The Journal of Ecology editors are delighted to announce that Sandra Lavorel is our Eminent Ecologist for 2017. In recognition of her work, we have asked Sandra to select 10 of her most influential Journal of Ecology papers to make up a special virtual issue.
I am honored by the Editors’ decision to put together a virtual issue of some of my Journal of Ecology papers and grateful for this opportunity to synthesize their main contributions.
These papers reflect a great history of collaboration with inspiring individuals who have helped me put to the test some ideas that often started as dreams.
The 10 papers I have chosen illustrate four major themes from my work that has built up from explorations of plant trait responses to environmental change and land use, to using functional traits of plants and associated organisms for quantifying ecosystem services across landscapes.
My earliest Journal of Ecology papers were the outcome of some first explorations, along with Sue McIntyre and other international colleagues, aiming to transfer and develop the plant functional trait concept from its initial focus on modelling global change effects on vegetation, mostly at large (continental to global) scales, to community- and landscape-scale responses to biophysical gradients and land use. At the time, moving away from a floristic approach to plant community ecology to a ‘taxonomy-blind’ approach was considered rather bold, if not heretic.
These were the early days of functional ecology, where the original definition from Peter Calow in the first issue of the BES journal Functional Ecology (1987) focused mostly on the organismal level, with only a final opening to the community level. The transfer of the plant functional type concept from large-scale biogeography required some conceptual and methodological advances so as to account for fine-scale, continuous variation in plant traits within broad, life form-based functional groups. This challenge was particularly important for working within single biomes or vegetation types, in our case, grasslands.
Our first two papers were based on field measurements using sampling designs that accounted for underlying sources of vegetation variability (e.g. geology, landscape position, slope) in combination with grassland management factors, mainly grazing regimes and fertilization. Both studies, set in the context of semi-natural grasslands of eastern Australia, shared extensive floristic data sets. The first one was based on a landscape scale survey (McIntyre et al. 1995) and the second one based on a six-year grazing trial.
For both studies our first task was to select a priori relevant vegetative and reproductive traits given that there were no standard lists or methods at the time – something to which we later contributed. While in the first study we focused on standard trait values from floras or from measurements under optimal conditions following the Comparative Plant Ecology standards from Phil Grime and collaborators (Grime, 1988), in the second study we started implementing field methods that were later to become part of international standards.
Both studies also shared analytical challenges that put us at the forefront of applications of new methods in biostatistics for ecology. Our analyses were guided by the desire to identify discrete response groups to combinations of abiotic and management variables. Our main findings were; 1) analyses of functional responses must work within a priori groupings such as separating grasses from dicots, or Raunkiaer life forms. Amazingly this is an evident prior that has somewhat been forgotten since, including by ourselves in later research. 2) Vegetative traits need to be analysed separately from reproductive traits. This was evidence for the Leaf-Height-Seed axes of variation (Westoby, 1998) already highlighted by Phil Grime (1979), and what we recently named the ‘global spectrum of plant form and function’ (Diaz et al. 2016). 3) Not all species fit neatly into plant functional types, and plant trait variation and responses are rather continuous. This was a first step towards current practice which focuses on continuous responses of quantitative traits.
We also used fixed values of traits for each species thus ignoring intraspecific variation, in particular in response to the very factors whose influence we aimed to measure, for instance soil fertility. This was to be later addressed in approaches considering trait variation across plant populations as part of sampling designs (Garnier et al. 2007).
The structure of such intraspecific variation in plant traits according to landscape factors also became a later focus of my research. This variation was quantified across a mountain landscape during Cécile Albert’s PhD, focusing on plant height and leaf traits (Albert et al. 2010b). We confirmed first that intraspecific trait variation is substantial and needs to be accounted for; second that populations across a landscape were the dominant source of variation although relations to abiotic factors varied across traits; and third, in a companion paper published in Functional Ecology (Albert et al. 2011), that interspecific variation remains greater than intraspecific variation. Intraspecific trait variation and its implication for community assembly and ecosystem functioning are lively research topics today.
You can read Sandra’s virtual issue on the journal website, or view the individual papers below.