Elymus diversiglumis
 
Variable-Glumed Wild-Rye

Floret Close-Up, Abaxial View

Rendek Elm Forest at Confluence of Smoking Tent Creek and Red Deer River,
East of Hudson Bay
30-July-2022

Note that most of the lemma awn has been removed from this specimen.

I believe this specimen is a variety of Elymus diversiglumis that overlaps with E. canadensis as described by Anna Leighton in the publication listed below:

"Some of our specimens overlap with E. canadensis in glume characteristics.  In these specimens, the glumes are setaceous-subsetaceous, 0.5 to 0.7 mm wide and sometimes 3-nerved, and ± equal in length at most nodes.  The 3 features, which, when taken together, are most useful to distinguish these intermediates from E. canadensis are wide leaves, short glumes and occurrence in moist wooded habitat."

This specimen was collected in a moist wooded floodplain of Smoking Tent Creek.

The following items are taken from keys in Flora of Saskatchewan, Fascicle 4, Grasses of Saskatchewan by Anna L. Leighton and Vernon L. Harms.  Family Poaceae is first divided into tribes, then the tribes are divided into genera, and the genera divided into species.  However, there are a number of tribes that are very difficult to distinguish morphologically.  These are grouped into a large, artificial tribe I call "Multitribe".  Multitribe is then divided into groups, and each group is then divided into genera.  The answers are in the order you would normally work through the keys.

Triticeae: Answers to key questions leading to this tribe. 
Mature inflorescence, if breaking into units, then the units not as below; NOT [Mature inflorescence breaking into spikelet units consisting of a sessile fertile spikelet, a hairy pedicel with or without a sterile spikelet at tip, and a hairy rachis joint, all arising at the same point (a node) in specialized panicle branches called rames]
Spikelets not as below; sterile florets if present, either located distal to the fertile floret(s) on the rachilla or paired and attached at the base of a single fertile floret, not paired with the upper glume as below; lemma and palea variously textured, enclosing the flower or not; disarticulation usually above the glumes; NOT [Spikelets usually dorsally compressed, appearing 1-flowered but containing 1 fertile floret and 1 sterile floret, the latter attached to the base of fertile floret opposite the upper glume, resembling the upper glume, and together with the upper glume enveloping the fertile floret; lower glumes minute (sometimes absent) to 3/4 as long as upper glumes and typically wrapping most of the way around the pedicel at base; fertile floret seed-like with chartaceous-indurate lemma and palea enclosing flower and fruit; disarticulation below the glumes with rare exceptions]
Spikelets 1 to many-flowered, subtended by a pair of glumes (only 1 on lateral spikelets in Lolium); palea margins enclosed or not; plants of dry or wet habitats; NOT [Spikelets 1-flowered, lacking glumes; margins of the palea tightly enclosed by the lemma margins on female or perfect florets; plants of wetlands, often emergent aquatic]
Inflorescence a terminal spike with sessile or subsessile spikelets attached broadside at nodes on opposite sides of the rachis; NOT [Inflorescence not as above; if a terminal spike, then the lateral spikelets attached edgewise to the rachis with inner (upper) glume wanting (as in Lolium)]

Elymus: Answers to key questions leading to this genus. Not all the answers apply to all members of Elymus, but they all do apply to a subset containing E. diversiglumis).
Spikelets 2 or more per node, occasionally 1 at some nodes; NOT [Spikelets 1 per node, occasionally 2 at some nodes]
Spikelets 2-7 per node with 1-7 (9) florets per spikelet; if spikelets 3 per node, then some spikelets with more than 1 floret; NOT [Spikelets 3 per node with 1 floret per spikelet; florets in central spikelet always fertile, those in the 2 lateral spikelets sterile and greatly reduced in plants that disarticulate at rachis nodes but may be fertile in plants with a continuous rachis]
Lemmas awned, the awns usually (5) 8-120 mm long, shorter (to 3 (5) mm) in plants with glumes distinctly bowed out (as in E. curvatus); NOT [Lemmas unawned or with awns up to 4 mm long; glumes not distinctly bowed out at base]

 

Diversiglumis: Answers to key questions leading to this species. 
Spikelets 2 or more per node (occasionally 1 at some nodes); glumes setaceous to linear-lanceolate; NOT [Spikelets 1 per node (occasionally 2 at some nodes); glumes lanceolate or wider (except E. glaucus which may have linear-lanceolate glumes)]
Rachis not disarticulating at the nodes at maturity; glume awns if long and outcurving then not forking near base; NOT [Rachis disarticulating at the nodes at maturity; glume awns 35-85 mm long, flexuous or outcurving, often forking into 2-3 divisions near the base]
Glumes usually setaceous or linear-lanceolate, with firm margins; glume bases thickened to terete, not overlapping, lacking apparent nerves; NOT [Glumes linear-lanceolate to lanceolate, with hyaline margins; glume bases flat, often overlapping, nerved to near base]
Glumes setaceous, 0.2-0.5 mm wide, 0-1-nerved distally, (1) 2-15 (20) mm long (entire length to tip), 1 or both less than 12 mm long, often distinctly unequal in length with 1 glume of a pair usually several mm shorter than the other, and occasionally obsolete; NOT [Glumes subsetaceous to linear-lanceolate, 1-5 (7)-nerved distally, 0.5-2 mm wide, 11-40 mm long (entire length including awns); if shorter then bowed out at base, and equal or subequal]