Tempura with udo, hogweed, sweet cicely and lovage

If you read online about the origins of the Japanese tempura cooking style, you will discover that it was introduced to Japan by Portuguese missionaries who possibly adapted it from Indian pakora in their colony of Goa. The word comes from the Latin tempora or ‘times’, which was used to refer to the Christian fast days on which meat could not be eaten and fish was a popular substitute. However, we in Scotland know that it is simply an excuse to deep-fry things in batter and that no further justification is needed. If they can then be dipped in soy sauce, so much the better.

I have been experimenting recently with a number of strongly-flavoured shoots from the carrot family, including hogweed (Heracleum sphondylium), sweet cicely (Myrrhis odorata) and lovage (Levisticum officinale), and one from the closely related aralia family – udo (Aralia cordata). In all the carrot relatives it is the young leaf stems that are eaten, preferably before the leaf itself has properly unfurled. Lovage is particularly strongly flavoured and is usually used as a herb, but it can be made milder by blanching it – excluding light from the growing stems so that they come up pale and somewhat chastened. Clay forcing pots are traditionally used for this but I find that an upturned bin works just as well (it needs to be large as lovage is quite a size).

With the udo on the other hand it is usually the pith – the flesh inside the central stems – that is used: the resinous skin is peeled or sliced off. In this case, however, I wanted to find a use for the tips of the shoots that are difficult to peel and include the young leaves. I thought that tempura might be a good way of harmonising all these strong flavours.

lovage - udo - sweet cicely - hogweed

lovage – udo – sweet cicely – hogweed

For my tempura batter I used plain wheat flour with a little corn flour, an egg and chilled water. I mixed it briefly so as to keep it light and fluffy. The stems were all cut up into bite-sized lengths, then coated in the batter and deep fried. You get lots of little bits of batter left in the oil once you take the large pieces out: it is good to get these out if you can or they will spoil your oil.


The verdict? I loved it! The strong flavours were all still there but tempered by the cooking and the mild batter and by the competition with the soy sauce. This is definitely a dish that I will be making again.



Back to my roots (and tubers)

One of this year’s experiments in the allotment has been to try to get more starchy roots, tubers and bulbs into the forest garden. I haven’t had any straightforward, unmixed successes, but definitely some interesting leads that I want to follow up.

One fruitful (rootful?) source of edible roots is the Apiaceae or carrot family, home to the carrot and parsnip amongst familiar roots, as well as a large number of the plants that we use as herbs and spices (cumin, celery leaf, coriander, parsley, fennel, dill…). In particular, I was interested to try the native members of this family that have a culinary reputation: sweet cicely (Myrrhis odorata) and native hogweed (Heracleum sphondylium). Both have other edible parts (with some care in preparation required for the hogweed) but that’s a topic for another post. This time I wanted to get to the root of the matter.

I have tried sweet cicely root before. Like the rest of the plant, it has a strong aniseed taste, although the flavour moderates with cooking. Even so, it is a root to use sparingly in dishes with milder ingredients rather than to tuck into a plateful of. Mature sweet cicely roots are truly massive – and so deep that they are next to impossible to dig out whole. Unfortunately they are also not very nice: I only find the first year’s roots palateable, so I think that the answer is to have a few favoured plants as perennials for shoots, leaves and seeds and grow the roots as an annual crop.

I sowed a few rows in a couple of different sites with different seed sources, intending to select for the largest and straightest roots. The results were interesting. Site one was in open vegetable beds. The roots were a decent size but almost all quite heavily branched. Site two was quite shaded and yielded smaller but much straighter roots. I’ll have to do some experiments to find out whether this was genetic or due to the shade. Anyway, super-cicely is not on the horizon quite yet, but I’ll keep working on it.

sweet cicely roots

sweet cicely roots

The situation with hogweed is much simpler to report. The boiled roots were revolting. Yuk. I think I’ll be leaving them to the hogs from now on – although it does occur to me that I didn’t get to try first-year roots…

Other carrot-family roots include Hamburg parsley, which I’ll write about soon, and skirret (Sium sisarum), a perennial which bears a multitude of sweet, pleasant-tasting roots. Harvesting it involves digging it up, taking off some of the roots and putting the rest back in the ground. Turnip-rooted chervil (Chaerophyllum bulbosum) is a biennial with very pleasant-tasting but rather small roots. There is some evidence that strains with larger roots existed in the past so I’ve been growing the biggest of mine on for seed.


Another two candidates were South American crops that I have tried before years ago and which I decided to have another go at: yacon and oca. Oca (Oxalis tuberosa) is a tuber-producing relative of wood-sorrel, so the tubers have that lemony oxalic acid taste when they are fresh, but fortunately if you leave them in the light for a while it goes away. The big difficulty with growing oca at this latitude is that they only start to tuberise at around the autumnal equinox, by which time the frosts might be only a few weeks away. This year there were no hard frosts until November and I decided to try a little experiment. The plants had grown luxuriantly and I felt that they still had a lot to put into tuberisation if they weren’t completely killed by the frosts. I dug up one plant, covered two in horticultural fleece and buried a fourth under a thick layer of leaves. A couple of days ago I dug up the protected plants to compare yields. They were:

325g first plant
400g each for fleece-covered plants
550g mulch-covered plant

I imagine that a statistician would tsk at the size of my trial, but the results do suggest that covering the plants was worthwhile. The leaf mulch was more effective at insulating the tops than the fleece as well as being cheaper. Overall though, the yields are still rather disappointing for the space taken up. I don’t think that oca will really fulfil its promise in Scotland until someone finds a strain that starts to tuberise immediately. It isn’t too much to hope for as potato suffered the same problem originally.

Oca tubers

Oca tubers

Yacon (Smallanthus sonchifolius) is also frost-sensitive but had managed to put on a more impressive set of tubers than the oca by the time the frost hit. Like the related dahlia (also edible but not very nice), yacon crowns have to be brought inside to overwinter. It has two kinds of tuber: vegetative ones which will grown the next year’s plants and storage tubers which you can pick off and eat. The great drawback of yacon is that, like so many otherwise-promising plants in the daisy family (Jerusalem artichoke/sunroot, chicory, dahlia, elecampane, scorzonera), its roots are packed full of inulin, a sugar which can’t be digested by humans but which our gut microbes happily use, creating rather a lot of gas in the process… The culinary virtue of yacon tubers is that they keep their crunch even when cooked, so a little bit in a meal is great for adding texture, but I wouldn’t want to try to live on them.

yacon tubers

yacon tubers

The pea family (Fabaceae) is better known for its seeds than its roots, but there is a small number of tuberous species and one fascinating plant called talet (Amphicarpaea bracteata) that has adapted its seeds to work as tubers. This year I tried growing aardaker (Lathyrus tuberosus, also known as the Fyfield pea or Dutch mice: the Dutch name means earth-acorn) and cairmeal (Lathyrus linifolius, also known as bitter vetch or heath pea: the Gaelic name is pronounced like Carmel).

The aardaker gave a small crop of inch-long, tapering tubers that did indeed put me in mind of both mice and acorns. I would rather grow them on than lose any of them by eating them, but everything I have read about them suggests that they are delicious. The yield wasn’t exactly large, but it was produced in a very small space and, like all the peas, they have the added virtue of fixing nitrogen from the atmosphere, so I definitely think that this is one to keep an eye on.

Lathyrus tuberosus

Lathyrus tuberosus

By contrast, the cairmeal tubers were tiny. I couldn’t even take a photo that showed them properly. I won’t give up on it but might just abandon it for a couple of years and see how it gets on. There are advantages to cairmeal that make me want to keep trying: it is shade tolerant and has a long history of edible use in the Highlands.

Silverweed (Argentina anserina) is another plant that is reputed to have been used in the Highlands but has a bit of a problem in the yield department. I’ve had it in my lawn for years but decided long ago that its roots weren’t worth the effort of digging them up.

This year, I thought it would be interesting to take a closer look at the growth habit of the plant and transferred a few to pots of compost. They are strongly runnering so it can be very difficult to dig up a whole plant from the ground. In the autumn I shook off the compost and got a great view of the roots – with their swollen sections often at the end of the root but sometimes with the root carrying on beyond them. On the plus side, the roots taste great. I’m going to keep growing them in pots in order to (a) stop them invading the rest of my garden and (b) see if I can isolate individual strains and compare them.



Finally, some negative results. Two plants that I grow for their edible flowers, king’s spear (Ashphodeline lutea) and day lily (Hemerocalis lutea) are said to also have edible roots so in the autumn I dug up a few and tried them. Neither seemed worth it. The burdock (Arctium lappa) that I sowed produced next to nothing, as it usually does. I don’t quite understand this as burdock grows well in the wild here. I consoled myself with the thought that since they are in the daisy family they probably aren’t that much good anyway [UPDATE: I was wrong. See here]. My hopniss (Apios americana), another member of the pea family, did what it always does: it used all the resources in the one tuber that I planted to produce a total of one solitary tuber at the end of the year. I’m beginning to think that I could achieve the same result by pickling it rather than by planting it. It was a similar situation with the pleurisy root (Asclepias tuberosa). I’m experimenting with a few plants in this genus but mostly I think the climate is just a little too cold for them here. My attempt to grow Chinese artichoke (Stachys affinis) took rather a blow when the supplier sent me what turned out to be the related wild flower betony (Stachys officinalis) instead. Finally, my rampion (Campanula rapunculus), which is said to have long, fleshy, white roots, produced nothing but fibrous roots. From talking to others, this seems to be a common problem: has anyone out there ever grown rampion successfully?

rampion - where's my roots?

rampion – where’s my roots?

Other rooty posts: Chufa/tigernut; Wasabi, horseradish and friends; Self-seeded parsnips; Growing pignuts; Dog’s tooth violet

Next installment: Back to my roots (and tubers) – 2016