This article originally appeared in the March 1996 Newsletter of the Liverpool Branch

Some Observations on Water Retention and Aeration

by Ray Allcock

Water retention and aeration are crucial factors in the cultivation of cacti and succulents in pots, and are best considered jointly. A first orientation can be obtained by considering a less than perfect current practice, wherein aeration is ensured through a liberal use of impermeable stone in the form of chippings or small pebbles.
Upon making an assessment of granite chippings graded to within the size range (3)-5-6-(8)mm I found a water retention after submersion of only 60ml per bulk litre of chippings. The amount of water held in the interstices by the forces of surface tension is small with chippings as large as these: the meagre quantity retained can therefore be accounted as due principally to the surface wetness of the chippings. The volume of the air space between the chippings may be ascertained by measuring the increase in weight upon pouring water into a beaker full of the dry material; - with this sample it came to 490ml/l. These chippings thus gave good aeration but poor water retention. In shape they were very irregular.

Chippings of flint (i.e. silica, silicon dioxide) come with a somewhat less irregular array of shapes. The same measurement procedure applied to a graded sample sold as hen grit revealed a substantially smaller unoccupied space, amounting to 370ml/l. In the case of close-packed spheres of uniform radius the proportion of space unoccupied can be shown mathematically to be exactly 1 - /(32) which is smaller still, only 260ml/l. Thus the air space achieved through irregular size-graded fragments is without doubt larger than that obtained with size-graded rounded pebbles.

In the first example the size grade given led to good wet aeration (430) with poor water retention (60). The opposite situation holds at small grain sizes. A trial using a potful of sharp horticultural silica sand with grain sizes within the range 0.25-1mm gave in the dry state an unoccupied space of 375ml/l (much the same therefore as the hen grit) and a water retention slightly in excess of 370ml/l. At this small grain size the aeration is thus essentially zero after a heavy watering. The full expulsion of all the air from the interstices leads to a state of ‘aquatic gridlock’, in which the draining off of the excess water is completely inhibited by surface tension.

The size of the grains is thus very crucial to the satisfactory aeration of sands and gravels and other media composed of impermeable fragments, and if the grain size is small, as in the above case (and a fortiori in a compacted sandy loam), then heavy soakings impede the drainage and block out the air, and are to be rigorously avoided.
Practical experience with cacti amply supports this last deduction - it is quite possible to get good growth in compacted sandy loam, but root rot usually ensues if the pots are watered to excess, since this induces the state of gridlock.
To establish a boundary between the two types of behaviour I conducted a trial with flint chippings in the intermediate range (1)-1.5-3-(4.5)mm. These taken dry gave an unoccupied volume of 420ml/l and a water retentivity of 220ml/l. Here no less than 90ml out of the total of 220ml was attributable not to surface wetness of the grains but rather to water occluded at their points of contact. The occluded water was put into view by vigorously stirring and beating the chippings, whereupon it ran out to be weighed. Even after this efflux the chippings still felt exceedingly wet.

In this last example the occluded water gives to the wetted chippings a quality of sopping wetness which I judge to be quite inimical to the health of the roots of desert cacti, and indeed several such which rotted from the roots upwards soon after I bought them were found upon depotting to have been accommodated in a mass of similarly small chippings interleaved with equally small fragments of peat, reminiscent after watering of the contents of a freshly brewed tea-bag! The plants looked very nice when acquired and were at that time obviously healthy, but in retrospect the reason for their unhappy demise in my care has become clear enough - the compost was a sure guarantee for quick death in my regime wherein, more for reasons of convenience than of theory, I act upon Professor Borg's advice and water generously but seldom! It will be seen that we have here quite a problem. Let us leave it for a moment and turn to soil-based composts.
A John Innes or other soil-based compost cannot be reckoned to be in good condition unless the bulk of its loam content and some of its peat appears in the form of small nodules or ‘crumbs’. The crumb structure imparts aeration and, concomitantly, the property of being ‘free-draining’. The crumb structure of loams is a variable attribute, but in all cases it is easy to damage it by excessive ‘thumbing’ or by over-generous and/or forceful watering. In times past this danger was well recognised, and to guard against it a loam full of fibrous dead grass roots was regarded as mandatory. Nowadays it is reckoned that the harvesting and preparation of loam in this state is an unaffordable luxury. Such is progress!

I have explored the air space and retentivity of a good (i.e. crumby, not powdery!) present-day representative of the John Innes formulation by starting it in a slightly moist and reasonably firmed down state in a plant pot, and then gently watering it with water laced with Chempak Wetting Agent, until an efflux from the drainage holes signalled me to stop. The increase in weight indicated a retentivity of 270ml/l. The drainage holes were then sealed and more of the water added, to fill up all the voids. In this way the residual air space was assessed as 150ml/l and, adding the two figures, the total unoccupied space in the dry state as 420ml/l.

The drainage holes were then unsealed, whereupon 140ml/l of water ran out. Thus with the wetting agent this compost proved to be free of the aquatic gridlock effect earlier mentioned - the flooding employed to fill all the voids succeeded in trapping only a mere 10ml/l. After all this mistreatment the compost was turned out and inspected, and it could be seen that the integrity of the clay nodules and the large voids had not suffered any noticeable damage.

Much of the air space in this sort of compost lies within the pores of the nodules of clay and within other very tiny cavities, and cannot be explored without the wetting agent. Indeed, when I first attempted this assay the four successive figures obtained without its aid were respectively 190, 130, 320 and 90. The two middle figures here are of course susceptible now to misinterpretation, since in fact the wetting is manifestly incomplete. In this wetter-free run the crumb structure was very noticeably damaged; at the end the compost gave the appearance of wet mud. MORAL: Never flood a J. I. compost, but if you cannot repress the urge to do so, then at least lower the surface tension forces by using a wetting agent!

Cactus enthusiasts often add some extra grit or gravel to John Innes composts in the hope that this may increase the aeration and drainage. In the case of a compost with good crumb structure, such as that just discussed, I would guess that added gravel will give merely an unprofitable occupation of space. Sharp horticultural sand would probably serve better, by helping the clay nodules not to stick to each other. At gravel sizes Perlite would also serve better, by bringing in its own internal air content and excellent properties of capillarity and wettability. Best perhaps would be some of each. If gravel is to be used then its principal role must, I think, be seen as a precaution against later degeneration of the crumbs; obviously it or Perlite etc. must be added in amounts generous enough to form a sort of supportive skeleton if this sort of long-lasting benefit is to be ensured.
When a slightly damp J.I. compost is watered moderately (as is appropriate to good cultivation) some of the wetness at first induced is rapidly absorbed into the interiors of the crumbs of loam and peat, and into the interiors also of the grains of any Perlite that might have been added. By this effect water initially occluded is removed from the voids, thereby replacing wetness by dampness without any diminution of water content. The resultant airy and long-lasting dampness is conducive to root health and root activity with all potted plants, and especially so with cacti and succulents. This consideration powerfully suggests the incorporation of fragments of absorbent material in place of sand and grit or gravel.

It becomes of interest therefore to investigate the retentivity and aeration properties of various readily available absorbent fragments. We will advisedly choose a grain size not smaller than 3mm so that, as evidenced by the results reported in earlier paragraphs, water occlusion becomes a minor factor in the overall behaviour and in the interpretation of the observations.

A good wetting undertaken without any wetting agent gave the following retentivities, which should be compared with the meagre 60ml/l obtained from a surface layer alone in the trial with granite chippings of comparable dimension:

graded fragments of soft low-baked red brick 200ml/l
graded Perlag (medium or coarse), Seramis 300ml/l
Biosorb (medium or coarse grades) 350ml/l
Perlite (medium or coarse grades) 400ml/l

The first of the above 5 products was prepared at home from 150-year old rotten bricks and was graded by sieving: the others are of commercial origin and come already graded. Perlag is a less-expanded and hard form of Perlite (expanded chemically inert volcanic rock), Seramis and Biosorb are composed of red clay expanded and baked.

It is evident that these materials can store up far more water than chippings or gravel. What is more, the initial surface wetness is rapidly and entirely absorbed into the interiors of the grains if watering is done from above in a normal and not excessive way. If water is allowed to soak in from below in unlimited amounts or poured on from above in unlimited amounts then the surface wetness will of course persist but even then, provided the grain size exceeds the stated 3mm, water will not be occluded in the spaces between the grains. None of these materials suffers structural degeneration in use. Thus we have here an effective way to avoid or to mitigate the problem of crumb structure deterioration and simultaneously to enhance the overall aeration, water retention and wettability.
I have not included Hydroleca in the above list, because the hydroponic purposes for which it is designed are quite different from those presently under discussion. It is composed of waterproof lightweight high-baked spheres thinly coated with a low-baked absorbent surface layer. The surface layer has excellent properties of capillarity but due to its thinness a retentivity of only 100ml/l or thereabouts. Also I have not included crushed sandstones and gritstones, both because of the danger of decomposition in use and also because some of them are lethally alkaline.

Due to their vascular structure fragments of wood charcoal show excellent wettability and high retentivity (about 300ml per litre of graded fragments above 3mm). However wood charcoal as left from the fire shows a high content of extremely alkaline substances (potassium carbonate, etc.) which must be neutralized before use. It is moreover a dirty material to deal with. Further, it has unique physicochemical properties which could upset the pH of the soil by abstracting and retaining ions which would better be left in the soil solution. Its traditional use in horticulture is to mop up toxic substances and excess acidity generated by the decay of organic potting media in the steamy conditions of stove houses! I am unable to say here whether or not it might be of any value with plants of the deserts.
Loam in nodules furnishes as mentioned an absorbent compost ingredient. A mix of 2 parts by bulk volume with 3 parts by bulk volume of 3mm+ absorptive fragments appears to maintain an adequate number of large air spaces indefinitely - under normal watering some compaction does occur, but due to a sufficiency of collisions between the fragments it stops well before all the voids have filled up. The visible air supply permanently ensured thereby is further augmented by the air contained within any unwetted or partially wetted fragments.

Accommodated in this way even the more difficult desert cacti can develop vigorous and healthy roots and beautiful top growth and maintain this over many years without repotting (other than simple potting-on) providing that the pH is kept in the requisite range and provided that a continuing supply of nutrients is not wanting.
The regulation of the pH needs a lengthier discussion than can be accommodated here. One point which cannot be ignored however concerns the use of crushed bricks. To remove the alkalinity often inherent in this material and further to reduce its pH to an optimal level each litre of graded fragments should be mixed before use with 15ml of horticultural sulphate of iron, with a little water to effect absorption of this chemical medication. This, or something chemically equivalent, is absolutely essential. Biosorb and Seramis give a neutral pH, nevertheless the results obtained with them will be greatly enhanced by giving them the same medication at half the rate, i.e. at 7½ml/l.

For further straightforward recommendations on the management of the pH our readers may be referred to the Liverpool Branch Newsletter Vol. 1 No. 1 of January 1995.