Sibudu Redux: Sedges and Rushes! Oh, my!

[I'm re-posting this under the expanded title, on the assumption that the faithful might think 'Redux' meant 'more of the same.' Since this is a substantive addition to my earlier post, I want to make certain everyone has had a chance to see it. So, ahead to the pasture!]

I think that if I see one more headline screaming 77 kyr old mattresses, whether insecticidal or not, I'll come down with a bad case of hay fever. After I suggested an alternative explanation for the presence of leaf mats at Sibudu 'Cave,' Gerrit Dusseldorp left the following comment on SA: 

'I enjoyed the post, but ... I think you have missed an important part of the argument. Namely, the younger part of the bedding in the sequence ~58 ka consists of sedges (Cladium mariscus according to the paper) characteristic of wetland environments. They would have to have moved up from the river valley to the abri. Hence, your alternative scenario does not hold for these deposits.'

As you might imagine, I beg to differ. It is simply false to claim, as Wadley et al. do in the Science article, that  

Cyperaceae (sedges)...and Juncaceae (rushes)...are normally plants of wet habitats and could not occur naturally within the dry rock shelter.

They should have said 'a dry rock shelter.' That would have been a statement that could endure scrutiny. However, nothing says that the rockshelter has always been a place bereft of ground or surface water. 
     In fact, team archaeobotanist Robyn Pickering is careful to point out (and unable to adequately account for) the rather enigmatic, but nonetheless frequent occurrence and relative abundance of gypsum formation in the Sibudu sediments. According to the excavators, this mineral could not have remained in nodular form in the presence of significant moisture--it would have dissolved. Thus, they say, it's a dry place. True enough. But how did the gypsum form in a place too dry for it to degrade? Pickering suggests that at some time cool coastal mists may have reached as far inland as the rockshelter and produced the circumstances necessary for gypsum to form in the deposits. But really, gypsum needs the prolonged presence of soil moisture to form. It's an empirical question whether or not fog would do the trick. So I'd have to say that the jury is still out. Unfortunately for the defense team, several witnesses for the plaintiff have come forward who can explain the presence of gypsum, as well as the hydrophilic plants that are the cause of so much commotion. 
     I'd hasten to add, however, that although running water can be ruled out as a significant contributor or disturber of the sediments at Sibudu, the possibility of ephemeral ponding can not be so easily dismissed. As I pointed out in my first post on Wadley et al.'s claims, the east to west slope trend is downward toward the rockshelter wall throughout most of the sequence. And, although the north to south trend is also downslope, the lay of the land near the wall would not preclude accumulation of water at certain times, which would explain formation of gypsum and the presence of hydrophilic plants.
     How so? A watercourse flows nearby, the riparian vegetation of which would be a perennial source for seed. Indeed, sedge and rush seeds are found throughout the sequence at Sibudu. Even if standing water were fairly short-lived, you can't rule out germination and growth to maturity of rushes and sedges within the dripline of the shelter. Its aspect is westward--plenty of warmth and sun to nurture the growing plants. The hypothetical standing water would only need to have persisted for a couple of months to play host to a thriving, but spatially limited, community of rushes and sedges. As for these genera being unable to survive in a tenuous hydrological regime such as an ephemeral pond, there is reason to believe that the species involved may not require permanent or even abundant water to survive.
     In fact a quick search of my friend Google gleaned an interesting brochure published in September of this year by the Botanical Society of South Africa's KwaZulu-Natal Coastal Branch, a portion of which is reproduced below. It's a brief description of the Society's outing to the ocean-side sand dunes near Durban's Beachwood Mangroves Nature Reserve.

The beach and dunes revealed more common coastal species eg. ... Juncus kraussii

As an example, have a look at one natural habitat of J. kraussii in the photo reproduced below. J. krausii is found in abundance in the sediments excavated at Sibudu Cave.

Juncus kraussii at home in South Africa. Note the distance that these plants are removed from the nearest source of water. Granted, they may not grow to 1.5 m as their healthier neighbours nearer the water. But they seem to manage allright.

In the same brief web search that yielded information about the tolerance of rushes for less-than favorable moisture regimes, I managed also to discover the following about sedges, published by the South African National Biodiversity Institute.  

Many species [of Cyperaceae] are deciduous, and survive the unfavourable season as rhizomes, corms or tubers. Several species grow in semi-arid areas, and are able to survive periods of drought due to succulent, water-storing leaf sheaths. In arid areas many species have overcome the problem of temporary moisture (such as in pans) by becoming annuals, completing their life cycles in a month or two.

As I am not an archaeobotanist I'm unable to say whether or not the Cyperaceae found at Sibudu fall into the same category as the ones described above. However, I do think I've made it clear that there's evidence for considerable genotypic variation and phenotypic tolerance of less-than perfect growing conditions within the numerous different species of rushes and sedges that grow in the region, including the rush, J. kraussii, that was recovered at Sibudu.
     The question really comes down to one of probabilities. What is the probability that gypsum formed at Sibudu on account of fog? What is the probability that, over the last 77 kyr, ephemeral ponds or moisture catchments formed at times in the low-lying areas of the large rockshelter, and the normally water-loving rushes and sedges, whose seeds accumulated in the shelter in abundance, managed to eke out an existence, like J. kraussi in the Durban dunes? What are the probabilities? What, indeed?

[Addendum 2011 December 30: I've heard from Gerrit Dusseldorp, the worker who prompted me to extend this mattress critique. He says I've convinced him. One down. Untold hundreds to go!]