| |
What
is TraitNet?>>>
TraitNet will coordinate trait-based evolutionary and ecological
research. Traits are biological
properties of species that influence individual fitness
and govern how species interact with their
biotic and abiotic environment. Traits are used across a
broad spectrum of disciplines, including
niche theory, community assembly, metabolic ecological theory,
phylogenetics, conservation,
and ecological stoichiometry. While each discipline has
developed its own operational
definitions, protocols, and databases, there is little coordination
across disciplines. TraitNet will
advance syntheses and analyses by coordinating integration
among disciplines. Its five primary
goals are (1) identify core hypotheses in trait-based research,
(2) identify critical data gaps, (3)
coordinate the standardization of collection and curation
of trait data, (4) assemble a database to
address core hypotheses, and (5) facilitate the development
of cross-disciplinary computational
tools for merging, disseminating, and sharing trait data.
This network, made up of 5 working
groups, will use electronic collaboratories, workshops,
training seminars, and electronic and
journal publication to address specific cross-disciplinary
hypotheses, but will be structured more
broadly by the goal of prototyping a universal trait database
entitled TraitBank.
In the face of global change, species traits take on additional
importance as
tools for predicting the ecological and evolutionary response
of species, communities and
ecosystems to changing weather patterns, exotic species,
and land use change. Additionally,
there is naturally significant overlap among trait-based
research disciplines so coordination will
substantially minimize redundant collection efforts and
maximize scientific productivity.
TraitNet’s website will be widely accessible to the
scientific community and the public, while
protecting the intellectual property rights of individual
investigators. TraitNet will further
diversity in science by maintaining a strong gender balance,
including under-represented groups,
and balancing participants among students, postdocs, and
senior and junior researchers. Its core
participants are listed in the table below.
Core
Hypotheses>>>
At the first TraitNet workshop, working groups will explore
and select two to three Core Hypotheses to serve as working
models for TraitNet. Here we provide three of the many possible
examples.
- Dimensionality of life-history trade-offs. While an
endless number of traits can be measured on individuals
and species, many traits are highly correlated with one
another, and it has been suggested that relatively few
trade-off axes can explain the majority of variation in
plant form and function (Grime 1979, Coley et al. 1985,
Charnov 1997, Reich et al. 1999, Hubbell 2001, Westoby
et al. 2002). How are species life histories constrained
by these fundamental trade-offs, how many axes of differentiation
exist, and how does the extent of these trade-offs vary
across environmental gradients and among biomes? These
key questions require data on multiple traits, collected
from multiple species, from multiple sites, and standardized
where different protocols were used.
- Mechanisms of exotic species invasions. The success
of invasive species has been often attributed to an escape
from natural enemies, whereby one would predict successful
invaders to have 'better' traits than the native species
they displace, such as greater height, lower R*, or lower
construction costs (Nagel and Griffin 2001, Bunker 2004,
Gaudet and Keddy 1998, Miller and Werner 1987, Seabloom
et al. 2003). Alternatively, the success of some invaders
has been attributed to novel traits, such as nitrogen
fixation (Vitousek and Walker 1989) or allelopathic effects
(Bais et al. 2003), that allow them dominate new habitats.
While both mechanisms certainly play a strong role, the
relative importance of each in driving species invasions
is not clear. An effective test would require species
trait data on plant invader species, on the native species
they may displace, on palatability to native herbivores,
and data on traits of potential natural enemies such as
body size, diet, and growth rates.
- Predicting species, community and ecosystem responses
to global change. Predicting the response of species to
climate change, pollution, and land use change is a key
challenge to ecologists. These predictions could be developed
by correlating species traits with either observed responses
to global drivers or across natural environmental gradients.
In either case trait data from a wide variety of species,
across multiple trophic levels, from a variety of habitats
would be required. Similarly, predicting the effects of
these global drivers on ecosystem function will require
additional trait data that mechanistically link species
with their per capita effects on ecosystem functioning.
(Etterson and Shaw 2001, Solan et al. 2004, Bunker et
al. 2005)
Scope
of TraitNet>>>
TraitNet aims to bring together species trait data from
a variety of taxa across different trophic levels and from
a variety of habitats and locations to address specific
interdisciplinary hypotheses (see examples above). Our initial
coverage will be greatest among terrestrial plants because
several focused plant trait networks are well developed
and will serve as useful starting points (Table 2). The
interdisciplinary hypotheses chosen for study by the network
of participants will likely require traits of herbivores,
predators, detritivores, and other trophic groups as well.
Additionally, TraitNet will not limit itself to terrestrial
systems and may select aquatic ecosystems or transition
habitats such as wetlands. To that end, we have assembled
a group of core participants that is weighted towards terrestrial
plant ecologists due to current trends in the literature
and also includes researchers who specialize in insect,
mammalian, microbial, aquatic and disease ecology. These
core participants will actively recruit additional investigators
within their respective areas of expertise. TraitNet participation
is expected to grow substantially once established and we
identify additional researchers who focus on other habitats,
taxonomic groups, and trophic groups. Table three describes
the diversity of taxonomic focus and expertise of core and
general TraitNet participants.
Data
integration challenges>>>
- Intellectual property rights - Intellectual property
rights are a critical issue for any research network and
even more so when data is aggregated from multiple sources.
Trait-based research progresses best when data sharing
is maximal, but currently the sharing of raw data is not
common except within groups. Workshops, collaboratories,
training sessions, and the TraitNet website will provide
a forum for discussion of the many issues surrounding
intellectual property rights and how they would affect
database tools, resources, and the design and implementation
of TraitBank in the future.
- Taxonomic standardization - Definitions of biological
taxa change with taxonomic revisions over time. For instance,
a single species may be split by one revision into several
species, and then lumped back into a single species in
subsequent revisions. A trait value measured on the lumped
species cannot be assigned to any one of the split species,
and a trait value measured on one of the split species
cannot be assumed to represent the entirety of the lumped
species. In addition, species are often cited with only
the name authority, but not the underlying taxon concept
reference. For these reasons, taxonomic names by themselves
cannot be considered a unique index for TraitNet datasets.
This obstacle applies to all data that are specific to
individual species, such as GenBank, VegBank, etc.
- Changes in trait collection protocols introduce
challenges that are similar to those introduced by taxonomic
revisions. A trait concept may remain fixed, but the protocol
used to quantify the trait may change as new protocols
are introduced. A trait database must be able to incorporate
revised trait collection protocols as they are developed
to ensure that data produced though all protocols for
a given trait concept are quantitatively comparable. For
example, wood density is the trait concept of mass per
unit volume. However the protocol to collect wood density
varies. Wood mass may be measured on oven-dry wood samples
or on air-dried samples with 12-15% moisture content.
Both metrics quantify wood density, but data from air-dried
samples must be corrected to account for the moisture
content. TraitNet will define trait concepts and associated
trait collection protocols. Each trait protocol for a
given trait concept must be quantitatively comparable
(Figure 3).
TraitNet will build on current ecoinformatic efforts to
address these issues. Ecological Metadata Language (EML)
has been developed by NCEAS' Knowledge Network for Biocomplexity
project and is widely considered the standard for documenting
metadata for ecological datasets. The SEEK project has extended
and formalized critical aspects of EML in the Observation
Ontology (OBOE), a formal model of scientific observations
that includes trait measurements. Thus TraitNet will use
and extend EML to specify the Taxon Concepts, Trait Concepts,
and associated environmental data (see figure 3 below).
Eventually, these trait concepts will be included in SEEK's
formal ontologies such as OBOE.
Please send us
your questions, comments and input to add to this page.
|
|
