54 
IAN D. LUNT 
the latter 11 species per quadrat. Except for one 
soil sample which could not easily be sieved, 
samples were passed through a 4 mm 2 sieve to 
remove coarse material and recognizable veg¬ 
etative propagules, particularly corms of Rom- 
idea rosea. A 10 mm deep layer of each soil 
sample was spread over potting mix in five seed¬ 
ling trays, each 8.96 x 10~ 2 m 2 in area, except for 
two quadrats for which sieved soil was sufficient 
for only four trays. Five control trays were filled 
only with potting mix. Trays were randomly 
arranged in an unheated glasshouse for 151 days, 
from 13 January to 24 May 1987. They were reg¬ 
ularly watered and seedlings were removed as 
identified. Densities of viable seeds were com¬ 
pared statistically by Students t-test (Sokal & 
Rohlf 1981). 
Danthonia and Isolepis seedlings were iden¬ 
tified only to genus, and in three cases pairs 
of species were not distinguished (Table 1). 
Cyperus lenelhis and Schoenus apogon were 
noted in the field but not as glasshouse germi- 
Species 
Density (viable seeds per 
m 2 ) 
Species-poor 
Species-rich 
mean 
range 
mean 
range 
* Vulpia bromoides 
3199 
268 
10848 
15087 
8796 
23950 
*Romulea rosea 
1483 
455 
3103 
2711 
170 
9497 
*Aira cupaniana 
128 
56 
170 
1759 
45 
3509 
Juncus bufonius/capitalus 
805 
112 
1786 
319 
27 
786 
Themeda triandra 
769 
447 
1375 
139 
98 
241 
*Briza maxima/minor 
81 
0 
313 
430 
9 
1149 
Wahlenbergia sp. 
232 
152 
304 
208 
45 
304 
* Cen laurium temiiflorum 
221 
71 
536 
186 
45 
509 
Isolepis spp. 
142 
22 
277 
199 
0 
777 
*Sagina procumbens 
152 
0 
591 
186 
54 
509 
Danthonia spp. 
26 
9 
54 
253 
100 
643 
*Cicendia quadrangularis 
47 
0 
170 
147 
36 
420 
Calocephalus citreus 
9 
0 
27 
172 
0 
500 
Crassula decuinbens 
46 
22 
89 
103 
0 
313 
*Gnaphalium purpureum 
39 
9 
67 
110 
0 
402 
*Hypochoeris radicata 
73 
45 
125 
53 
0 
107 
Oxalis sp. 
73 
9 
212 
36 
0 
98 
Juncus subsecundus 
9 
0 
27 
70 
0 
201 
* Trifolium campestre/dubium 
32 
9 
71 
7 
0 
11 
*Cerastium glomcratum 
16 
0 
54 
16 
0 
36 
*Leontodon taraxacoides 
14 
0 
36 
10 
0 
27 
Deyeuxia quadriseta 
3 
0 
11 
24 
0 
67 
*Lolium rigidum 
12 
0 
22 
11 
0 
44 
* Vicia sp. 
2 
0 
9 
7 
0 
18 
Helichrysum apiculatum 
2 
0 
9 
3 
0 
11 
* Trifolium subterraneum 
131 
0 
473 
* Bromus hordeaceus 
84 
0 
170 
* Trifolium angustifolium 
7 
0 
27 
*Sonchus oleraceus 
15 
0 
22 
Lythrum hyssopifolium 
2 
0 
9 
Leptorhynchos squamatus 
14 
0 
54 
Eryngium ovinum 
5 
0 
18 
*Hainardia cylindrica 
3 
0 
11 
Con volvulus erubescens 
3 
0 
11 
*Plantago coronopus 
3 
0 
11 
Solenogyne dominii 
2 
0 
9 
*Erodium sp. 
2 
0 
9 
Hypericum gramineum 
2 
0 
9 
Table 1. Mean and range of number of viable seeds per m 2 recorded from species-poor and 
species-rich T. triandra grassland. Species nomenclature follows Forbes & Ross (1988); 
asterisks denote exotic species. 
