Study site
The research interested, from April 2004 to September 2006,
Khersonesskoe (44°35'12"N; 33°24'00"E) and Koyashskoe (45°02'31"N;
36°12'20"E) lakes in the Crimea peninsula (Ukraine), Nartë (40°32'13"N;
19°25'48"E) saltworks (Southern Albania), Vecchia Salina (40°18'06"N;
17°43'56"E) at Torre Colimena (Gulf of Taranto, Apulia, Italy), Pantano
Grande and Pantano Roveto (36°48'29"N; 15°06'02"E) at Vendicari Nature
Reserve (Sicily, Italy) (Figure 5).
Collections
of resting stages from the sediments of the 2 contiguous lakes at
Vendicari (Sicily, Italy) allowed to compare results coming from nearby
habitats with those coming from very distant ones (e.g. Italian and
Crimean lakes).
Sampling procedures
Samplings were collected during summer months corresponding to the
minimum water level, or even its absence to take only samples not
submerged. Three replicate sediment cores (diameter, 7.5 cm; depth, 6
cm) were obtained from each lake by using a core sampler (20 cm length)
and stored at 4°C in a refrigerator for 1 year. For the analysis, each
sediment core was cut into 3 cm thick layers. Each layer was
ultrasonified to break the larger particles of sediment and then sieved
at two mesh sizes (212 and 45 μm). The
sediment collected by both sieves was centrifuged at 1,090 g in a 1:1
sucrose-distilled water solution for 3 min. The supernatant derived
from the centrifugation of the two sieve fractions was analysed to
separate cysts.
Cysts were reported as number per 100 cc of sediment. The most abundant resting stages (2 types for the fraction >212 μm; 3 types for the fraction >45 μm)
were used in hatch experiments in the laboratory. Sets of 30 cysts of
each morphotypes, taken from each layer (from the superficial to the
deepest one), were stored in 3 cc wells raised with 2 cc of original
water filtered at 0,45 μm. To avoid bacterial growth, in each well, 20 μl
of an antibiotic mix (streptomycin/penicillin 1:1) was added. Resting
stages were submitted to different storage conditions in thermostatic
rooms (an "equinox" simulation, with 13°C and 12 hL:12 hD photoperiod
and a "early summer" simulation, with 24°C and 14 hL:10 hD photoperiod)
at 4 different salinity values (46‰, 36‰, 26‰, freshwater) obtained by
diluting the original-site water. Hatching plates were checked daily to
test the presence of active stages, which were counted and removed for
identification.
The related water column, on each lake, has been sampled in
different periods of the year (at least 2 seasons), and in different
years (between 2002 and 2006) to compare its faunal composition to each
other. Zooplankton samples were collected monthly (three replicates),
with two plankton nets (mouth diameter, 25 cm; length, 65 cm; mesh
size, 200 μm and 50 μm)
towed horizontally, equipped with a water-flow meter at the mouth. The
Italian hypersaline lake Vecchia Salina and Crimean ones were already
studied in the past (see data in [10] and [16], respectively).
Data analysis
Data were analysed by multivariate statistical techniques with a
non-parametric approach because of the wide disparity in density of
some cysts in different lakes. The significance of the spatial
variation in "cyst banks" composition was tested using a One-Way
Analysis of Similarities for replicated data (ANOSIM) routine in PRIMER
(Plymouth Routines In Multivariate Ecological Research) version 6β R6 (PRIMER-E) [21].
For multivariate analyses, the absolute densities of each
morphotypes were fourth root transformed, to severely down-weight the
importance of the very abundant species so allowing the less dominant,
and even the rare morphs, to play some role in determining similarity
among samples.
Stress values were shown for each MDS plot to indicate the goodness of representation of differences among samples [22]. A One-Way similarity percentages procedure (PRIMER SIMPER routine, Clarke [22]) was used in order to obtain the percentage contribution that each taxon provided
to Bray-Curtis similarities measures. A cut-off criterion was applied
to allow the identification of a subset of species whose cumulative
percentage contribution reached 80% of similarity value. SIMPER
analysis consented to identify the species responsible for the
biological characterisation of the "seed banks" stored in each
investigated lake.
Answers to all your Biology Questions
