Dear MARMAM subscribers,
on behalf of my colleagues I am pleased to share with you our recent
publication:
Piero Manfredi, Luigi Marangi, Alessia Rossi, Giovanni Santangelo
(2016). An improved model life table for the Indian River Lagoon
bottlenose dolphin population and remarks on early mortality. Marine
Mammal Science. 32(4): 1522-1528.
http://onlinelibrary.wiley.com/doi/10.1111/mms.12334/abstract
Abstract:
Substantial dolphin population mortality occurs during the first few
years of life (Mann and Watson-Capps 2005), with evidence of up to 50%
of newborns eliminated by age 4 (Stolen and Barlow 2003, Mannocci et al.
2012). In particular, the first year of life is the most cryptic period
for marine mammals, making it important to understand the underlying
mortality process in depth. An important source of information for
age-specific bottlenose dolphin mortality is the stranding data set of
the Florida Indian River Lagoon population (IRL), for which age at death
at completed years was reconstructed by counting dentinal growth layer
groups (Stolen and Barlow 2003). A theoretical life table was built on
such data by Stolen and Barlow (2003), by fitting Siler’s parametric
model (Siler 1979) to the age at death of 220 stranded individuals,
yielding useful information on the mortality pattern of this population.
In the Siler model the age-specific hazard of mortality is given by the
sum of three components: an exponentially declining one; a constant one;
and an exponentially increasing component, which is senescence-related.
The Siler model provides an excellent fit for many real populations,
supported by clear biological interpretations of its parameters. A
possible shortcoming of Stolen and Barlow’s approach is that the Siler
model was proposed for the IRL data on a priori grounds, i.e., as a
flexible model for a wide range of long-lived species, without
preliminary discussion on its full appropriateness for the data
considered. Indeed, retaining Stolen and Barlow’s (2003) baseline
assumption of a stationary population (i.e., a stable age distribution
with an intrinsic growth rate equal to zero; see Caswell 2001),
inspection of the resulting observed age-specific hazard of mortality in
the first few years of life does not robustly support the initial
exponential decline postulated by the Siler model: the mortality rate is
high (about 18%/yr) at age 0, it declines to about 10% at age 1, it
climbs back to more than 20% at age 2, and then stays constant in the
region of 12% in the two subsequent age groups, before declining quite
rapidly to a bottom rate (around 3.5%/yr on average) which remains
roughly constant between age 6 and 15. Similar patterns have been
observed also in other published data (Mannocci et al. 2012). The
above-described pattern in the data indicates as more plausible the
alternative hypothesis of a roughly constant hazard of mortality in the
first 5 yr of life, followed by a rather rapid decline before bottoming
out, prior to the final onset of the exponentially increasing phase. We
therefore propose, as an alternative to the Siler model, a new model,
termed ReLogit for ease, replacing the exponentially declining component
for early mortality with three-parameter “reversed-logistic” curve.
Please contact me by email ([email protected]) for the full text.
I will be glad to send you a pdf copy.
Kind regards
Alessia Rossi
--
PhD Alessia Rossi
Biology Department (Zoology - Animal Ecology) - Demography and
Conservation Laboratory
via A.Volta, 6 - 56126 Pisa, Italy
Cell: +393494743502. e-mail: [email protected] skype: musicale85
_______________________________________________
MARMAM mailing list
[email protected]
https://lists.uvic.ca/mailman/listinfo/marmam
