Page 5 - Tomasello_Borra_etal2009

Seagrass meadows at the extreme of environmental tolerance
performed using a Bayesian method (Rannala & Moun-
tain 1997) and tested over 1000 replicates, utilizing the
software GENCLASS ver. 2.0 (Piry et al. 2004).
A CFA (correspondence factorial analysis) based on
allele frequency was achieved with the software GENETIX
ver. 4.1 (Belkhir et al. 1996–2002) to visualize genetic
relationships among populations and individuals.
Growth and flowering performance analysis
Estimates of growth performance of Posidonia oceanica,
expressed in terms of annual rhizome elongation and
annual leaf formation, were obtained by the lepidochro-
nological analysis (Pergent 1990). This reconstructive age-
ing technique is based on the analysis of cyclic variations
in the sheath thickness. Cycles are delimited by two con-
secutive minimum sheaths’ thickness, corresponding to a
1-year period (lepidochronological year). Lepidochrono-
logical years were dated starting from the rhizome apex
(sampling year) downward and then matching the cycle
of sheath thickness with the corresponding rhizome seg-
ment. Annual leaf formation and rhizome elongation
were derived, respectively, from the number of sheaths
and the length of rhizome segment per lepidochronologi-
cal year. Moreover, the shoot age was determined by add-
ing the number of annual rhizome segments from the
bearing axis and the most recent year (Pergent &
Pergent-Martini 1990). This method also made it possible
to detect and date the past flowering occurrence per
shoot across time by finding floral stalk remains between
the sheaths (Pergent et al. 1989). Only the most recent
lepidochronological year has been considered for compar-
ison of growth performance among sites, as the temporal
horizon of the time series varied significantly.
Statistical analysis
Differences of mean growth performances among sites
were assessed by linear model (LM) (Underwood 1997).
If the age of shoots sampled was different among sites,
the possible confounding effect of shoot age has been
taken into account by including it as covariate in the sta-
tistical model (Calvo et al. 2006; Tomasello et al. 2007).
Before parameterization, residual homoscedasticity were
assessed using the Levene test (Glaser 1983). When the
heteroscedasticity of residuals was present, the response
variable has been transformed to fit the assumptions of
linear models. In the parameterization used for LM, it
was required to refer all results to the intercept, a baseline
conventional category that, in this case, was set with the
Favignana site. Finally the effects were presented in the
original scale to facilitate the interpretation of results.
For flowering occurrence, the analysis was made taking
into account two data sources: lepidochronological series
292
Tomasello, Di Maida, Calvo, Pirrotta, Borra & Procaccini
obtained in this study and existing lepidochronological
series collected in previous studies carried out in the same
area (MURST 2003). This allowed us to increase the time
range and the number of retrodated shoots from 29 to
55 years and from 243 to 1009, respectively. To detect the
pattern of flowering occurrence in space and time, the
following indices were calculated: (i) meadow flowering
frequency (FF) as the fraction of years that a given mea-
dow has flowered, (ii) shoot flowering probability (Pf) as
the fraction of flowering stalks per total number of shoot
annual segments, (iii) meadow flowering intensity (FI) as
fraction of shoots presenting stalks within a particular
year, and (iv) annual flowering prevalence (FP) as the
fraction of meadows flowering in a given year (Diaz-Al-
mela et al. 2006). If the shoots coming from both data
sources were collected in the same sampling site, a unique
value per each index was calculated. Otherwise, if they
were collected in more than one station per site (within
less than 1 km distance or at different depth), the index
values were computed by averaging the values of different
stations (Diaz-Almela et al. 2006).
All indices were estimated according to minimum sam-
ple size specific for each index proposed by Diaz-Almela
et al. (2006, 2007). Statistical analyses have been carried
out using SPSS 14.
Results
Genetic variability and structure within sites
A total of 80 alleles were identified for the 13 micro-
satellite loci utilized. According to the total number of
alleles, the theoretical number of genotypes was
Ng = 1.12 · 10+14, showing the high polymorphism of
the markers utilized. Seventeen genotypes from the total
of 218 samples analysed were not complete for one to
two loci and were excluded from the assessment of the
basic statistics. Of the 201 complete multilocus genotypes,
192 were distinct (Table 1). Within-population pairwise
analysis of allele sharing showed that, in the Atolls
population, in 20 cases multilocus genotypes (MLG) were
different only in one allele. Only one case was recorded
among all the other populations analysed (i.e. Re´cif site).
Nevertheless, among the 20 pairs present in the Atolls
site, only in one case did the psex value suggest a merging
of two MLGs in one MLL (Table 1).
Clonal richness, expressed as R, ranged from 0.79 in
the Atolls to 1.0 in all the other populations except the
Re´cif (Table 1). Average clonal diversity for the two Stag-
none lagoon sites (R = 0.86; Atolls, Re´cif) was lower than
for open sea sites (R = 1.00).
Expected heterozygosis (Hexp) and observed hetero-
zygosity (Hobs) values ranged, respectively, from 0.36
Marine Ecology 30 (2009) 288–300 ª 2009 Blackwell Verlag GmbH