Current trends in trait-based ecology
In total, we identified 822 studies relevant to traits-based ecology from >200 journals published as early as 1978. Over 50% of the studies came from just 18 journals. The number of studies applying traits-based approaches to describe ecological pattern and process has grown exponentially over the past forty years, with a marked increase in the rate at which traits-based ecological studies have been published in the last decade (Figure 2). Most studies (59%) focused on the role of traits at a single level of environmental filtering, while only 3.5% of studies examined traits in the context of more than two levels of filtering. One study, Usseglio-Polatera (2000 - Freshwater Biology ), examined traits in the context of deriving environmental filtering from coarsest (i.e. abiotic environment matching) to finest (i.e. trophic interactions) filters. In total, 435 studies examined more than one type of trait (i.e. physiological, morphological, behavioural, or life history), 176 considered more than two trait types, and only 26 (<0.5%) studies considered traits that represented all four categories simultaneously.
Most of the research to date focuses on relating variation in morphological and life history features of vascular plants to abiotic environmental filtering processes within terrestrial ecosystems using observational techniques (Figures 3 and 4; 34% of all papers focused on plants, 30% on plant morphology); this result that is perhaps unsurprising given that traits-based research originated in plant ecology community (McGill et al. 2006). In contrast, traits-based investigations within marine and freshwater systems comprises just 30% of the studies combined, with a focus on observational studies of traits in the context of abiotic environmental filtering processes, primarily for fishes (Figures 3 and 4).
Trait-based research outside of plant-based research has overwhelmingly focused on size as the characteristic of interest (Figure 5, Table S3, Figs S3-S14). In fact, we identified 131 different metrics of size used across studies, representing a range of length, mass, or volume measures—including vegetative height and cone length (plants), snout-ventral length or instar size (animals), and biovolume (cells). Body size is the most common trait measured in studies of animal taxa, and also the most variable in terms of its definition and measurement, with metrics ranging from wet or dry mass, to inference from the length of the organism (involving multiple methods of measurement depending on the study and taxa under consideration). In contrast, specific leaf area (SLA; measured as leaf area per dry mass) is the most common trait used within plant trait research and is estimated via a single s standardized method across studies and plant taxa (Evans & Hughes 1961; Spence et al. 1973). In fact, SLA is the most common single trait examined in the studies reviewed here (127 occurrences over the 822 studies), with measures of body size a close second at 108 occurrences.
Beyond size, a plethora of morphological, behavioural, life history, and physiological features have been applied to trait-based biodiversity research (Figure 5). In total, we identified 2,561 unique traits within the literature we reviewed. While only 5% of these (131) represented aspects of organism size, size-based traits were used within 405 studies (49%) and each of these 131 traits were used by 3.7 separate papers on average, while non-size traits were each used only 1.7 times on average across the 822 studies. Accounting for variation in trait names (i.e. ‘time to maturity’ vs. ‘age at maturity’), and instances where a similar trait was assessed via multiple metrics (e.g. ‘trophic guild’ could be measured as a categorical variable with levels ‘resource’, ‘primary consumer’, ‘secondary consumer’, or ‘tertiary consumer’, while in a separate study ‘trophic group’ could be measured in levels of ‘primary producer’, ‘herbivore’, ‘omnivore’, or ‘carnivore’, both relating to ‘trophic role’) allowed us to attribute traits that confer information about the same process into conceptual groupings, revealing 196 ‘secondary’ trait classifications (Supplementary Material; Table S3). Of these, 14% are used in a single study while 34% are used in more than 10 studies. Of the 2,561 unique traits, morphological traits (Figure 5; n = 1,163) included aspects of organisms’ physical form (e.g. body size and shape, or the presence and form of dentition or spines) and biochemical composition (e.g. nitrogen or carbon content). Key behavioural traits (Figure 5, n = 626) include aspects of organisms’ activity (e.g. movement rates or nocturnality) and habitat use (e.g. vertical habitat position within forest canopies or water columns, range size or edge position). Life history traits (Figure 5; n = 585) describe growth, abundance, survival, and reproduction (including reproductive mode, timing, and frequency), while physiological traits (Figure 5; n = 187 traits) conferred information about organisms’ environmental habitat requirements (e.g. moisture or temperature tolerances), and resource acquisition (e.g. photosynthetic rate).
The diversity and identity of traits applied to ecological research depend on the environmental filter under investigation (Figure 1; Figure 5A-D, Figure S7&S8) and the ecosystem of interest (Figure 5I-K, Figure S9&S10). Interestingly, relatively narrower assemblages of traits are used in plant compared with animal research (Fig 6A,Fig S1A&S2A), with several potential explanations: (i) plant ecology has focused on traits for longer, and thus may have made relatively more progress in distilling a specific set of traits that represent key processes (versus animal trait research, which still needs to be distilled in this way), (ii) traits that represent key processes have more standardized measurement methods than traits for other taxonomic groups (i.e. plant size measured as SLA vs the variety of animal body size measurements), and (iii) differences in size of the trait space between taxonomic groups may simply reflect differences between sessile/mobile organisms; Compared with mobile taxa, sessile plants and fungi (represented in ‘other’; Figure 6A, Figure S12) have a relatively more narrow set of strategies for resource acquisition, defense, dispersal, and reproduction. In addition, perhaps also due to the longer history of traits approaches to plant ecology, research linking plant functional traits to their hypothesized ecological function is somewhat more developed, conferring a more complete understanding of the ecological processes that can be inferred from specific traits.
We also find that relatively narrow assemblages of traits have been applied in experiments, meta-analyses, reviews, and theoretical work compared with observational research (Figure 6B), the latter being the majority of studies in this review, and the likely starting place for first identifying and linking traits to important aspects of species distribution and interactions. Likewise, a relatively narrow assemblage of traits are applied within multi-ecosystem studies (i.e. those where findings are relevant broadly across ecosystem types; Figure 6C). However, the multi-ecosystem research reviewed here primarily represent theoretical models, which apply traits that can be estimated universally across taxa and systems such as body size and shape, and feeding mode, potentially explaining why the assemblage of traits applied in this body of work is more homogenous compared with marine, freshwater, or terrestrial focused-work (Figure 6C, Figure S14).