likely that these methods willresult in practical conservation applications in a near future. However, sincethe data are normally available only for a few taxonomic groups, this methodrequires a taxon surrogacy approach (Ryti 1992; Prendergast et al. 1993;Howard et al. 1998; Garson et al. 2002) which may not effectively allow fordifferences in basic components of community structure (e.g. within site orbetween site species richness). In forest ecosystems of Tuscany, the selection ofsites which maximize the vascular plant species richness was not efficient inmaximizing the bryophyte species richness. This may be due to the limitedcorrelation of within-site species richness combined with the totally uncorre-lated patterns of rare species. A similar result was found for fungi in reservesets optimized for vascular plants in 25 forest stands of Tuscany (Chiarucci etal. 2005). Again in that case, fungal species composition was related to vascularplant species composition, though with lower explained variance, and therewas no correlation between stand scale fungal and plant species richness(Chiarucci et al. 2005).Although the issue of the spatial scale certainly affects taxon-surrogacy ap-

proaches used in reserve selection this was not considered here. Two basiccomponents of the spatial scale can affect the results: the scale at which thedata are aggregated, the plots in this case or geographic cells in others, and thegrain in which the data are really collected. Our bryophyte data were gatheredby 24 small sampling units within a plot and almost certainly do not providecomplete lists of species within the plot. However, it is almost impossible tohave complete species lists for several large sampling units and the use of alarge number of smaller sampling units is a necessity, especially for bryophytesand other difficult-to-sample taxa.

Conclusion

The present study demonstrated that compositional gradients of woody plantsand total vascular plants at the plot scale are well reflected by the composi-tional patterns of bryophytes in forest ecosystems of Tuscany, central Italy. Areduced predictive value of species richness of woody plants and total vascularplants was found for bryophyte species richness at the plot scale and an evenlower performance was found in using plants for the optimization of reserve

542

site selection. In addition, bryophyte species rarity was uncorrelated with plantspecies rarity. This reduced performance of plant species richness as indicatorsof bryophyte species richness indicates that specific data on bryophyte speciescomposition are needed when some interest is given to this group and thatcross-taxon congruence is scarcely usable for detecting their patterns, at least inMediterranean forest ecosystems.

Acknowledgements

Paper No. 29 of the research project MONITO – (MONitoraggio Intensivoforeste TOscane), funded by the Regional Administration of Tuscany and theEuropean Community on the Regulations EEC 2157/92 and 3528/86. Theproject was also supported by a grant of the University of Siena (PAR Progetti2001) to the first author (AC). We also wish to acknowledge Alfonso Riva,Luisa Frati, Francesca Casini, Antonio Gabellini, Marco Cetoloni, DanieleViciani, Federico Ghini, Manuela Boddi and Giacomo Nicoletti for collabo-ration during the sampling activities. A special thank is due to Duccio Rocchinifor completely managing the GIS and GPS systems employed during the re-search. A very special thanks to the ever patient Barbara Anderson forattempting to correct the English.

References

Aleffi M. and Schumacker R. 1995. Check-list and red list of the liverworts (Marchantiophyta) and

hornworts (Anthocerotophyta) of Italy. Flora Mediterranea 5: 73–161.

Bartolozzi L., Bonini I., Boretti R., Bussotti F., Cenni E., Chiarucci A., Cozzi A., De Dominicis V.,

Ferretti M., Grassoni P., Landi G., Leonzio C. and Vignozzi G. 2002. Forest ecosystem moni-

toring in Tuscany (Italy): past activities, present status and future perspectives. J. Limnol. 61:

129–136.

Bernetti G. 1987. I boschi della Toscana. Giunta Regionale Toscana. Edagricole, Bologna.

Chiarucci A., De Dominicis V. and Wilson J.B. 2001. Structure and floristic diversity in permanent

monitoring plots in forest ecosystems of Tuscany. For. Ecol. Manage. 141: 203–212.

Chiarucci A. and Bonini I. 2005. Quantitative floristics as a tool for the assessment of plant

diversity in Tuscan forests. For. Ecol. Managem. 212: 160–170.

Chiarucci A., D’Auria F., De Dominicis V., Lagana A., Perini C. and Salerni E. 2005. Using

vascular plants as a surrogate taxon to maximize fungal species richness in reserve design.

Conserv. Biol. 19: 1644–1652.

Church R.L., Stoms D.M. and Davis F.W. 1996. Reserve selection as a maximal covering location

problem. Biol. Conserv. 76: 105–112.

Colwell R.K. and Coddington J.A. 1994. Estimating terrestrial biodiversity through extrapolation.

Philos. Trans. R. Soc. Lond. B 345: 101–118.

Cortini Pedrotti C. 2001. New check-list of Mosses of Italy. Flora Mediterranea 11: 23–107.

Csuti B., Polasky S.,Williams P.H., PresseyR.L., Camm J.D., KershawM., Kiester A.R., Downs B.,

Hamilton R., HusoM. and Sahr K. 1997. A comparison of reserve selection algorithms using data

on terrestrial vertebrates in Oregon. Biol. Conserv. 80: 83–97.

Garson J., Aggarwal A. and Sarkar S. 2002. Birds as surrogates for biodiversity: an analysis of a

data set from southern Quebec. J. Biosci. 27: 347–360.

543

Gotelli N.J. and Colwell R.K. 2001. Quantifying biodiversity: procedures and pitfalls in the

measurement and comparison of species richness. Ecol. Lett. 4: 379–391.

Gray J.S., Ugland K.I. and Lambshead J. 2004. Species accumulation and species area curves – a

comment on Scheiner (2003). Glob. Ecol. Biogeogr. 13: 473–476.

Howard P.C., Viskanic P., Davenport T.R.B., Kigenyi F.W., Baltzer M., Dickinson C.J.,

Lwanga J.S., Matthews R.A. and Balmford A. 1998. Complementarity and the use of indicator

groups for reserve selection in Uganda. Nature 394: 472–475.

ILOG 1999. CPLEX 6.5. ILOG. Gentilly, France.

Ingerpuu N., Vellak K., Kukk T. and Partel M. 2001. Bryophyte and vascular plant species

richness in boreo-nemoral moist forests and mires. Biodivers. Conserv. 10: 2153–2166.

Moore J.L., Folkmann M., Balmford A., Brooks T., Burgess N., Rahbek C., Williams P.H. and

Krarup J. 2003. Heuristic and optimal solutions for set-covering problems in conservation

biology. Ecography 26: 595–601.

Negi H.R. and Gadgil M. 2002. Cross-taxon surrogacy of biodiversity in the Indian Garhwal

Himalaya. Biol. Conserv. 105: 143–155.

Nekola J.C. and White P.S. 1999. The distance decay of similarity in biogeography and ecology. J.

Biogeogr. 26: 867–878.

Palmer M.W., Earls P.G., Hoagland B.W., White P.S. and Wohlgemuth T. 2002. Quantitative

tools for perfecting species lists. Environmetrics 13: 121–137.

Panzer R. and Schwartz M.W. 1998. Effectiveness of a vegetation based approach to invertebrate

conservation. Conserv. Biol. 12: 693–702.

Pharo E.J., Beattie A.J. and Binns D. 1999. Vascular plant diversity as a surrogate for bryophyte

and lichen diversity. Conserv. Biol. 13: 282–292.

Pharo E.J., Beattie A.J. and Pressey R.L. 2000. Effectiveness of using vascular plants to select

reserves for bryophytes and lichens. Biol. Conserv. 96: 371–378.

Pignatti S. 1982. Flora d’Italia. Edagricole, Bologna.

Polasky S., Csuti B., Vossler C.A. and Meyers S.M. 2001. A comparison of taxonomic distinctness

versus richness as criteria for setting conservation priorities for North American birds. Biol.

Conserv. 97: 99–105.

Prendergast J.R., Quinn R.M., Lawton J.H., Eversham B.C. and Gibbons D.W. 1993. Rare spe-

cies, the coincidence of diversity hotspots and conservation strategies. Nature 365: 335–337.

Pressey R.L., Humphries C.J., Margules C.R., Vane-Wright R.I. and Williams P.H. 1993. Beyond

opportunisms: key principles for systematic reserve selection. Trends Ecol. Evol. 8: 124–128.

Ricotta C., Carranza M.L. and Avena G. 2002. Computing ß-diversity from species area curves.

Basic Appl. Ecol. 3: 15–18.

Rodrigues A.S., Cerdeira J.O. and Gaston K.J. 2000. Flexibility, efficiency, and accountability:

adapting reserve selection algorithms to more complex conservation problems. Ecography 23:

565–574.

Rodrigues A.S.L., Andelman S.J., Bakarr M.I., Boitani L., Brooks T.M., Cowling R.M., Fishpool

L.D.C., da Fonseca G.A.B., Gaston K.J., Hoffmann M., Long J.S., Marquet P.A., Pilgrim J.D.,

Pressey R.L., Schipper J., Sechrest W., Stuart S.N., Underhill L.G., Waller R.W., Watts M.E.J.

and Yan X. 2004. Effectiveness of the global protected area network in representing species

diversity. Nature 428: 640–643.

Rodrigues A.S.L. and Gaston K.J. 2002. Optimisation in reserve selection procedures – why not?

Biol. Conserv. 107: 123–129.

Ryti R.T. 1992. Effect of the focal taxon on the selection of nature reserves. Ecol. Appl. 2: 404–410.

Saetersdal M., Gjerde I., Blom H.H., Ihlen Per.G., Myrseth E.W., Pommeresche R., Skartveit J.,

Solhoy T. and Aas O. 2004. Vascular plants as a surrogate species group in complementary site

selection for bryophytes, macrolichens, spiders, carabids, staphylinids, snails, and wood living

polypore fungi in a northern forest. Biol. Conserv. 115: 21–31.

Sanders H.L. 1968. Marine benthic diversity: a comparative study. Am. Nat. 102: 243–282.

Scheiner S.M. 2003. Six types of species-area curves. Glob. Ecol. Biogeogr. 12: 441–447.

544

Tardif B. and DesGranges J.-L. 1998. Correspondence between bird and plant hotspots of the St.

Lawrence River and influence of scale on their location. Biol. Conserv. 84: 53–63.

Virolainen K.M., Ahlroth P., Hyvarinen E., Korkeamaki E., Mattila J., Paiivinen J., Rintala T.,

Suomi T. and Suhonen J. 2000. Hot spots, indicator taxa, complementarity and optimal net-

works of taiga. Proc. R. Soc. Lond. B. 267: 1143–1147.

Williams P.H. 1999. Key sites for conservation: area selection methods for biodiversity. In: Mace

G.M., Balmford A. and Ginsberg J.R. (eds) Conservation in a Changing World: Integrating

Processes into Priorities for Action. Cambridge University Press, Cambridge, pp. 211–249.

545