A modular approach to raised beds
As I work through my design and layout for the new plot, the challenge of deciding what size, and how many, raised beds to accommodate has continually left me wondering whether the balance is right - but I may have finally found a solution.
On my original plot, I had a number of raised beds of various sizes - some 1 x 1m, some 3.6m x 0.5m, some 3 x 1m and quite a few 0.9 x 1.8m. Generally, this 'worked' - though in the last year the larger beds were taken up with perennials such as strawberries and rhubarb, and the smaller 1x1m beds accommodated fruits such as blackberry and blueberries. For the most part, I was left with 0.9 x 1.8 beds for most annuals.
This posed a few problems - firstly, there are inevitably some things you want more of than others - I had 2 beds planted with onions, for example. Conversely, sometimes you don't need such a large bed - and sometimes you only need a small space to begin with, but for a crop that will eventually out grow it (squashes are a good example of this).
As I've been working on the layout for the new allotment plot, I've been keen to avoid falling in the trap of only having set sizes left at certain points in the year - I wanted something that could adapt easily. Obviously the 'easy' solution here would be to just have straight rows, or do away with raised beds altogether. However, 'tidiness' and 'compartmentalising' is important to me - I also find this approach makes weeding easier (or, at least, seem so). I do appreciate that for many, this is of little importance.
My previous raised beds have been simple stakes in each corner, with timber boards screwed onto them. In most cases, this worked well - though with the larger beds they would quickly warp, even with supporting takes added along the lengths - the longer the timber, the less likely it is to be perfectly straight. For that reason, the approach I've settled on uses shorter lengths, regardless of the length of the bed.
The additional benefit of this new approach is that damaged or rotted elements can easily be replaced without having to dismantle the entire bed, or build a new one altogether. Finally, it can also be constructed without any nails or screws too.
The basic idea is simple - a number of short 'posts' (100mm x 100mm in my case) with channels the height of the timber you want for the sides - in my case I've chosen 150mm at 22mm thickness. The channels therefore are 24mm wide (extra 2mm to allow for some expansion in the timber, without the risk of them 'wobbling') and 150mm high. The posts themselves are 375mm high, with the remaining 225mm in ground to anchor them in place. The side and end timber then simply slot into these channels. In my case I'm working on the basis of beds 1.2m wide and just under 4m deep - using 8 posts per bed and timber 1.2m long (the 375mm height of the posts was dictated by the fact this is the size of a 3m length of 100x100mm, cut into 8 pieces).
Whilst this could work just as well on a bed with only 4 posts and 4 'boards', the main reason I'm doing this is to take advantage of those posts in between the ends - the 'middle' posts have 3 channels instead of the 2 in each corner. The extra channel points inwards, and is a little shallower - we don't need a full 150mm high board on those (I've decided a simple 25mm board would suffice, as this is mostly for organisational and 'tidiness' sake). This leaves me with the option of having, for example, three 1.2m beds or one 2.4m bed and one 1.2m bed, or a single 3.6m bed
Taking this approach, we could for example split the bed into three 1.2m beds and planting a pumpkin in the middle section, with some quicker cropping plants either side - as the pumpkin grows and needs more space, those dividers can be removed and the bed size doubles or triples immediately.
I do appreciate two things with this - firstly, it's probably a solution to a problem that doesn't exist for many. Secondly, it's probably not something you could adequately do with 'scrap wood'. But, for those like me who appreciate some level of tidiness and organisation (and are probably 'on the spectrum' somewhere!), this is the most cost-effective and flexible solution I have been able to come up with - it's certainly the approach I'll be taking.