Low Caryopsis Production of the Toothbrush Grass (Microchloa altera) from Katanga (DR Congo) Could Limit the Revegetation of Trace Metal Contaminated Lands by Seeding

In Lubumbashi (Katangan copperbelt, DR Congo), atmospheric fallout from a copper smelter has contaminated soils with several trace metals (TMs) (e.g. up to 50000 mg kg^-1 Cu). For large areas downwind of the smelter, the vegetation has been destroyed and replaced by bare soils (Munyemba 2010, Mpundu 2010). These areas are subject to active erosion by rainfall and wind, which further disseminates TMs, generating several environmental and human health hazards (Banza et al. 2009). Phytostabilization appears as a suitable technique to reclaim those bare soils through the restoration of a permanent vegetation cover. This, in turn, may reduce TMs dissemination in the environment, limit human exposure, and improve the landscape (Berti and Cunningham 2000, Pilon-Smits 2005).

Toothbrush grass (Microchloa altera), a perennial bunch grass which grows only on metal enriched soils in SE Katanga (Duvigneaud and Denaeyer-De Smet 1963, Faucon et al. 2010), has excellent potential for land reclamation and phytostabilization (Shutcha et al. 2010). It has excellent survival and growth on soil artificially contaminated with Cu at 2,500 mg kg^-1 as CuSO₄ and rapidly forms dense tufts with high stabilization potential. Experimental trials have used vegetatively propagated materials, however, large-scale implementation of phytostabilization using this species will obviously require organized seed production.

In natural populations of toothbrush grass from Katanga, spikelets are the dispersal unit and their dissemination occurs three times a year (January-February, April-May, and July-August). Spikelets are 4-5.5 mm long and contain two florets, one hermaphrodite and one male. However, our preliminary observations showed that mature spikelets were often sterile (i.e. do not contain a caryopsis), a serious concern in the context of large scale seeding. In addition, we know nothing about their germination requirements. In this study, our objectives were: (i) to assess the proportion of spikelets containing a fully developed caryopsis in three populations considered as the potential seed providers around Lubumbashi and (ii) to assess their germination rates.

We collected spikelets from three natural populations of toothbrush grass located in Lubumbashi (11°39'S, 27°28'E, annual rainfall: 1200 mm, dry season: April-October) in three contrasting edaphic conditions: 1) on mine deposits at 'Mine de l'Etoile'; 2) on soil contaminated by atmospheric fallout from Cu smelter at 'Gécamines/Penga Penga'; and 3) on slag from a processing plant near the Cu-smelter. Spikelets were collected in 2007 and 2008 on mature spikes during the three dispersal periods (January-February, April-May and July-August).

In the first experiment, to assess the variation of the proportion of spikelets containing a fully developed caryopsis among the three populations, we established a collection in 2008 at the Agronomy Faculty of the University of Lubumbashi (normal soil contaminated by Cu at 500 mg kg^-1 with CuSO₄.5H₂0). Ten vigorous and visually healthy individuals were collected in each site. Each individual was divided into 20 transplants (clone) and planted the same day. The collection comprised of two blocks (3m × 38m) and each block contained three plots (3m × 10m). Each plot received individuals from the same site. Ten transplants of each individual were planted in each block as the clonal line of the individual. However, only 33% (203/600) of transplants had survived after the first year of transplantation. We collected spikelets in 2011 on all clonal lines (47) present in the collection.

In natural populations, we separated spikelets containing a fully developed caryopsis and empty spikelets based on their specific weight. This procedure is called 'spikelet concentration' in this study. Seven lots of spikelets were concentrated as follows: after separation from stalks, spikelets were sieved through a 2 mm mesh and finally spikelet concentration was performed using a ZigZag blower (Selecta Machinefabriek BV) at the ILVO (Merelbeke, Belgium). The proportion of heavy spikelets was calculated as the ratio of weight of heavy spikelets: weight of all spikelets. We manually counted the number of spikelets containing a...