In general, the presence of ground cover plants greatly reduced weed invasion as opposed to leaving the soil bare. Although ground cover plants are not totally guaranteed to prevent weeds establishing, average figures in Table 3.12 across all plots showed that leaving the soil bare resulted in over 30 times more weeds than when ground cover plants were present.
The five best performing species of ground cover plants which resisted weed invasion best were A. inermis, C. acerosa, M. axillaris, S. mexicanum, and A. reptans. These species varied widely in terms of spread and height, suggesting that the physical dimensions of the ground cover species used were of no great importance in determining resistance to weed establishment.
Due to the presence of thick sawdust mulch, the weed species which germinated amongst the ground cover foliage were mainly species with wind-blown seeds which landed on top of the mulch. Weed seeds in the soil bank would have been prevented from germinating by the mulch. Most of these weed species have also been documented as being sensitive to light for germination (Conium maculatum (Baskin & Baskin 1990); Crepis sp. (Darwent & McKenzie 1978); Holcus lanatus
(Williams 1983); Sonchus asper, Sonchus oleraceus and Taraxacum officinale (Milberg et al. 2000)) except for Trifolium repens whichgerminates in any light conditions; and Oxalis corniculata which has very low light requirements for germination (Holt 1987). However, as Trifolium repens is less likely to be disseminated by seed carried in the wind, there was only one incidence of T. repens.
As discussed in the literature review, response to light conditions for seed germination is mediated by the ratio of red to far-red light (R: FR). It may be expected that the R:FR for the five best
performing species to be relatively low, so as to inhibit weed seeds with a light requirement for germination. A comparison of R: FR in November with the mean weed mass collected in the plots for the species A. inermis, A. reptans, C. acerosa, and M. axillaris in Table 3.14 shows that indeed to be the case. However, for S. mexicanum, the R:FR in November was the second highest in the list. This suggests that the R:FR in November was not necessarily a good predictor for weed suppression. This is because the weeds collected in November would have germinated in preceding weeks or months.
To explain the relative success of those five species of ground covers in weed suppression, one needs to examine the R:FR of an earlier time frame, such as that in June. Table 3.14 shows that the R:FR in June for three of the better weed suppressing ground cover species in the trial, namely
A. reptans, C. acerosa and S. mexicanum, were amongst the lowest in the trial group. This suggests that ground cover species which maintained a dense canopy of live foliage in winter performed best in suppressing weeds in spring.
Table 3.14 Comparison between mean weed mass and recorded R:FR.
Species
Mean weed mass (g m-2) Ratio of red to far red light (R:FR) under canopy Under foliage Open ground Jun 10 Mean Nov 10 Mean A. inermis 0.026 b 5.5 b 0.287 de 0.164 f A. reptans 0.303 ab 65.3 ab 0.089 f 0.117 f C. acerosa 0.041 b 16.5 b 0.137 ef 0.184 ef G. lanigera 2.519 ab 9.1 ab 0.328 d 0.970* a J. procumbens 3.145 ab 93.9 ab 0.223 def 0.128 f L. diffusa 4.154 ab 75.0 ab 0.868 b 0.679 bc
M. axillaris 0.017 b n.a. n.a. 0.199 def 0.122 f
O. planiscapus 1.541 ab 27.3 ab 0.182 def 0.212 ef
P. capitata 8.198 a n.a. n.a. 1.268 a 0.504* cd
P. prostrata 2.662 ab 45.1 ab 0.586 c 0.469 cd
S. mexicanum 0.000 b 45.0 ab 0.065 f 0.779* ab
V. peduncularis 2.578 ab 292.1 a 0.265 de 0.384* de
*Means are significantly different from June 2010 readings using paired t-test at 5% critical level.
An exception to the above statement is for A. inermis and M. axillaris, which had a higher R: FR in June but yet managed to resist weed invasion admirably. A. inermis and M. axillaris both had
an R: FR level in June similar to that of G. lanigera, J. procumbens, O. planiscapus and V. peduncularis, which had a mean weed mass of about 1.5-3.1 g m-2. A possible reason for the better than expected weed resistance by A. inermis and M. axillaris is that blocking of light was not the sole factor which resulted in resilience against weed. The extensively sprawling growth habit for these two species, which root regularly at stem nodes, may represent considerable competition against weed seedlings that do germinate for growth resources found underground. As pointed out in the literature review, multiple resources under strong competition results in synergistic effects for the more aggressive species (Donald 1958).
A sprawling network of rooting stems is not sufficient to compensate for poor R: FR reduction in resisting weeds. P. capitata has a similar habit and spread across the allowed space within the plot in a matter of 3-4 months. However, it had the worst weed suppressing performance, allowing 8.2 g m-2 of mean weed dry mass. This amount of weed mass was similar to the weeds collected in the open space of some plots. The dismal performance of P. capitata in preventing weed invasion can be attributed to the loss of foliage during cooler months, which meant that red light could not be blocked by the leaves, allowing seed dormancy to be broken. Given the variable amounts of weed invasion in open ground conditions, ground cover plants with a deciduous habit can be considered to produce a level of weediness akin to no ground cover planting during the period after defoliation.
Deciduous habits are not the only seasonal trait which can affect light blocking and R: FR variations under the ground cover foliage canopy. The summer-time flowering of G. lanigera, S. mexicanum, and V. peduncularis caused significant rises in the R: FR in November from June. The transmission of red light through S. mexicanum canopy nearly quadrupled from 1.2% in June to 4.6% in November, and may be expected to increase further later into the summer season. This was due to the plant producing flowering stems with less or smaller leaves, thus reducing the available surface area with red-light absorbing pigments (see Figures 3.15 - 3.17). Red light has a stimulatory effect in seeds with light requirements for germination. Despite S. mexicanum showing good weed suppression in results collected in November, this performance is unlikely to be repeated in late summer as the high R: FR in November suggests that weed seed germination will not be inhibited in the weeks through summer. Similarly, in the case of G. lanigera and V. peduncularis, the flowering period led to more red light penetration through its canopy. This may be due to overall reduction in red-light absorbing surface area as the plant positions its flowers in a conspicuous manner which may cause gaps in the foliage canopy. This leads to higher R:FR, and Table 3.14 shows that
G. lanigera had a 0.97 R:FR in November compared to 0.328 in June; similarly for V. peduncularis, the R:FR increased from 0.265 in June to 0.384 in November .
Some plant species become relatively dormant during the cooler months. This seemed to be the case for P. prostrata and L. diffusa, where the overall visible light penetration was greater in June than in November. With P. prostrata, the foliage was visibly reduced during the cooler months as noted in Tables 3.9 and 3.10. Visual ratings for foliage density of L. diffusa may have been less consistent as this drought tolerant species struggled with root rot or excessive moisture during the trial. This opening up of the canopy would have contributed to the higher R: FR in June 2010 than November 2010.
There was some difficulty encountered when assessing the mean weed mass for the
J. procumbens plots. Despite consistently high readings of light blockage and one of the lower R: FR measured, there was still considerable weed mass collected from J. procumbens plots. This is attributed to the irregular shape and plant margins of the species, which made assessment of which weeds constituted “open space” or “under foliage” positions difficult. The weeds considered to be found “under foliage” were mostly near the plant margins and may have germinated in relatively unshaded conditions.
Figure 3.15 (left) Top view of Sedum mexicanum in vegetative state Figure 3.16 (middle) Top view of Sedum mexicanum in flowering state
Figure 3.17 (right) Sample flowering stem on left side, placed next to sample vegetative stem on right side. Note the tighter whorled foliage on the vegetative stem on right side
Overall, weed suppression by ground cover plants appeared to be dependent on maintaining a dense foliage canopy capable of greatly reducing red light penetration to achieve low R:FR conditions which inhibit germination from species with a light requirement for germination. The ideal ground cover plant should therefore be an evergreen perennial with dense foliage not given to seasonal variation for dormant growth periods or reproductive phases. Where a dense foliage canopy cannot be maintained, additional competitive traits for growth resources other than light, such as having an extensive rooting system just below the soil surface, will be an advantage.