Lecturer: Bas Kooijman
Affiliation: Dept Theoretical Biology, Vrije Universiteit, Amsterdam
Seminar: Univ. Oslo, CoE Center for Ecological and Evolutionary Synthesis 
   (CEES).
Date: 2011/02/10
Title: Add my pet, a data and parameter collection 
   revealing adaptive trends in evolutionary energetics
Extended abstract:

The standard DEB model is the simplest in the family of models implied
by Dynamic Energy Budget (DEB) theory for metabolic organisation
(Kooijman 2010). It deals with a single food source, a single reserve
and a single structure, while isomorphy is assumed. The standard model
is appropriate for most animal species. The covariation method (Lika
et al 2011, 2011a) was developed to estimate all parameters of the
standard DEB model from simple data on the energetics of a
species. The method exploits the covariation of parameter values as
implied by DEB theory on the basis of physical and chemical
principles. This lecture is about our experience with the covariation
method.

At this moment (2012/01/19), data has been collected for 109 species
from 10 animal phyla, including all 13 chordate classes: the "add
my pet" collection. The collection has been set up as part of the
biannual DEB tele-courses. The model and the data concern the full
life cycle of the species (embryo, juvenile, adult stages) and
includes the ageing module of DEB theory as well as an extension of
the standard model that allows for the frequently observed metabolic
acceleration (Kooijman et al 2011). This acceleration starts at (or
soon after) birth and ceases at metamorphosis, a life-history event
that can occur in the juvenile period.

The data has been classified in zero- and uni-variate data. All
together more than 100 different data-types are presently involved,
relating to feeding, digestion, respiration, growth, reproduction,
survival (life span), life history events (hatching, birth,
metamorphosis, puberty, death), length-volume-weight relationships,
chemical composition, effects of temperature, food availability and
changes in food availability. These data-types has been schematised
and survey has been composed that shows what data for what species has
been used to estimate parameters. The completeness level of the data
has been marked for each species (ranging from zero, where only
information on maximum body size is available, to ten, where data on
all aspects of the energy budget during the life cycle is available);
typical completeness levels in the collection are around 2.5 till
3.5. The goodness of fit has also been marked for each species
(ranging from minus infinity to ten, where all predictions match the
data in all decimals); typical fit levels in the collection range from
7.5 till 9.3. So the fit is typically very good indeed.

Each species has three files: 
 - the mydata-file sets the data (with references), 
   estimates the parameters and presents the results
 - the predict-file codes for the predictions for the data
   and is used by the parameter estimation routines 
 - the pars-file sets the parameter values 
   and computes a list of over 100 implied properties 
   on many aspects of energetics and life history, 
   including population dynamics.
The predict-file can account for specific conditions that apply to
particular data as specified in the mydata-file; the pars-file does
not account for these details and is meant for comparative
purposes. Templates for these files are available (mydata_my_pet,
predict_my_pet, pars_my_pet) and give explanations for each of the
steps. These templates can be copied and edited, while the existing
collection offers many examples for extensions to the template files.

All files make intensive use of software package DEBtool (which has
over 1000 functions and a manual), and the results are collected in
the file Species.xls. All files are freely downloadable offering full
transparency from data to conclusions; all computations can easily be
re-done by the user and will be re-done when more data become
available for a species. Full documentation of the collection, the
method and the theory is available.

The add_my_pet collection of metabolic data and parameters is by far
the largest of this kind that presently exists, illustrating the
general applicability of the theory. The lecture will discuss the
set-up, a few case studies and some emerging patterns in parameter
values. Some of these are remarkable and new in the context of the
existing literature because of the interpretation of respiration data.

The existing literature treats respiration sometimes as a quantifier
for metabolic activity, sometimes as metabolic losses (maintenance),
with, possibly, contribution by specific dynamic action (heat
increment of feeding). Dioxygen consumption is taken equivalent to
carbon dioxide or heat production. This interpretation is basic to the
Scope for Growth method and, what I call, Static Energy Budget (SEB)
models. These models include almost all work in ecology on
energetics. It is at the root of a lot of confusion on explanations
for why respiration scales (approximately) with body mass to the power
3/4. None of these approaches account for metabolic memory and, as a
consequence, typically ignore the embryo stage (where this memory is
essential).  Embryos decrease in mass while increasing respiration; a
well-known fact that is frequently ignored. DEBs differ fundamentally
from SEBs by the treatment overheads and (metabolic) memory.

Dynamic Budget Theory has a very different perspective on respiration. 
It classifies metabolic transformation in three underlying processes: 
- assimilation (the conversion of food to reserve), 
- growth (the conversion of reserve to structure) and 
- dissipation (the conversion of reserve to excretion products only). 
Under aerobic conditions dioxygen serves as non-limiting
substrate in these transformations; its consumption follows from the
principle of conservation of chemical elements. Dioxygen, carbon
dioxide and heat fluxes are not assumed to be proportional to each
other and differ substantially from proportionality in non-animal
taxa, as is well-known in microbiology. An implication of the DEB
interpretation of respiration is that the overhead costs of growth
contribute to dioxygen consumption, making the method of Scope for
Growth invalid. DEB theory views eggs as wrapped reserve that is
excreted, with hardly any structure. An implication is that
reproduction might represent a substantial energy drain to the
individual, but hardly contributes to dioxygen consumption (no
chemical transformation is involved). This demonstrates that
respiration cannot be viewed as a quantifier for metabolic activity.

DEB theory explains the observed scaling of respiration with mass as
the result of a decreasing contribution from growth for intra-specific
comparisons (where the ontogeny of an individual is followed) or as a
result of the fact that big-bodied species require relatively more
reserve (metabolic memory) for inter-specific comparisons. Reserve
does not require maintenance in DEB theory, only structure does and
does it proportional to the mass of structure. The goodness of fit
with empirical data is very good.

The covariation method can decompose respiration in its various
contributions, somatic maintenance being only one of them; other
contributions comprise maturity maintenance, maturation and overheads
of assimilation, growth and reproduction. My lecture will discuss some
observed patterns in somatic maintenance costs among animal taxa and
the observed distribution of the allocation fraction (kappa) of
mobilised reserve to the soma. It turns out that the typical fraction
is much larger than the fraction that maximises reproductive
output. This empirical observation should have a big impact on
theories in evolutionary ecology.

More information about research on DEB theory can be found at
  http://www.bio.vu.nl/thb/deb/

The add_my_pet collection and DEBtool can be found at the electronic DEBlab
  http://www.bio.vu.nl/thb/deb/deblab/

References:

Lika, K. and Kearney, M. R. and Freitas, V. and Veer, H. W. van der and 
  Meer, J. van der and Wijsman, J. W. M. and Pecquerie, L. and 
  Kooijman, S. A. L. M.
  The `covariation method' for estimating the parameters of 
  the standard  Dynamic Energy Budget model I: 
  philosophy and approach.
  J. Sea Res. (2011) 66, to appear
  http://www.bio.vu.nl/thb/research/bib/LikaKear2011.html

Lika, K. and Kearney, M. R. and Kooijman, S. A. L. M.
  The `covariation method' for estimating the parameters of 
  the standard Dynamic Energy Budget model II: 
  properties and preliminary patterns.
  J. Sea Res. (2011) 66, to appear
  http://www.bio.vu.nl/thb/research/bib/LikaKear2011a.html

Kooijman, S. A. L. M. and Pecquerie, L. and Augustine, S. and Jusup, M.
  Scenarios for acceleration in fish development and the role of metamorphosis.
  J. Sea Res. (2011) 66, to appear
  http://www.bio.vu.nl/thb/research/bib/KooyPecq2011.html

Kooijman, S. A. L. M. 2010
  Dynamic Energy Budget theory for metabolic organisation.
  Cambridge University Press, third edition
  http://www.bio.vu.nl/thb/research/bib/Kooy2010.html