Summary of the proceedings of the TraitNet's first workshop help at Columbia University, New York City, December 7-9 2007
The goal of the workshop was to identify the key needs of the community both in terms of informatics resources and trait-based scientific knowledge. The workshop was attended by 30 ecologists, evolutionary biologists, and information scientists from a diverse array of institutions.
A set of core needs for community development were identified as well as working groups to address these core needs, which were:
Ontology working group
This working group is tasked with building an ontology that will define a common language for plant functional traits. An ontology is critical both for ensuring that we use is common language and also for developing a cyber infrastructure based on relationships and definitions embedded in the ontology. The ontology will facilitate communication between the domain scientists (i.e., trait data owners and users) and the information scientists. At the meeting the group explored the need for ontologies by TraitNet and began to develop a plant trait ontology that will build on efforts in related fields such as the Arabidopsis and maize plant biology research communities. Integration with these additional ontologies is both an opportunity and a challenge. For TraitNet to be integrated in the larger bioinformatics community our ontology must extend these established ontologies, yet at the same time we must be sure that this does not impose critical limitations on the practicality of our ontologies.
Data structure working group
This working group is tasked with defining internal data structures required for effective trait data and annotation, as well as identifying external data sources that will commonly be used in conjunction with species trait data. Key internal data include trait observations, environmental metadata, and taxonomic data. Key external data include: geospatial data, digital elevation models, soils data, climate data, land use data.
Theory working group
This working group is tasked with identifying theoretical applications of the TraitNet database such that the database can be designed with these applications in mind.
Abstract: The predictive power of ecological evolutionary theories is linked to the way that they are developed; the best approach promotes the application of specific theories across a hierarchy of biological organization. We argue that this approach begins with a model of individual performance (fitness) that is rooted in the phenotype (traits) (and ultimately the genotype). By considering the performance of individuals as a function of their traits and the dependencies of these traits on the environment and community structure individual performance can be scaled up to predict population and community dynamics. This approach can also link individuals to ecosystem processes like NPP when resource use and uptake traits are considered. We find that most prominent community theories are ameliorable to being ‘retrofit’ with trait-based models of individual performance, and that thereafter they may be applied across a broader range of inference scales. Crucial to the development and evaluation of a trait-based theoretical approach is the collection of a wide variety of data such as traits, environmental trait plasticity, individual phenotypic states such as size, and population, community, and ecosystem states and processes. We discuss the development of the TraitBank (currently underway) in light of testing trait-based theory macroecologically. We suggest that TraitBank be flexible enough to include not only a wide variety of trait data but also the associated ecological data required to test cross-scale trait-based theory. Furthermore, we identify stochastic and function valued traits often ignored trait categories that must me measured and incorporated in trait databases. Stochastic traits are those that influence discrete or binary outcomes such as germination while function valued traits are highly dynamic traits related to resource use and uptake such as the relationship between photosynthesis and light-levels.
Working group on the incorporation of external trait databases
This working group focused on strategies to incorporate pre-existing trait databases. Key challenges include adapting TraitNet and TraitBank to handle the more varied data structures that have been previously implemented in external databases such as Leda, Bioflor, and TRY (formerly the PFT-FTI). In order to accommodate a wide variety of databases, several strategies may be implemented. These include a) direct incorporation of the database into TraitNet with coordinated updates to both the source and to TraitNet, b) a confederated database whereby TraitNet executes searches across its own internal and in parallel across the associated external databases, and c) simply providing a link to additional databases rather than direct search capabilities. Strategy (c) would be implemented only for databases that have fundamentally incompatible data structures or where intellectual property considerations prevent direct searches.
Working group on the integration of TraitNet with synergistic efforts in evolutionary biology
This working group worked to identify related efforts in evolutionary biology and outline opportunities for building collaborations between these groups.
Participants of TraitNet1