Diverse ecological functions and the convergent evolution of grass awns

Abstract

The awn of grasses is a long, conspicuous outgrowth of the floral bracts in a grass spikelet. It is known to impact agricultural yield, but we know little about its broader ecological function, nor the selective forces that lead to its evolution. Grass awns are phenotypically diverse across the extant ~12,000 species of Poaceae. Awns have been lost and gained repeatedly over evolutionary time, between and within lineages, suggesting that they could be under selection and might provide adaptive benefit in some environments. Despite the phylogenetic context, we know of no studies that have tested whether the origin of awns correlates with putative selective forces on their form and function. Presence or absence of awns is not plastic; rather, heritability is high. The awns of grasses often are suggested as adaptations for dispersal, and most experimental work has been aimed at testing this hypothesis. Proposed dispersal functions include soil burial, epizoochory, and aerial orientation. Awns may also protect the seed from drought, herbivores, or fire by helping it become buried in soil. We do not fully understand the fitness or nutrient costs of awn production, but in some species awns function in photosynthesis, providing carbon to the seed. Here we show that awns likely provide an adaptive advantage, but argue that studies on awn function have lacked critical phylogenetic information to demonstrate adaptive convergent evolution, are taxonomically biased, and often lack clear alternative hypotheses.

Keywords: Poaceae; adaptation; agriculture; evolutionary ecology; exaptation; florivory; geniculate; grass ecology; herbivore defence; herbivory prevention; spikelet.

Figure 1

Simplified diagram of a grass spikelet to illustrate the position of nonfloral structures. The four example spikelets show positions where awns commonly develop and the possible variations in number and form. Awns can develop from lemma or glumes or more rarely paleas. The type of awn can be mixed within a single spikelet, e.g., bent and straight.

Figure 2

Presence or absence of awns in Poaceae regardless of developmental origin (e.g., glume or lemma) optimized on a phylogeny from Saarela et al. (2018). Outer names indicate subfamily classification; inner names indicate tribe level classifications. Dark green: awned species or genera; blue: no awn. An awn is defined as present if it is ≥1 mm long. Full species names are in Appendix S2.

Figure 3

Morphological origin of awns on the spikelet, optimized on a phylogeny from Saarela et al. (2018). Outer names indicate subfamily classification; inner names indicate tribe level classifications. Light blue: species or genera without awns; dark green: lemma awn; yellow: glume awn, pink: palea awn, orange: both lemma and glume awns. Arrows specify taxa in which glume awns appear to have evolved independently of any other awn.

Figure 4

Type of awns optimized on a phylogeny from Saarela et al. (2018). Outer names indicate subfamily classification; inner names indicate tribe level classifications. Light blue: species or genera without awns, orange: geniculate (twisted and bent) awns, dark green; straight awns, yellow: mixed geniculate and straight awns, pink: others (branched or coiled). Arrows specify places where geniculate awns have evolved in lineages of straight awns.