Specifically, these are the Conebushes in the Arid (Membranaceae) and Sun (Nuciferae) sections, which produce nice large, round, juicy fruit which they drop on the ground for rodents to eat. This strategy has been a contradiction at variance with all other strategies, which seem to be designed to keep seeds safe from rodent predation. Thus Pincushions and Pagodas (and Spiderheads, Sceptres and most of the minor genera) have antfruit (elaiosomes) which attract ants who take the seeds underground into the nests where they are safe from fire and rodent predation. It is well known, thanks to William Bond and Peter Slingsby, that any seeds left above ground, as occurs when the Argentinian Ant is present and eats the antfruit without burying the seed, are eaten by rodents within days. The other strategy has been adopted by Sugarbushes, many Conebushes and Featherbushes: here the seeds are protected in cones, and the seeds are only released after a fire into a landscape devoid of rodents for germination after the first good rains. And, again thanks to William Bond and co-workers, we know that natural fires at the end of summer (February- March) result in far better recruitment of these species than fires in spring or early summer, because of the much longer time these seeds are exposed to predation by rodents (and birds) before conditions are suitable for germination in autumn.
Why then should some four species of Arid Conebush and 18 taxa of Sun Conebush be so stupid as to have seeds designed for feeding rodents? The answer was simple: like Oaks, Chestnuts and other plants which produce nice large nuts, they must be designed for caching. The strategy is simple! Produce lots of nice plump seeds over a very short period. The short period means that large populations of seed predators cannot be supported and that for a short time there will be a large oversupply of fruit which the predators cannot cope with. But predators are not stupid. Some of them store these seeds for future use - they cache (or bank) them. The success of this strategy relies on the fact that such caches must be well hidden, otherwise other animals may use them. Thus our cacher goes to a lot of effort to ensure that the seeds are well hidden. And this is where the plants score. Some of the caches will be too well hidden - not even the cacher will find them again. Then the cacher may die, or be usurped from its territory by another animal, and the seeds will be safely stored in an ideal place for germination. In addition, the cacher may cache more than needed to survive the lean periods, and some caches may not be consumed before the next glut, so that they will be forgotten when the cacher starts laying his next horde.
There is only one problem with this strategy: There are no "squirrels" in Fynbos. All the rodents eat seeds and do not cache them. At least that is what was known.
Now Jan Vlok has found the elusive beast who caches! It is none other than the Cape Spiney Mouse, Acomys subspinosus. What is it? Well Smithers in Mammals of the southern African subregion has this to say of the beast:
G de Graaf (Rodents of southern Africa) says much the same.
In summary, all we knew about its habits before Jan Vlok started his investigation was that it lived on land and came out at night.
But what is the Cape Spiny Mouse? The genus Acomys contains two species in southern Africa, but some authorities recognize 38 forms including 26 distinct species. The generic name means "sharp-pointed mouse" and refers to the genus's distinctive feature of sharp, spiny (or bristly) hairs especially on the back. The Cape Spiny Mouse occurs in the Cape Floral Kingdom, (all the localities identified in G de Graaf are from Mountain Fynbos) as the race subspinosus. In Sudan, Uganda, Kenya, Somalia and Tanzania some seven other races occur. The species is replaced by the Spiny Mouse Acomys spinosissimus in Tanzania, Zambia and Zaire to Botswana, Transvaal and Mozambique. And this, apart from the description given below (and a lot more details on its dentition, skull morphology and fleas), is all that we know about the Cape Spiny Mouse.
The Cape Spiny Mouse can easily be identified by the exceptionally tough gutter hairs (the broad, channelled bristles which form the firm outercoat of rodents and give the fur "body and bounce") on the back, especially towards the tail, which give a spiny feel to the pelage. Other hairs are softer, but still bristly. The underfur is fine and sparse, and scattered with guide hairs (long bristles which just stick out of the fur - well supplied with nerves in the skin, they are used to detect movement). The animal is greyish-brown on the back, becoming lighter, and more rusty, on the flanks. There is an abrupt change from the flanks to a dirty-white colour on the underside. The hands and feet have fine white hairs. The hand has four digits, the thumb reduced to a stump with a nail (not a claw like the other digits). The foot has five digits. There are six nipples, 1 pair on the rib cage and 2 pairs on the abdomen. The animal is about 170-180 mm long, half of this being tail. The mass is about 20g. The abnormally narrowed teeth and strongly reduced third Molar are diagnostic of the species, but unless you want a nip from the incisors, don't rely on this feature except for dead specimens.