**Marine Biology, in
press**

**Oxygen
consumption of the semiterrestrial crab ***Pachygrapsus
marmoratus* in relation to body mass and temperature - an information theory
approach**
**

** Stelios
Katsanevakis,
****John
Xanthopoulos, Nikos Protopapas**,
**George Verriopoulos
**

Department
of Zoology-Marine Biology, School of Biology, University of Athens,
Panepistimioupolis, 15784 Athens, Greece

**Abstract**

*Pachygrapsus
marmoratus *is a semi-terrestrial
crab and the most common grapsid crab in the intertidal belt of rocky shores
throughout the Mediterranean Sea, Black Sea and northeastern Atlantic. In this
study, the combined effects of temperature (*T*),
body mass (*M*), and sex (*S*) on
the routine oxygen consumption rate (*R*)
in *P. marmoratus* were quantified. The
blotted wet body mass of the specimens ranged between 43 mg and 18.0 g, and five
test temperatures were used between 13.5 ^{o}C and 28.0 ^{o}C. Six candidate models that reflected different assumptions regarding the
dependence of *R* on *S* and *T* were compared. Model
selection was based on Kullback-Leibler’s information theory and Akaike’s
information criterion (AIC). The
model
had the highest support by the data
(*E* is the activation energy, *B *= 8.618x10^{-5} eV K^{-1}
is Boltzmann’s constant, *T*_{a} is the absolute temperature in
degrees Kelvin, and *b* the allometric scaling exponent); for *P.
marmoratus *it was found that
. No sex dependence of *R* was supported by the data. Following a multi-model inference approach, the mean (± SE) allometric
exponent
was
0.750 (± 0.013) having a 95% (bootstrap) confidence interval of 0.726 to 0.774.
Thus, it was established that *P.
marmoratus* follows Kleiber’s ¾ law, as seems to be generally true
for intertidal crabs. The allometric exponent was independent of temperature as
has also been reported for many other marine invertebrates (at normal
temperatures). Q_{10} values were relatively low, indicating wide
thermal tolerance of the species. Model selection based on information theory is
recommended for respiration studies, as an effective method in finding a
parsimonious approximating model. Multi-model inference by model averaging,
based on Akaike weights, is an effective way to make robust parameter
estimations and deal with model selection uncertainty.