#883 Oecologia (2016) 181:39–53 DOI 10.1007/s00442-015-3430-3 HIGHLIGHTED STUDENT RESEARCH Do key dimensions of seed and seedling functional trait variation capture variation in recruitment probability? Julie E. Larson1,6 · Roger L. Sheley2 · Stuart P. Hardegree3 · Paul S. Doescher4 · Jeremy J. James5 Received: 22 November 2014 / Accepted: 16 August 2015 / Published online: 4 September 2015 © Springer-Verlag Berlin Heidelberg 2015 Abstract Seedling recruitment is a critical driver of and cluster analysis to identify major dimensions of trait population dynamics and community assembly, yet we variation and to isolate trait-based recruitment groups, know little about functional traits that define different respectively. PCA highlighted some links between seed recruitment strategies. For the first time, we examined and seedling traits, suggesting that relative growth rate whether trait relatedness across germination and seedling and root elongation rate are simultaneously but indepen- stages allows the identification of general recruitment dently associated with seed mass and initial root mass strategies which share core functional attributes and also (first axis), and with leaf dry matter content, specific leaf correspond to recruitment outcomes in applied settings. area, coleoptile tissue density and germination rate (sec- We measured six seed and eight seedling traits (lab- and ond axis). Third and fourth axes captured separate trade- field-collected, respectively) for 47 varieties of dryland offs between hydrothermal time and base water potential grasses and used principal component analysis (PCA) for germination, and between specific root length and root mass ratio, respectively. Cluster analysis separated six recruitment types along dimensions of germination Communicated by John M. Stark. and growth rates, but classifications did not correspond to patterns of germination, emergence or recruitment in the Highlighted student research statement Utilizing germination, field under either of two watering treatments. Thus, while belowground and leaf traits of 47 grasses, we show that trait variation during recruitment is multidimensional in its influences we have begun to identify major threads of functional var- on seed and seedling function. Furthermore, we detected an iation across seed and seedling stages, our understanding unexpectedly high degree of intraspecific variation in the nature of how this variation influences demographic processes— of trait interactions. However, weak links to recruitment highlight particularly germination and emergence—remains a key a key knowledge gap in trait-based ecology. This study provides a framework for further explorations of trait interactions and gap in functional ecology. influence during critical life history phases of seed germination and seedling recruitment. Keywords Emergence · Germination · Plant functional type · Roots · Survival Electronic supplementary material The online version of this article (doi:10.1007/s00442-015-3430-3) contains supplementary material, which is available to authorized users. * Julie E. Larson 4 College of Forestry, Oregon State University, Corvallis, OR [email protected] 97330, USA 5 Sierra Foothills Research and Extension Center, University 1 Environmental Sciences Graduate Program, Oregon State of California Division of Agriculture and Natural Resources, University, Corvallis, OR 97330, USA Browns Valley, CA 95918, USA 2 United States Department of Agriculture-Agricultural 6 Present Address: Schmid College of Science and Technology, Research Service, Burns, OR 97720, USA Chapman University, Orange, CA 92866, USA 3 United States Department of Agriculture-Agricultural Research Service, Boise, ID 83712, USA 1 3 40 Oecologia (2016) 181:39–53 Introduction explore and integrate trait variation within and across these key life stage transitions (but see Gardarin et al. Functional traits are used to describe ecological variation 2011; Thompson et al. 1996). Seed mass is most broadly among plant species and provide a mechanistic understanding studied among seed traits. Though it is typically com- of species distributions, community assembly, and how these pared to seedling function and survival (Moles and influence ecosystem function and services (Bello et al. 2010; Westoby 2006), evidence suggests that light and tem- Garnier and Navas 2012; McGill et al. 2006; Westoby and perature stratification requirements for germination may Wright 2006). This line of work has been instrumental in iden- be negatively and positively associated with seed mass, tifying trait tradeoffs that limit independent trait variation and respectively (Milberg et al. 2000; Pearson et al. 2002). uncovering potential ecological strategies favored under differ- Phenological patterns of germination and seminal root ent environmental conditions (e.g., Bernard-Verdier et al. 2012; growth could also influence recruitment probabilities and Cornwell and Ackerly 2009; Wright and Westoby 1999). While are tightly tied to requirements for minimum soil tem- there has been sustained headway in describing key dimen- perature, minimum water potential, and hydrothermal sions of trait variation in plants, efforts to integrate this under- time accumulation (Bradford 2002; Gummerson 1986; standing in a manner that facilitates practical advances in eco- Harris and Wilson 1970). Furthermore, the size, shape, system conservation and restoration have been slow to develop and density of coleoptiles or cotyledons could influence but represent one of the most critical areas of plant functional survival of emerging seedlings under harsh abiotic condi- trait research (Funk et al. 2008; Laughlin 2014). tions (Ganade and Westoby 1999; Gardarin et al. 2010). One key gap limiting application of trait-based frame- A major challenge in identifying general recruitment works is our lack of understanding of traits driving vari- strategies is that traits and ecological dimensions related ation in recruitment outcomes (Garnier and Navas 2012). to germination and emergence may be unrelated to those The vast majority of our current understanding of func- influencing seedling resource capture (Grime et al. 1997; tional trait spectra and strategies is centered around plant Leishman and Westoby 1992). However, early work by tissue economics and the influence of these traits on Grime et al. (1981) and Shipley et al. (1989) documents at resource capture, conservation and growth (e.g., Lambers least one cohesive spectrum linking large-seeded species and Poorter 1992; Reich 2014). Along a tissue economics to both slow germination rates and slow seedling growth spectrum, species constructing thinner or less dense leaves rates across several life forms. Ultimately, if germination or roots create more absorptive surface area per unit of bio- and emergence are key processes influencing recruitment mass, resulting in more rapid return on tissue investments outcomes, it is essential to identify and incorporate related and higher relative growth rates (RGRs) (Reich 2014; Ship- traits into a broader understanding of ecological strategies ley 2006). At the opposite end, species constructing thicker and variation among plants. or denser leaves and roots tend to have slower growth but Whether trait variation across multiple seed and seedling greater tissue longevity and stress tolerance, which could stages can be used to identify general ecological strategies promote survival (Poorter and Bongers 2006; Ryser 1996). which ultimately predict recruitment outcomes is unknown. Species may also vary in the plastic response of these It has been suggested that functional groups of species attributes to changing environmental conditions, which sharing key trait attributes may have common responses to could be key in differentiating performance among coex- environmental stressors and ecosystem effects (Boutin and isting species (Valladares et al. 2007). Given the potential Keddy 1993; Lavorel et al. 1997). These trait-based groups, for early seedling survival to influence recruitment rates or functional types, should provide a simpler, generalizable (Gómez-Aparicio 2008), these seedling functional traits alternative to species-based models and management while could provide important insight into demographic patterns. capturing community responses better than groups based However, there is also growing evidence that a bulk of simply on life form or life history. However, it is unclear plant mortality can occur prior to vegetative stages, during whether functional types could effectively predict demo- germination and emergence (James et al. 2011; Leishman graphic responses, which are difficult and time-consuming and Westoby 1994; Sharitz and McCormick 1973). In order to measure. to understand the implications of trait variation for popula- The first two objectives of this study were to look for tions, communities and ecosystems, it is critical to expand major axes of variation and tradeoffs among seed and our understanding beyond plant growth and resource cap- seedling functional traits of dryland grasses, and to use ture to include key traits influencing transition probabilities these trait data to identify recruitment types (i.e., func- across early life stages. tionally similar species or varieties identified via cluster A wide array of traits may influence germination and analysis) among candidate restoration grasses. We then emergence processes, but few studies have attempted to asked whether these recruitment types captured variation 1 3 Oecologia