A trait is any aspect of an organism that defines it with respect to a concept. A trait may be color, when examining heat absorption; accent, in the case of humans when attempting to approximate nationality; specific leaf area (i.e., area to dry mass ratio of a leaf) when interested in investment in photosynthetic machinery over carbohydrate revenue; metabolic rate, when attempting to quantify the pace of living of an organism.
The concept trait, which was already used by Darwin (1859), was initially used almost exclusively as a proxy to individual performance. Examples include the matching between seed size and the size and shape of beaks in Galapagos finches. However, the advent and impressive development of community ecology has expanded the term trait away from its original usage. And it must be emphasized, there is nothing wrong with that a priori; word meanings evolve all the time.
Certainly, a trait-based approach in this discipline has resulted in the discovery of important findings such as the detection of niche-based community assembly processes (Kraft et al. 2015) or the identification of key plant traits driving ecosystem productivity (Garnier et al. 2004). In summary, a key advance of trait-based ecology is the ability to scale up processes, and one may not always need demography for that…
One can, however, find oneself (we do!!) attending conferences where the adjective “functional” is prefixed to the noun “trait” perhaps a bit too freely. Why is this an issue, potentially? Well, functional means that the trait itself either has a specific function on shaping organismal fitness (e.g., McGill et al. 2006; Violle et al. 2007), or it can be used as a proxy to it, perhaps through their modulation of fitness components.
What if one is measuring a trait that simply does not approximate a fitness component or the performance of an organism? It should not be a big deal to examine traits for ecosystem functioning among other questions…though we would argue that a system functions in a specific manner due to the emergent properties (e.g., turnover) of the demographic dynamics of the species that compose the community.
We contend that in most cases, only a careful and quantitative examination of the relationships between traits and fitness components can help scientists benefit from a “trait approach” to their specific questions. Equally important, is the question of how the enormous interspecific variation in plant traits that has been found relate to key processes at the organismal level, such as competitive ability (Albert et al. 2011; Violle et al. 2012).
To contribute to such an urgent and timely need, at the 2017 annual meeting of the Ecological Society of America in Portland, we are organizing a symposium titled “Towards a unified framework for functional traits and life history strategies in plants”.
Due to the acquisition of large volumes of data on traits and vital rates (e.g., survival and growth), during the last decade researchers have implemented global analyses on the covariation of traits on the one hand, and vital rates, on the other. As a result of such efforts, important discoveries have been made, including the leaf-economics spectrum (Wright et al. 2004), or the wood-economics spectrum (Chave 2009), among others. Despite the independence in approaches and data used in these global analyses of functional traits and of vital rates, remarkable similarities in how plant life is structured are starting to emerge.
Recent findings (see fig. 1) report a couple of axes explaining most of the variation in the vast plant biodiversity repertoire, one related to plant size, and the other to individual/organ turnover (Díaz et al. 2016; Salguero-Gómez et al. 2017). Moreover, global patterns for the unification of traits and vital rates are starting to pop up too (Salguero-Gómez 2015). The invited speakers in this symposium, leaders in the fields of plant traits and demography, will provide a synthetic framework to integrate plant shape, function and strategies using big data.
Our own contributions to this field have led us to build bridges between demographers (R. Salguero-Gómez) and functional ecologists (C. Violle), and to start evaluating single-trait to single-vital rate correlations in the plant kingdom. For instance, using data from TRY (Kattge et al. 2011) and COMPADRE (Salguero-Gómez et al. 2015), it was shown that not all traits commonly regarded as functional in the plant kingdom are actually so (Fig. 2), and that their functionality is fitness-component specific (Adler et al. 2014).
Also, leaf nitrogen content, the main driver of the leaf-economics spectrum (Wright et al. 2004), turns out not to have much of a predictive power in the relative importance of survival or growth of individuals in a population in over 200 plant species worldwide, but it was positively correlated with the effect of fecundity. The opposite was true for seed mass, which was not correlated with the elasticity of population growth rate to growth or fecundity, but was positively correlated with the survival elasticity.
Even in the models where the strongest correlations were found, the coefficients of correlation were low. We have suspected for quite some time now that this is so because plant populations ecologists and functional ecologists have not historically worked together in the field. Consequently, big data correlative exercises like this may inevitably blur exiting underlying controls of anatomy onto demography due to the great geographic distance (and thus difference in microhabitat conditions) between the collection point of traits and vital rates. In an effort to build predictive models of life history strategies and population performance using anatomic and physiological traits as proxies, the StrateGo Network (Fig 3) was recently launched.
StrateGo is liaising with field researchers worldwide to test the following hypotheses: (1) shifts in functional trait values precede changes in population dynamics, and thus traits are good proxies to demographic projections, (2) not all traits are functional; trait functionality depends on the life history of the species along the fast-slow continuum and reproductive strategies continuum (Salguero-Gómez et al. 2016), and (3) trait interactions of orthogonal axes such as wood density and leaf Nitrogen describe emerging properties result of trade-offs in resource allocation that single traits cannot.
StrateGo, a globally distributed network, sits on top of a much older network: COMPADRE. The COMPADRE Network is constituted by all the researchers who have contributed plant population dynamics data in the shape of matrix population models (Caswell 2001) or integral projection models (Easterling et al. 2000).
Researchers who have collected or are currently collecting demographic data are being encouraged by StrateGo to collect a set of traits in the same, comparable manner. However, StrateGo is also open to researchers who have collected, are collecting or plan to collect functional trait data and would like to evaluate plant population dynamics too at the same study site(s). Incidentally, wanna join? Contact here.
Also cooking in the backstage, we are planning a special feature, together with Olivier Gimenez and Dylan Childs, for the British Ecological Society. Manuscripts will be submitted to Journal of Ecology, Functional Ecology and Journal of Animal Ecology. The special feature, inspired on the aforementioned ESA symposium, will aim to provide novel frameworks to link traits and life history strategies within plants, animals and microbes.
Dr Rob Salguero-Gómez (University of Oxford, UK, and Associate Editor of Journal of Ecology)
Dr Cyrille Violle (CEFE CNRS Senior Researcher, France)
2 thoughts on “When are traits functional?”
Pingback: Tuesday taster: 08/08/2017 – Functional Ecologists
Pingback: Fewer results, more open questions: Insights from ESA by Ignasi Bartomeus | Journal of Ecology Blog