The way we lay out our vegetable beds should not be random or haphazard as it can have a strong effect on the behaviour of various crops and make a great difference to our workload. For example simply by aligning rows north south so that the land on either side of a tall row receives equal sunlight we can get very different results to east west arrangements where plants on the north side of a tall row live in perpetual shade whilst those on the south roast (with luck and in the northern hemisphere). We can use all such modifying cases to our advantage depending on our climate and choice of crops, or inadvertently let them work against us.

Perhaps first we should also consider why we so often grow in rows when other options are available. It appears that for millennia most vegetable and field crops were sown broadcast and seldom grown in rows except those few that were transplanted. In the 18th Century Jethro Tull advocated his improved seed drill; this was economical as it placed seed uniformly along multiple straight rows resulting in less inter-crop competition. It also allowed the use of a multiple hoe drawn by horses that could weed between the rows. Up till that time weeding, if done at all, was generally ineffective, thus his system was a breakthrough. By eliminating weed competition AND ensuring each crop plant had adequate space the returns went up significantly.

The row system may have emerged from this as gardeners cultivating their small plots emulated the almost magic like methods they saw the farmers adopting. Or perhaps they continued emulating the strip farming of land that was then disappearing from the scene. In either case in English speaking countries we started growing in long rows, spaced at unequal distances apart to allow for the needs of different crops. Only sweet corn escaped unscathed as it needed wind pollination and just had to be grown in blocks.

Meanwhile on the Continent most gardeners seem to have followed the traditional broadcast sowing system with their small beds or blocks, perhaps subconsciously reproducing the patchwork of fields of their continually subdivided farms. And why change, the broadcast sowing system works better on a garden than a farm scale as weeding and spacing can be done economically by hand. And it turns out that many plants prefer to be in blocks, or clumps rather where they can meet their leaves and form a weed excluding micro-climate.

Block planting also lends itself to showing up companion effects. With rows each row has two different crops on either side extending uniformly along with it. If either crop is hostile or friendly in some way to the central row the effect will not show up -so good performance will be put down to the gardener's skill and bad to the year as per usual. However with blocks then each chequer has different neighbouring crops on each of it's four sides who are at varying distances from all the plants constituting it. Thus plant effects on each other soon show up as good and poor growth across the patch.

There is no inherent right or virtue attached to either scheme, row or block, they are just different methods suited to different crops and situations. Equally the fixed bed system is different to the usual row arrangement of the plot itself. With many conventional plots the rows of crops and paths are continuously replacing each other. This in itself is no problem, it is easy enough to keep an accurate record of a row system on a plot to help rotate crops. The problem is that all the trampling up and down between rows all season compresses the paths and necessitates digging the whole plot over before the next crops need the ground. The paths need to be dug as the following year their space may be used for growing crops.

The fixed bed system concentrates the rows to beds of four or so feet across so that all work can be done from paths on either side. The paths are permanent so need never be dug immediately saving one fifth to one sixth of the total digging work. Lengthways paths need be no wider than a foot or so and most find cross paths are plenty wide enough at two feet. Generally the length is for convenience but more than fifteen foot is so long there is a tendency to jump across rather than to walk round.

There is one other big advantage that comes with fixed beds, or fixed paths as the system should be called. Not only do you not have to dig the paths at all, ever, but you do not need dig the beds so often either. Annual digging is not actually beneficial, most trials have shown one year in seven is more than sufficient. Excessive digging burns off humus and allows nutrients to leach away so save the effort and turn the compost heap instead. The only good reason for annual digging was to restore the soil after compaction from treading during the year!

Of course you can keep digging if you enjoy the exercise. However I have never dug most of my forty beds since I came here nearly twenty years ago, a few of them were dug initially as comparisons but I rapidly gave up the task. There really was no point to it. Anyway every four years or so in the course of rotation the Potatoes follow Roots and for both these crops the disruption of harvesting is more than sufficient cultivation to break up the soil and destroy ants nests, mole runs and so on.

Inevitably beds start to rise as compost, plant residues and worm tunnels permeate and expand their bulk. I helped mine initially by transferring top-soil from the paths-to-be onto the beds but with the normal process of time any fixed beds slowly become raised beds. Especially as the fixed paths become trodden down and compacted further increasing the difference in height.

Raised beds, once only some inches higher than the paths, give vastly improved root drainage and aeration and up on top the cold air drains off and the micro-climate becomes much improved. This makes all the difference with over-wintering crops and early sowings. It is not necessary to raise the sides or make them vertical with bricks or planks, the natural slump or slope of the soil is much cheaper but does limit the height difference obtainable. Very great height differences can cause erosion problems and the higher the bed the drier it will be -useful if you want to grow moisture hating crops. But with steep slopes you cannot get a loose mulch to stay in place! Having a bed with sloping soil sides gives moister more sheltered positions for some plants as compared to them all being up on the flat top of a mechanically raised bed. I also found fixed sides of wood or tile to be hiding places for pests, I prefer bare soil or a mulch instead. Currently I'm using straw to mulch the paths, sides of beds and around all crops that will take it so that I both conserve moisture and save weeding.

Regardless of the sides being fixed or free, on each bed the plants can be grow in rows as per usual, or they can be grown in small or large blocks as you wish. Likewise the record keeping becomes simpler in one way as each bed can have it's own record and can be regarded as a separate plot independent of all the others. Moreover then each plot can be directly represented on a card or a page in a book.

I keep cards; each of my forty beds has a card that tells me the crops grown on each bed over the last eighteen years. I not only plan to rotate the crops over as long a period as possible but when that crop's family returns I try to vary the actual cultivar. So for example when 'Brassicas' return to a bed I look to see whether I grew cabbages or cauliflowers there last time; this time I'll choose kales or sprouts and so on. When 'Roots' return I can follow carrots by other root-crops thus giving the carrots an eight year gap.

One question I'm often asked is how you cope with rotation when you also want to use companion planting? There are some crops that need more careful rotation than others; carrots, potatoes, brassicas, onions. It is best to rotate these first and then look next at what to plant with them. It is obviously easier if you know what grew where before. I have another stack of cards, one for each bed for each year. Each card is a plan of the bed with the position and variety of each crop's row, block or every individual plant that grew on there that year. So even when the same crops return after many years to the same bed the larger individuals such as brassicas can be placed in other positions to their original ones.

As I often do trials of many different sorts of the same crop, such as eight varieties of French beans, then a simple plan on paper is much safer than relying on individual labels. As much as I prefer the convenience of individual labelling the vagaries of nature (cats and birds etc.) ensure they often go missing or walkabout. A card or a plan in a book or on a pad is so much more certain. And when a potato volunteer appears three years running despite my efforts to eat all it's offspring I can look back and see which sort it was I originally planted on that spot.