基本HTML版本
Intl. J. of Mol. Evol. and Biodivers. 2011, Vol. 1, No. 1-5
http://ijmeb.sophiapublisher.com
4
arid climates, and also in wet, cold regions of Spain
(Hódar et al., 1996). These traits could suggest a
similar adaptation for
A. erythrurus, P. hispanicus
and
P. algirus
. However, if we were to quantify the other
habitat variables (presence of rocks, soil composition,
and prey abundance) and run further analyses then
possibly our results would yield different outcomes, as
our two habitats may show to be more diverse.
Likewise, Michael et al (2010) discovered that rocky
environments are related to reptile abundance.
The Generalized Linear Model showed that the
interaction between habitat, wind speed and air
temperature had no effect one escape distance. In
future experiments, it would be wise to take
calculations of temperature and wind speed even at
sites where no lizards are found. Then a Generalized
Linear Model can be constructed on the effect of these
interactions terms, together with habitat and escape
distance on lizard abundances.
To conclude, lizards did not have a preference for one
particular habitat since there was no difference
between the escape distance, vegetation cover and,
therefore, no difference between lizard abundances
between each of the sites. This implies that these three
species are occupying the same space, and feeding on
the same arthropod populations and foraging in the
same area. Thus, species competition to determine if
these species have (or potentially will) evolve certain
adaptations to the environment needs further studies.
3 Methods
3.1 Study area
Observations were conducted during spring (April 1-4,
2011) in the sparsely vegetated habitats of the sloppy
desert in Sorbas Karst Gypsum Natural Monument,
and in a sandy (Hughes, 2011, unpublished data) dry
river valley near Sorbas, Almeria Province, South-east
Spain. Dominant vegetation in the region is
Stipa
Tenacissima
,
Anthyllis cytisoides
,
Retama Sphaerocarpa
,
Garigue shrubs, and tall maquis-like vegetation in the
river valley (Hughes, 2011, unpublished data).
3.2 Field recordings
Between the hours of 0930-1330 and 1430-1730 hours
(G.M.T.+1), we haphazardly walked at a constant pace
of about 15m per minute in a zigzag pattern (Martín
and López, 2003) for 1.5 hours in each of two study
sites (20m apart) in both the gypsum desert and the
dry river valley. Approximately every four minutes on
the first day, a 1m
2
quadrat sample was taken of
percent vegetation cover. Twenty quadrats were taken
in total for each site. At every lizard sighting, the date,
time, identification, size, and color of species,
adult/juvenile, microhabitat with presence of prey, air
temperature/humidity of sunny/shaded areas, escape
distance to refuge, wind speed (Schoener, 1970;
Pianka and Parker, 1972), and aspect were recorded.
Digital photographs aided in further identification of
lizard and vegetation species. A compass was used to
record the aspect (Fisher et al., 2003), and a
thermo-hydro-anemometer was used to record the
temperature, humidity and wind speed. A transect line
was used to measure the escape distance from the start
point (where we first spotted the lizard) to the refuge.
4 Data analysis
All of the raw data showed non-normal distribution,
which when tested with a Shapiro-Wilks normality
test for both logged and square root transformed data
still showed non-normal distribution. Therefore,
non-parametric tests were rendered. The differences
on habitats were analyzed via Fisher’s Exact Test. A
Pearson’s correlation between escape distance and
temperature, and also escape distance and wind speed
were analyzed. Mann-Whitney Test was used to test
for a significant difference between each species in
their relative escape distances. Kruskal-Wallis tests
were used to analyze the difference of lizard numbers
for each site; the vegetation cover for each site; and
escape distance for each site. Fisher’s Exact Test was
performed to dictate the abundance of each species in
each habitat, and also to test for significance of lizard
abundances in each habitat as a function of time. A
Generalized Linear Model was constructed to test for
significance of the effect of the interactions between
wind speed × air temperature × habitat (and all of the 2
way interactions) on escape distance. Data analyses were
performed using SPSS 17 and Microsoft Excel 2007.
References
Andersson M., Krockenberger A., and Schwarzkopf L., 2010, Experimental
manipulation reveals the importance of refuge habitat temperature