I updated my ”Botany Breakdown” of the Legume Family recently, with a focus on understanding the possibilities of legume trees and leguminous forages in my climate. That felt somewhat obligatory since legume trees figured so prominently in the the early permaculture movement (for better or for worse), which is was my gateway to agroforestry. There’s also something intuitively appealing about legume trees when you understand the African savanna origins of humanity. Legume trees are inseparable from tropical savanna.
But as I’ve spent more time in the temperate oak savannas of Western North America, I’ve realized that we have our own legume tree analogues- native actinorhizal shrubs. Most of our native nitrogen-fixing woody plants don’t form tall giraffe-pruned vase-like silhouettes, but they play very similar ecological roles. Perhaps our native woody nitrogen-fixers get less attention because these taxa are unique to this part of the world, while legumes are pan-tropical.
Frankia & Actinorhizal Plants (resurrected link to an excellent database)
Perhaps the primary Old World analogue of our region’s actinorhizal shrubs are the Cytisus species, including Scotch broom. These are in a distinct subgroup of the legumes called Genisteae which includes lupines. One could think of Cytisus as lupine bushes (like the true bush lupines of our Pacific coast). Unlike most of Legume Family, the Genisteae are uniquely able to form woody growth in cool northern climates. Genisteae is part of the larger Faboideae group which contains many of the northern legumes, notably Pseudoacacia, Caragana, Trifolium, and Medicago. These plants are not in the same group as the classic acacia-type trees more common in the tropics. Pseudoacacia and Caragana are not analagous to acacias- they are more like giant trefoils. I’ve noticed that Pseudoacacia will thrive wherever forest lupine does and function like the woody version of a lupine in forest understories.
We are lucky that Cytisus species are not our main nitrogen-fixing shrubs! These bushes are high in alkaloids and they are a nightmare for agriculture. Our native actinorhizal species are an unsung blessing.
Nitrogen fixing trees and their shrubs have some important applications, but it’s especially obvious in the local context that the presence of nitrogen fixing plants does not translate to protein available to animals or humans. The best possible forage systems, as far as they go, are boring grass/forb systems on soil that retain and cycle nitrogen without legumes. Nitrogen fixers, at best, can produce forage only approaching the quality of boring pasture grasses. Where legumes excel is in improving forage yield on poor or depleted soils where the best pasture grasses can’t thrive. Trees are generally less palatable than forbs, despite higher total biomass production. Trees tend to defend their greater protein productivity with toxins, antinutrients, height, or thorns. And I suspect that aspens or willows on fertile soil produce more and better forage than any hardy legume tree. The reason to plant trees is not total protein yield. Trees are for shade, drought feed, winter feed, mixed yields, or for overall ecological stability. Tree forages loose much of their appeal where it is too cold for evergreen trees.
One exception to forb superiority might be intensive coppiced systems. Because trees can grow so much faster than grass and forbs, regular cutting of tender tree growth might outyield the ideal forb system. Importantly- the most vigorous and palatable trees are adapted to grow above the browse line and stay there, so this could only be an anthropogenic situation. It is also theoretically possible to get the protein production of trees into the soil for better forb yields through cutting. In this case, toxicity would be irrelevant.
I am curious to experiment with woody actinhorizal forage systems maintained by burning. Burning results in surface nitrogen loss but might stimulate nitrogen gain in soils via root shedding. And burning can cycle minerals otherwise tied up in woody growth.
One way to cheat evolved tree defenses would be to use trees that evolved on islands without megafauna. This is the case of the famous tagasaste tree of the Canary Islands. The mainland Cytisus relatives of tagasaste trees are too toxic to make great forage. The actinorhizal trees of California’s Catalina Islands could likely be be used like tagasaste. One could also mimic the island effect through a selective palatability breeding program.
Another evolutionary trend to consider is that plants from forests tend to rely on growth rate rather than toxins to escape browsing pressure. But pastures are more work to maintain within forested regions, and forest plants will be fragile and thirsty outside of forested regions. It would be interesting to compare the palatability and growth rate of Ceanothus thysiflorus (a vigorous forest species) and C. intergerrimus (a similar savanna species).
I’d like to focus on Cercocarpus, Purshia, and Ceanothus here because they map best to Aridoamerica. Alus and Eleagnus probably each deserve seperate articles. That leaves some odds and ends in Roseacea and Myricaceae.
Look at these global distribution maps of Cercocarpus (left) and Purshia (right)
A map of North American Ceanothus distribution would be similar except for two Eastern species. Almost all of the roughly 53 species are western/southwestern. The Cerastes subgenus is more Aridoamerican than the Ceanothus subgenus.
Ceanothus distribution maps
Diversification of Ceanothus (Rhamnaceae) in the California Floristic Province
Ceanothus may be worth sorting into functional groups.
Vigorous Semi-Evergreen Ceanothus (Cold Tender)
Ceanothus integerrimus (California, Southern Cascades, Arizona)
–Tallest Ceanothus
Ceanothus arboreus (Channel Islands)
Ceanothus thyrsiflorus (Coast of California and Oregon)
Hardy Deciduous Ceanothus
Ceanothus sanguineus (Northern Rockies and Cascades)
Ceanothus martinii (Nevada)
–Thorny Ceanothus
Ceanothus cordulatus (Sierras and Cascades)
Ceanothus fendleri (Four Corners States)
Hardy Scleriphyllous Ceanothus
Ceanothus velutinus (Mountains North and West of AZ/NM)
Desert Scrub Ceanothus
Ceanothus cuneatus (Californian chapparal)
Ceanothus pauciflorus/greggii (hot desert)
Ceanothus greggi var. franklinii (rare Utah/Upper Colorado endemic)
Prostrate Ceanothus (Evergreen)
Ceanothus prostratus (Sierras and Cascades)
Looking at these functional Ceanothus groups, probably the ones worth experimenting with as tagasaste-type forages are the vigorous semi-evergreen types. The hardy deciduous types might be worth seeding and some management on rangelands. The scleriphyllous and desert scrub types are of interest only as desert restoration plants for keeping jackrabbits and the infrequent mule deer alive. The prostrate species are of interest as ground-covers in gardens, and possibly in soil improvement for timber species. Ceanothus blossoms range from white to lavender to almost blue, and are as beautiful as lilacs.
……………………………………………………………………………………………………………………………..
Cercocarpus functional groups
Resprouting Broadleaf Cercocarpus (from most to least vigorous/cold tender)
Cercocarpus betuloides var. catalina (Channel Islands)
Cercarpus betuloides (Southern Oregon, California, Arizona)
Cercocarpus var. macrourus/douglasii (Klamath/Siskiyou)
Cercocarpus montanus (Southern Rockies)
Narrowleaf Cercocarpus
Cercocarpus ledifolius (Intermountain)
C. var ledifolius (curl leaf)
C. var intermontanus (leaf not curled)
C. var intricatus (SW dwarf from)
Cercocarpus brevifolius (Mogollon Rim)
The hardiness of various Cercocarpus betuloides subspecies should be better investigated. For forage production, the most vigorous and types which are hardy enough should be used. The narrow-leaf types do not resprout after fire and are not amenable to heavy browsing. Hyprids between groups may be of interest for the driest areas.
…………………………………………………………………………………………………………………………….
Purshia
Purshia distribution map
Purshia tridentata to 16′ (Intermountain)
Purshia stansburiana to 24′ (Southern Intermountain)
Purshia glandulosa to 15′ (Nevada and SoCal)
Purshia mexicana 6’+(Arizona/Mexico)
Purshia plicita (Mexico, not hardy, pink flowers)
The tree Purshia’s are tall enough to be pollarded for cut forage.
………………………………………………………………………………………………………………………………
The combined genera of Cercocarpus, Purshia, and Ceanothus could be arranged on continuum of moisture requirement which would be roughly equivalent to distance from the forest edge. The vast majority of these species would not be found far from a ponderosa or Jeffery Pine. Ceanothus tend to be denizens of frequent fire conifer forest and chaparral. Purshia is more of a shrub-steppe genus with an intermediate tolerance for fire. Cercocarpus is occupies both fire-prone chaparral and fire-proof high desert rock. But there outliers from these tendencies. Purshia tridentata can be a proper cold steppe species like big sagebrush. And there are hot desert species like Ceanothus pauciflorus. Only in the transition to hot desert and the Southern Plains do we see significant niche space for Mimosideae legume trees. Most of the cool-arid woody Faboideae, like Psorothamus, are toxic.
All of rangeland manager’s concern for protein begs the question- Which herbivores are best suited to access the protein in native rangeland plants. In the case of our actinorhizals the answer is not cattle. If were were concerned about intercepting the most protein from our rangeland we could give due consideration to locusts, moths, mormon crickets, prairie dogs, ground squirrels, packrats, jackrabbits, pronghorn antelope, mule deer, elk, bighorn sheep, donkeys, llamas, camels, peccaries, sage grouse, prairie chickens, quail, and doves. And we should think about the niche space left by extinct herbivores like Columbian mammoths, ground sloths, southern muskoxen (think takins), Harrington’s mountain goat, pronghorn relatives, giant marmots, and others. There might even be room for hardy rhinos and ratites somewhere. The rangeland management involved in feeding even a fraction of these herbivores would be indistinguishable from conservation biology. I’m going to go back an bold those species I think would digest actinorhizal shrubs well. Of these, goats/sheep, llamas, camels would be most sensible to farm. Bison or yaks fit better than cattle. Donkeys fit better than horses. Takin and rhino husbandry might have a niche. The other species suite game ranch management.
As with woody forages generally, the key to sustainability is multi-year rest periods for recruitment. This is well studied in the case trembling aspen as forage. Multi-year rest periods would be absurd for grasslands. Woody forage is just different.
above: the extinct shrub ox of Southwest North America, relative of the muskox and takin
Resilience Land Care 