N-fix 8: Albizia julibrissin

Terrible Tree #1 — Mimosa (Albizia julibrissin) What’s wrong with it: Weedy, short-lived, insect- and disease-prone, invasive roots, unattractive most of the year.

Comment: Yes, I know. You grew up with mimosas in the yard (sniff), they remind you of Meemaw’s garden (sniff, sniff), and they’re so pretty when their fluffy pink flowers open in early summer. But let’s get real. The flowers last about two weeks. Then they’re replaced by scads of these large, ugly, brown seed pods that hang there until the next spring. So for two weeks of beauty you get 50 weeks of gross. Plus, seedlings from your tree will sprout in everyone’s yard within a quarter-mile.

-Southern Living Magazine

Once the naysayers settle down to let me get a word in edgewise: Mimosa -specifically Albizia julibrissin or silk tree -in my climate at least, does not own up to any of these accusatory standards. To boost, it’s quite a proficient N-fixer.

As to their longevity: There are several specimen of silk tree in my area, many of which are over 30 years old, some, in really protected areas, nearly twenty feet tall, and only all the more gorgeous for it. Perhaps we are outside the range of its pests. On the other hand, winters around here often frost silk tree back to the ground. Such dieback actually extends their likely lifespan, because the wood is renewed. For plant cells, this is the equivalent of turning back the biological clock, and can be done over and over again.

Short lived flowers? Perhaps it’s the cooler weather, but the bloom lasts far more than two weeks around here. Some do only bloom for three or four weeks, but several I see start in late July, and continue without stop until October.

I should mention some trees of course have more color to their plume-like blossoms. The one below my neighbor found as a seedling even further north, from a parent tree whose flowers were this undesirable color. It came out alright I think.

As to “ugly,” “brown,” pods: They’re tan, if not gold. I find them very attractive. In fact, even less attractive pods like these on any plant are often highly sought by advanced garden designers for “winter interest” they provide. Of course, a barren landscape in winter is more a northern problem than a southern.

Seeding all over the place: In the south silk tree is considered a noxious weed, per its willingness to pop up everywhere. It’s doing its job of turning open space into riparian field, to become productive forest. Shame on you for hogging up petroleum, riding around a growling hog to keep open spaces it’s just trying to repair. But I’ll stop my rant there.

If this seeding were an issue in northern climates I would only be all the more delighted to grow this plant; but I can say from experience that we only find a seedling or two every year, quite close to the parent tree. These do not always survive to adulthood, and so are quite precious, usually given as gifts.

There are rumors of a few trees in this climate that seed especially well, but these are interestingly enough, ugly white-flowered specimens.

These considerations are for tall trees though. My neighbor has allowed her tree to reach the upper story of her forest garden between a magnolia,  Montmorency cherry, and serviceberry tree. For most food forests, silk tree has much more practical functions.

Soft design aside, the tree makes N. It grows footage of branches every year, which in our climate helps it bounce back from winter-kill. In the case of a food forest, this means lots of delicious mulch for healthy soil.

The USDA cites this plant as a medium level nitrogen fixer. One study noted it made far more nodules than other N-fixers, although only had medium level bacterial activity as pictured below. (Notice all the species aside from silk tree are decidedly tropical). View the full study here.


This study in particular pointed out that nitrogen fixation is heavily dependent on sunlight for enough energy to run the nitrogen-fixing process. As a result, cloudy climates may not see such vigorous growth. But Ohio certainly isn’t the most sunny state; and many fast growing silk trees I have come across do so in half day’s shade.

I recommend taxing the tree of its growth for mulch at about time of bloom. You could just snap off the flowering branches with your bare hands. In the case that your tree has not reached blooming age but is getting out of reach for easily collecting mulch, just wait until about midsummer to begin snapping off branches.

One more practical accusation to address is the invasive roots. This refers to roots choking out plants below it, and roots becoming new plants. I should also mention that many accuse mimosa of allelopathy -manufacturing poisons for the surrounding plants (See Notes on Juglone for more info on allelopathy). Soil type, and climate can have major effects on how these attributes are expressed. I must report I find they don’t show up with the local specimen. My neighbor in particular has ferns, hosta, and tree peony below her mimosa. They have no complaint but for the slight dryness due to the thick branches above. Otherwise, they’re quite happy.

Studies have been conducted on mimosa and its relatives’ ability to inhibit germination in seeds. Although only a limited range have been placed in controlled studies, this study, for example, did not find any inhibition of germination.

View the rest of the text this image came from here.

The best way to propagate mimosa, aside from seeds which just need water, warmth and soil in most cases, is actually to dig up pieces of the root and let them sprout branches. I have tried cuttings, and they don’t root. If you are hacking up the ground with a shovel around the base of a silk tree, then perhaps you’ll get a sprout. We never have been lucky enough for this to happen around here.

When placing mimosa in design, I give it especially visible positions, where people can interact with it. For one client, I have mimosa at the corners of a grotto design where they will be allowed to grow as large as winters allow. The intent is to create a dome of arching branches. In winter these will quietly rustle with golden pods.

Placed where you might brush by the leaves, silk tree is especially entertaining, because it folds its leaves in response – a character which in plant terminology is called thigmonasty. The tree also folds its leaves at night, or in the dark before storms. This response to light is called photonasty. Below are some photos taken by my client of her young trees.

My client asked why the mimosa had stopped folding its leaves when she petted it. The fact is they can learn.  In The Hidden Life of Trees a study was cited which used dripping water as a means for stimulating a mimosa’s thigmonasty, causing it to shut. After many days of applying water in regular intervals, stopping, applying again, it learned this was not a threat, and got on with photosynthesis.

An intelligent tree like this is most desirable for adding to food forests as pleasure for the eye, delight for the curiosity, as fountains of mulch and fertility, for vibrant, healthy, cold climate jungles. Thank you for letting me state my position.


N-fix 7: Vetch

I was rather dissatisfied when finished with the Top ten ground covers for food forests series because I didn’t include an N-fix in the lineup of dynamic mulch. Vetch would be my first choice.

Crown vetch (Securigera/ Coronilla varia) is indeed the tyrannical ruler of the genus. It’s industrial grade ground cover, produces medium levels of nitrogen, stretches 3-4 feet tall, and eats as much as 60 ft of new territory in all directions via rhizome spread every year.


I’ve only seen this number in reports. I’m guessing the ground where these measurements were taken was the most fertile sandy loam on the face of the planet, or someone dropped a bag of fertilizer. I’ve never seen it spread more than four feet a year -if it spreads at all. many readers have likely seen this plant along highways where governments have seeded it. Erosion is a thing of the past once this perennial is established, so they have encouraged its use.

Once established though, they can start new colonies via seed. As a result, many governments have removed it from their official list of recommended plants.

It’s a little late. I have some patches appearing by the road at the edge of my food forest, and one of my clients has several healthy patches around their property. I have in fact installed this plant in one client’s system, because it really is quite useful, and easily controlled, if you know its habits.

The one system where I installed this for instance was a very high production food forest with long rows of shrubs and fruit trees running along swales (not my designs. I was brought onto the scene in the later stages). The owner wanted a system that could be managed by laborers simply weed-whacking the place every month or two. With this mentality, the owner had started by laying black woven plastic mulch under the plantings -despite my disapproval. It didn’t take more than two moths before weeds found their way through. He was aghast when I showed him. Finally open to my suggestions for a dynamic mulch, crown vetch was the perfect candidate.

As a dynamic mulch, it can exclude grass. Sometimes, if the grass is well established, the two may persist together. If well established, it’s nigh inpossible for weeds to get a foothold in the crown vetch’s domain.

It grows well seeded among daylilies, and most any plant that grows more than three or four feet -a food forest cover.

The crown vetch in my food forest has to creep across the upper grain patch before it can invade any of my plantings. This is one of the most poor spots in the food forest even the crown vetch can’t stomach crossing. I simply tax it for its nitrogen rich growth about the time it comes into bloom. Because it can be pulled so easily, I simply yank up whole armfuls to feed other parts of the food forest. As you can imagine, the patch is expanding at a snail’s pace.

My clients have done the same, and find their patches stay put.

Hairy vetch (Vicia villosa) and wood vetch (Vicia sylvatica) are both annuals, but have scarcely less ability to spread. Hairy vetch is used commonly used as a ground cover in vegetable gardens to be tilled in as a green manure. It and wood vetch are also medium level n-fixers. I have seen vilossa reach about four feet tall as it meanders around, such as below with wine-raspberry. Reports say it can reach as much as eight feet. So long as it doesn’t bloom and seed, there isn’t a problem.


I actually let the thing seed in my food forest because it grows right among weeds, grass, you name it -without complaint. I love it. When it blooms, I nab as many as I can for mulch, leaving the rest to seed. I really just wish I had more.

I came hit the jackpot the other day at my step-grandmother’s garden. She has the wood vetch, and a lot of it. I gathered a bunch of the pods as pictured below, which popped, flinging seeds every which way even as they sat in the bowl.


I planted them in a problematic part of the food forest to see how well they might compete with the grass. Perhaps they’ll clear the place for me in a similar way sunchokes can. in the bed I found it growing so happily, it was hardly three feet tall. I’ll just yank them up or cut them down before they make any very much seed.


Growing Amorpha

Deep purple petal over bright orange stamens of Amorpha fruticosa melt into green, spotted little crescents of seed. These ripen in the sun to a dark brown, then white-gray and hold. I kept thinking these seeds would fall off, but even in January when I lately picked some, they were tightly attached.


They have a peculiar ability to sprout without cold -unlike most woody perennials that need months of subfreezing temps. They just need heat above 70 degrees F, and up pop little green leaves. Usually I soak the seed for a couple hours before sowing into flats. I start them in February in some years, March in others, but for my climate these both mean heat has to be provided.

I have to be careful with the dry air of indoor heating to keep the seeds wet, so usually cover with some plastic, and water often. We begin heating a small portion of our greenhouse about that time for garden vegetables, so these seedlings can soak up real sunlight from day one.

They are wise little seeds, and spacing their sprouting time -which outdoors would be a fail-safe against late frost and other catastrophes. For me, it’s a great convenience. Out of one ‘source flat’ as I call it, sprouts pop within three days after planting, but keep popping up for several weeks.

Usually I wait until the first true leaves show before I begin transplanting, then clear the flat of any sprouts with true leaves once a week.

Out of the hundreds I have grown, I find it’s best to start the seeds with potting soil, or compost with good levels of nitrogen. From here I separate into small pots or cell flats no larger than 2 inches across, filled with the same kind of nitrogen rich potting soil they sprouted in.

The heat and rate of drying in smaller pots, where the roots can quickly reach the bottom and be air pruned, has given superior results for me. They still develop very deep taproots once in the ground, but this root pruning while in the pot is helpful -in part because it stimulates more branching of the root system. Planting in extra large pots with nitrogen rich soil, many seedlings rot, and must be replaced two or three times over before each pot successfully grows a plant. On the other hand I have tried planting them this early in nitrogen poor soil, and they make little headway.


I think this best mimics the situation they would find in nature. Forests and grasslands have a thin layer of nutrient rich, fluffy soil on the surface usually, which quickly becomes clay or whatever the base soil of the area. I want to get the seedlings into nitrogen poor soil to induce nodulation (aka hosting nitrogen fixing bacteria as evidenced by the formation of little nodules). This is spurred on by a lack of nitrogen in the soil. The catch is it takes time for the young plants to find the bacteria and get the symbiosis set up.

Nature’s way seems to be nutrient rich soil at first, then less rich soil as the plant gets bigger, the roots deeper. My contrived biomimicry that gives best results is moving the seedlings once they have filled their small pots and gotten a bit root pruned (not pot bound, as in roots turning back on themselves) into larger pots of whatever size you choose, filled with nitrogen poor soil about 1/3 rd coarse sand. I usually mix nutrient-rich rock powders, such as carbonitite or granite, into this before filling the pots.

Usually I transplant into 4inch pots at this point so they are filled with their roots in a couple of weeks -about the time nodules start to form. Usually this is early June -plenty of time for establishment before fall. Those I don’t get in the ground the first year go into gallon pots by August, which they usually have amply filled by next spring.

I try to avoid keeping Amorpha in pots more than a year. They grow best put in the ground as soon as possible after they have acclimated to the nitrogen poor soil. After years of refining this method, I’ve had transplants pushing 5ft by the end of year one -well on their way to exploding every spring with growth, providing some of the best organic matter for fueling your plant projects.

Where to get the seeds? You might have a plant nearby, which I recommend you snatch some seeds from. Otherwise they’re very affordable, and widely available from Sheffield seeds (my first choice), Oikos Tree Crops (They advertise A. californica, but I’ve gotten their seed -and plants, and compared it against pictures and attributes on the USDA plant database, and they have the name wrong. It’s species fruticosa), or even Amazon if you shop there.


N-fix 6: Senna hebecarpa

One of the few hardy Cassia/Senna’s, Senna hebecarpa. Photo credit: George H. Bruso found on Ladybird Johnson institute

The method of finding hardy nitrogen fixers I used to comprise my list is pretty simple: choose a genus of nitrogen fixers (I say genus because there are some weirdos that can’t be found from the higher orders and phylum) and look through the whole genus. Often, there will be some member that is hardy in your area.

I consider one of the greatest prizes I found this way to be Senna/Cassia hebecarpa.

Cassia is a genus almost exclusively tropical, and mostly trees, which are gorgeous.

Cassia alota, commonly called “Candlstick Cassia” for its looks. Can’t grow it. Click for source.

Like a lot of the non-hardy N fixers cassia are woody yet legumes and utilize Rhizobium to fix their usually medium level nitrogen.

S. hebecarpa, on the other hand, is herbaceous, and clump forming. Like lespedeza it grows up over the summer and dies back to the ground in winter. Although senna usually stops around 4 ft.

A more established clump of S. hebecarpa. I can’t wait until mine looks like this! Click for source.

This change in height allows for some niche-in-time filling because it makes sunlight and mulch abundant in spring when most perennial vegetables are growing fastest, but in summer and fall covers everything up to protect from wind and make a buffer against quick changes in temperature extending the warm season.

Another plus senna has over lespedeza is its taproot rather than fibrous shallow root system. Senna is native to the eastern US prairies, which were created and partially maintained by droughts that kept the water table below the reach of tree roots. So it makes sense senna would go deep to get its sustenance, leaving the upper ground for shallow rooted plants to play in, so plant away!

I haven’t bothered starting this from seed yet since I found decently priced second year roots available from Prairie Moon Nursery.

The happiest of my three S. hebecarpa this year. It barely bloomed but made a nice four feet of growth.

Companion Plants sells first year seedlings, and lots of ornamental greenhouses sell it because it’s really quite pretty with it’s yellow flowers characteristic of the genus.

Senna marilandica is different from S. hebecarpa only in how its pods break open in fall. So if you have a hard time finding the one the other is an excellent stand in.

Another characteristic I like about this plant is its thigmonasty -the tendency to fold it’s leaves when exposed to touch, or when the sun goes away for night or before a storm (technically photonasty), or when it’s overly hot (thermonasty). Amorpha and several other plants do the same, but senna with its larger leaves is more notable. It’s caused by electrical sensors in the plant modifying the potassium in the leaves causing an efflux of water from the cells, reducing turgor pressure, and thus reducing leaf rigidity.

Being so ornamental and interesting to watch I’ve placed these in very visible spots in the food forest. Of course, I’d make it visible if only to be reminded I have a Cassia in my northern food forest as it seems I’ve got the collector’s bug.




Growing annuals

I’m always writing about perennials and no till, but I actually do a lot of annual gardening. My family’s farm has a CSA for about 20 shares that uses about 3 acres of tilled gardens.

Mortal Tree is my exposition on alternative methods, so of course I wanted to try growing annuals in a passive, no till setup.

That’s what the keyholes at the front were supposed to be; with perennials thrown in for propagation, and greater efficiency.

After two years of the first “system” (or lack thereof) there are two problems: lack of fertility, and weeds. Of course everyone has those problems, but the system was supposed to keep these to a tolerable level.

As it is, aside from mache, annuals just aren’t satisfied with the fertility. This year I didn’t harvest anything but what the perennials willingly supplied.

Ideal weed level is pulling a few weeds as I inspect the garden on a pleasant evening. Instead, I found myself clearing whole beds only to have them full of weeds again in a month.

I’ve scrutinized the system and found two problems: lack of mulch and lack of fertility in general.


The year of high fertility

I got the idea I wasn’t cycling enough nutrients from the lack of vigor the annuals and their self seeding progeny showed. I wasn’t sure how much more I needed. Normally the tillage and compost in the big gardens makes everything grow without complaint. Now that I’m trying to make this work with comfrey and other in-system nutrients, without tillage, it’s not.

Click to view

Martin Crawford has several tables in Creating a Forest Garden that really pin down nutrients and how much different plants need, and how much different sources offer. He has a light, moderate, and heavy cropping category, then annuals.

Most of the perennial vegetables he places in the light cropping category. It takes about two cut comfrey per square meter to sustain these plants, which is about what I am applying. To sustain annual cropping takes 60 cut comfrey for that same area. Problem found.

The amount of fertility I’m accustomed to working with in the annual gardens is simply an unnatural surge of nutrients. Compost is extraordinarily rich in nutrients compared to bulk green mulch.

Also, tillage forces more oxygen into the soil, breaking down those nutrients at a faster rate. The keyholes had this advantage at first because I dug out the paths and piled up the soil to make the beds.

Problem is, the mulch apparently needed for the annuals would drown most of the perennials. In response I’m moving all the perennials out, and making the keyholes completely annual.

The best comparison I have for this so far is a keyhole bed I have near the Willow Garden in its fourth year of no till.


It has a dug out path and raised beds too. It has one comfrey for its mulch and fertility source. As a result, I’ve had to bring in more mulch to sustain the system.

For instance, I brought in a lot comfrey from the Willow Garden to drown out some quack grass (Agropyron repens) that had moved in. It was rather effective at suppressing it. Besides some vegetable mallow over the summer, it was enough to grow some nice cabbages.

In the food forest, I’m not supplying 60 cuts of comfrey per keyhole per year. I’ve got 12 beds with an average of 1.3 square meters each. It would take several hundred plants.

When faced with a large surge of energy in a design, I always try and disperse the blow across many sources.


Young regrowth on a coppiced amorpha

According to Martin’s fertility tables, one amorpha, based on mature canopy size of about six feet in diameter, provides the equivalent nitrogen of 20 cuts of comfrey, or 10 comfrey plants; lespedeza about the same. I’ll throw some of these in the patch so the roots sloughing off after coppicing can feed the comfrey. They need nitrogen themselves.

This still requires more space than I’ve allocated to the annual’s mulch patch. So I’ll grow some annual cover crops on the keyhole beds once in a while to fill the gap -careful to choose crops that will die when cut, frosted, or heavily mulched since there won’t be any tilling.

To really cinch the deal I have the food forest rabbit’s manure. I let a bucket of it sit out to catch rain, and harvest the resulting “tea” to feed establishing beds right now. In time it can be exclusively for the annuals.

Achieving this much mulch in the food forest will take a while. I’m moving out the perennials first, and planting the whole thing in lots of annual cover crops. The first plant to start yeilding mulch will likely be comfrey, though I might just mulch their own patch the first year to ensure they are established. The next year the N-fixers will be ready for light coppicing, then full production. I’m assuming three years before that point, but I’m quite excited to the see the results. I’ll keep you posted as it goes along.

N-fix 5: Lespedeza

In the last N-fix I touched on the use of  autumn olives  (Eleagnus umbellata) in reclaiming strip mines. I wanted to continue this theme with autumn olive’s southern counterpart lespedeza.

Winter Lespedeza. Photo © Chris Evans, River to River CWMA, Bugwood.org
Winter Lespedeza. Photo © Chris Evans, River to River CWMA, Bugwood.org

Despite its similar occupation, Lespedeza bicolor is very different from autumn olive. It’s not actinorhizal, it’s a legume, the more common kind of N fixer. It’s hardly woody, it’s herbaceous, growing a full 6-12 ft every year only to die back to the roots in winter in my climate. Its root system is rather shallow, and quite fibrous. It’s a medium level nitrogen fixer, more in the range of amorpha. About the only characteristic it has in common with autumn olive is it grows like crazy in soil worthy of horror movies.

It’s been an invasive species that colonized and then poured out of the southern strip mines for decades. Lucky for us in the north, and probably the reason why I’ve never seen a lespedeza in any strip mines around here, lespedeza needs a long season to ripen its seed –a minimum of 160 frost free days.

You might wonder then how it has a range far in to Canada? There are exceptions. In fact the USDA selected for fast maturation of seeds in L. bicolor the most notable being ‘Natob’ which ripens seed a whole month earlier.

There’s an interest in the plant making seed because it was intended to feed wildlife such as quail. For this, and grazing animals the USDA mentioned switchgrass (Panicum virgatum) makes a good companion. Switchgrass grows very tall during hot weather so ditto for harvesting mulch.

So far I haven’t bothered with L. bicolor and have stuck to the more showy but similar L. thunbergii. My Neighbor has two cultivars, ‘Alba’ and ‘Gibralter’ in her yard, well over a decade old, and has never had them reseed or set seed on the ‘Gibralter’ as far as she can tell. Neither is a hybrid, but seed doesn’t seem to ripen for her.

Lespedeza thunbergii ‘Alba’ in my neighbor’s garden. The gold in the background is a variety of black locust (Robinia pseudoacacia).
My own little ‘Gibralter’ with the resident aster.

The two lespedeza in my food forest are side shoots from her ‘Gibralter’ I transplanted. Unfortunately they weren’t successful in growing to six foot like hers; they topped off and bloomed very prettily at three feet this year. I assume they’ll get that size once well-established.

I would like more though, because lespedeza’s drastic change in height every year allows for a lot of niche planting with tons of room available in spring, then partial shading in summer, and an all out eclipse in fall. Then too, the canes will last through most of the winter, creating a wind break for more tender plants.

‘Gibralter’ nearly finished with flowering in my Neighbor’s yard.

The only possible problem could be the shallow root system. I haven’t experimented much with it, so can’t say how much it suppresses nearby plants, and my Neighbor mostly has other shrubs around hers. Best bet is probably spring bulbs such as camassia closest to it. Then a little further away where sun will last until at least mid summer you could plant Turkish rocket which is very taprooted and doesn’t do much after May. Good King Henry has about the same schedule but its roots don’t pierce quite as deep.

Two of our native species, L. capitata and cuneata are both much shorter, capitata topping off at four and cuneata at a two feet usually.  Cuneata is considered an invasive species in some states though, such that its sale has been banned in them. I don’t plan on growing them unless included in a pasture mix.

I prefer ‘Gibralter’ not only for its biomass, but its beauty. It has very deep pink color with an almost iridescent blue and purple hue on the tips if you look close. Often it’s placed near a pond in ornamental gardens to give the effect of a green and pink waterfall. Here is an excellent example

Overall, lespedeza is quite a diamond in the rough with whom I’m pleased to keep company.


N-fix 4: Eleagnus umbellata

It always surprises me how many people don’t notice the plants along our highways. I can’t be the only one driving down the road remarking the excellent pairing of plants nature makes in the median strips. Of course a lot of people bemoan that we have a poor range of woody nitrogen fixers so perhaps they don’t realize our median strips are clogged with one of the best woody nitrogen fixers known. Please say you recognize the common name: autumn olive?

The most common form of autumn olive: in an open field turning to woods.
It’s as though this plant were invisible to the general public. It is in fact one of the few elite plants to hit high nitrogen fixer status. This category starts at 160 lbs N fixation per acre, and doesn’t even have an upper limit.

 This is the first N-fixer in the series that is not a legume. It is actinorhizal. The reason being it doesn’t use Rhizobium bacteria to fix N; the whole group, made up almost entirely of woody trees, use a filamentous bacteria called Frankia, which has quite a high salt, water, and drought tolerance.

Combine this with an early leafing, and in general shade tolerant disposition of Eleagnus, and you have a specimen rugged and ready for any terrain.

In fact its ruggedness is what gave it the green card to emigrate from East Asia to Northern Europe and the US. We needed something that could rehabilitate strip mines -places where land is stripped of its soil hunk by hunk to get coal under it rather than building an underground mine.

I’m not sure how widespread the practice is, but I have a lot of it around me -mostly done 50 years ago. Even today after rehabilitation, many of these sites look like a scene of Mordor (in fact many movies are made in strip mines for that affect) with grey and black 20150411_075129_LLSrock and coal, and orange clay. Often there are huge ravines left that were used as personal dumps. Then most springs in them are some variety of toxic –either gray or orange red water.20150411_075931_LLS

So if this gives you a picture of what Eleagnus was brought to America to live in, and that it thrived in, you get my meaning for rugged.

Obviously any food forest where fruit trees are expected to grow will be a walk in the park for Eleagnus.  All the better because it is a fruiting shrub.

I remember the expression on my friend’s face when I first introduced him to the shrub. He was at first wary of eating off a random bush in the middle of a half wild field, but once he tasted the berries his face lit up with surprise and copiously praised them saying they would be amazing in a salad.

I completely agree. The flavor to me is like a tart apple, although it has a more intense sweetness than any apple I’ve tasted. Some species like E. multiflora, commonly known as Goumi in Japan, naturally has much larger fruit, and about the same flavor I’m told. There are several varieties of E. umbellata bred-up for size and flavor, such as ‘Ruby, and ‘Amber,’ which I have; although I haven’t been able to taste any yet to say how much better they are.

A more sensuous experience Eleagnus has to offer are its flowers. I love their scent-and others do too; one woman said she’d pay me if I could dig her a wild bush to smell in her yard. It lands somewhere between a floral and a gourmand in terms of perfume, being a vanilla spice scent. Thick as it is, I’ve caught the scent wafting down wind a whole fifty feet from the plant.

A very tree-like Eleagnus reaching for the light at the edge of a strip mine.
As for its form, it’s highly variable. In wandering strip mines I’ve seen some with six inch single trunks that are more fitting as trees stretching for the upper story, and shrubs with dense canes coming up from the base as a thick bush. Coppicing in the sun will produce the latter effect, limbing up in shade the former.

Limbing up in shade is one place Eleagnus does very well because it can take shade and still produce a decent amount of N for the space it takes. For the most nitrogen and fruit though, it’s better to let them get sun in the overstory, or have some sun in the understory where you can pick them.

Notice at the top the Eleagnus is wrapping around the tree like a vine. A novel way of growing a shrub.
So after all that, I just hope you don’t get into any wrecks studying your new found highway inhabitant. Perhaps it’s better to study them in your food forest. They’re far more productive that way.