Impact of creeping vegetable cover crops on weed flora composition during the early growth of Hevea (Natural rubber) saplings in a tropical rain forest zone

Weed management is amongst the most expensive agronomic practice in immature rubber plantation. Field experiment was conducted during the rainy seasons of 2004 and 2005 at the Rubber Research Institute of Nigeria. The purpose was to assess the impact of some creeping economic vegetable crops (vegetable cowpea – Vigna unguiculata L. Walp. ssp sesquipedallis, Egusi melon – Cucumeropsis manni Naaud., Fluted pumpkin – Telfairia occidentalis Hook. F and Pumpkin – Cucumbita pepo) compared to the conventional cover crop (Centrosema pubescens) on weed flora composition when grown with rubber saplings. The experiment was laid out in a randomized complete block design with four replications. Weed flora comprising of 33 species under 17 plant families was observed across the field. Panicum maximum, Elusine indica, Brachiaria lata and other grass and sedge species dominated in the conventional cover crop plots. While ephemeral weed species like Talinum triangulare, Peporomia pellucida, Acalypha cillata and other broad leave weeds dominated in creeping vegetable crop plots, except the Telfairia plot which had weed flora closely similar to those of the conventional cover crop plots. The weed spectrum, density and dry matter weight were significantly lower in the Curcubita pepo and Cowpea plots compared to the conventional cover crop plots. The effect of the cover crops on the plant height and stem girth did not differ significantly within the first 12 months. Thus, it could be said that with the exception of the Telfairia occidentalis, the creeping vegetable crops could be of importance for the control of noxious weed flora in immature rubber. Key word: density, growth performance, intercrop, species, spectrum Introduction Regions of optimum natural rubber production lie within the equatorial belt of the low land humid tropics. The climatic conditions (adequate rainfall distribution, less fluctuation in temperature and relative humidity) throughout the year in the region promote Journal of the Rubber Research Institute of Sri Lanka, (2009) 89, 33-45 *Rubber Research Institute of Nigeria, P. M. B 1049, Benin City, Edo State, Nigeria.


Introduction
Regions of optimum natural rubber production lie within the equatorial belt of the low land humid tropics. The climatic conditions (adequate rainfall distribution, less fluctuation in temperature and relative humidity) throughout the year in the region promote The preponderance of weed does not only interfere with the crop's utilization of above and below ground resources, but also with agronomic practices, especially during the first few years of plantation establishment (Kuan et al., 1993;Esekhade et al., 2004). However, the extent of interference and damages varies with species. Kuan (1993), classified various weed species found in immature rubber plantation into beneficial, acceptable or tolerable and noxious ground covers, based on their effect on the young rubber and field management.
Studies (Teoh et al., 1982;Esekhade et al., 1996) have shown the cost effectiveness of chemical weed control in immature rubber fields relative to manual. However, total reliance and/or repeated application of herbicides does not only result in rapid up surge of some more deleterious and chemical resistance species (Teoh et al., 1982), but also constitute environmental hazard (Moss and Rubin, 1993).
The use of leguminous cover crops as live mulch in weed control in rubber plantations has been found to be environmentally friendly and economically viable among large estate holder. To the smallholder farmers who accounts for over 75% of the global natural production (IRSG, 2000), the incorporation of leguminous cover crops into their cropping system is still unacceptable. This is mainly due to lack of any monetary or physical benefits in terms of food supply or else.
There are some economic creeping vegetable crops of the humid tropical origin that possess similar growth characteristics like most of the commonly used creeping legumes in plantation crop management, except for nitrogen fixation. Documented evidence of the economic values of some of the creeping vegetable crops could be found in Schipper (2000). Clay and Oberle (1996), Akinyemi et al. (1997) and Nwagwu et al. (2000) have in separate locations and studies reported on the weed control efficacy of some creeping vegetable cover crops in arable crop field when used as live mulch. There are paucity of information on the use of creeping vegetable crops of economic importance in the control of weed in rubber plantations. Since the work of Kuan (1993) has shown that degree of damage and interference of weed in immature rubber plantation depends on the species composition, it is the objective of this study to assess the impact of some creeping vegetable cover crops (Egusi melon, pumpkin, Telfairia and vegetable cowpea) on weed floral composition relative to the conventional cover crop (Centrosema pubescens) during the first 14 months of Hevea plantation establishment.
Weed flora composition under creeping vegetables

Materials and Methods
Field experiment was carried out during the raining seasons of 2004 and 2005 at the Rubber Research Institute of Nigeria, Iyanomo. Iyanomo is near Benin City and lies between latitude 6 0 00 1 and 7 0 00 1 N and longitude 5 0 00 1 and 6 0 00 1 E. The rainfall pattern is bimodal with an annual average of 2516mm and a mean maximum and minimum temperatures of 31.5 and 23.6 0 C, respectively. The soil is acid, derived from coastal plain and it is described as being very deep and well drained Typic Kandiult -an Ultisol (Ojo-Atere et al., 1987). The surface soil (0 -20 cm depth texture is loamy sand with low level of exchangeable bases (Table  1).
Land preparation for the experiment involved manual clearing of the three-year old vegetation followed by burning. Each plot measured 136.68m 2 with six stands of budded stumps (NIG 801 clones) being planted at the spacing of 6.7 x 3.4m. The treatments consisted of rubber saplings inter planted with: a These intercrops were compared to the conventional rubber sole (with Centrosema pubescens). The experiment was laid out in a randomized complete block design with four replications, thus occupying a total land area of 0.23ha.
The creeping vegetable crops were planted at the onset of rains (April) each year at the spacing of 1m x 1m and at 1m away from each rubber stand. The rubbers (budded stumps) were introduced three weeks later following the stabilization of rains. Manual hand picking of weeds from each plot was carried out once (at four weeks) after planting of the cover crops.
At two, four and six months after planting of the budded stump in 2004 and at similar periods after the second season (2005) planting of cover crops, weed samples were collected. The weed sampling in 2005 coincided with the 10, 12 and 14 months after planting of the budded stumps.
Weed samples in each sampling period were taken from 1m x 1m quadrats systematically laid down at five different spots in each plot for the determination of weed spectrum and density. At the Laboratory, weed samples were separated by species and identified. The composition of the weed flora across the various treatments, was analyzed by calculating the average relative abundance of each species for each year using the mean of the sample period as: Relative abundance = (relative density + relative frequency)/2. (Derken et al., 1993). Where, Relative density = number of individuals of a given species within the quadrats per treatment divided by the total number of individuals of all the species within the plot and.
Relative frequency = proportion of quadrats in which the species occur per plot divided by the total frequency values of all the species.
Dry matter weight of the weeds was determined after being oven dried at 85 0 C for 46hrs. The weed spectrum and density were determined by counting the number of individual weed species and number of weed stands, respectively per quadrat. The effect of the creeping vegetable crops on the Hevea sapling was determined by measuring the height and stem girth (10cm above the bud union) of the plants at two, four, six, nine, twelve and fifteen months after planting. Data collected on weed density, dry matter and spectrum, and Hevea growth performance across the treatments were separately analyzed and means compared using the LSD test at 5% level of probability.

Effect of creeping vegetable crops on weed species composition, spectrum, density and biomass (dry matter weight):
Weed communities made up of 33 species under 17 plants families were observed across the treatments throughout the study period. Majority (mainly grasses and sedges) was found in the conventional cover crop plots ( Table 2). The relative abundance of individual species differed widely with the type of creeping vegetable crop planted across the sampling periods.
The weed species that had the highest relative abundance value in the conventional cover crop plot were Panicum maximum and Brachiaria lata during the initial six months after planting (2004), while species like Elusine indica and Optismenus bumanic were the most prevalent between the 10 th and 14 th MAP. The result from rubber/telfairia plots followed similar trend, except that the relative abundance value of Cyprus rotundus with a mean value of 25.7% between the second and sixth MAP, exceeded those of Brachiaria lata as observed in the conventional cover crop plots at the initial six MAP. Generally, the relative abundance of majority of weed species observed in this study dropped at the 4 th and 12 th MAP and later increased slightly at the 6 th and 14 th MAP in that order.
Across the creeping vegetable cover crop treatments, with the exception of the Telfairia plots, the weed species that had the highest relative abundance value were Talinum triangulare, Chromoleana odorata and Acalypha cillata (Table 2). However, between the 10 th and 14 th MAP samplings period, Chromoleana odorata was no longer recorded. High relative abundance values for other weed species like Commelina diffusa and Braciaria diffusa were recorded in the Rubber/Cucurbita pepo and Rubber/Telfairia plots in that order.
As shown in Table 3, the number of weed species per quadrat (spectrum) was significantly higher in conventional cover crop plots relative to those in Rubber/Cowpea and Rubber/Cucurbita pepo at two and four months after planting (MAP) sampling periods. At the period of six and 14 MAP, all the creeping vegetable cover crop plots recorded a significantly lower weed spectrum relative to the conventional cover crop plot.
The effect of the creeping vegetable cover crop on weed density relative to the conventional one is presented in Table 4. The weed density (number/m 2 ) narrowed significantly between the conventional cover crop plots and those of the creeping vegetable crops except for the plots under Telfairia, which did not differ significantly from the conventional at the second and fourth I K Okore and F A Akpobome MAP. However, from the sixth MAP, the values observed from the Telfairia plots were significantly lower compared to the conventional cover crop plots. The effect of the cover crop on weed dry matter ( Table 5) mirrored those of the weed density, with the exception of the values recorded at 14 th MAP. At that period (14 th MAP), the weed dry matter weight recorded in the Rubber/ conventional cover crop and those of Rubber/Telfairia and Rubber/Melon, did not differ significantly.

Effect of creeping vegetable crops on the performance of Hevea saplings:
The effect of the creeping vegetable cover crops compared to conventional one, on rubber sapling height did not differ significantly during the first nine months of growth. From the 12 th month, rubber sapling grown with either cowpea or Curcubita pepo had significantly higher heights relative to the conventional cover crop plot ( Table 6). The result on the effect of the treatments on the plant stem girth followed similar trend except that significant difference only occurred at the 15 th MAP.

Discussion
Preponderance of grasses (Panicum maximum, Elusine indica) and sedges (Cyprus rotundus0) were recorded in the conventional cover crop plots relative to the creeping vegetable cover crop plots ( Table 2). The highest percentages of these species were recorded during the initial two MAP. This may have been due to the initial vegetation of the fallow resulting from the age. Higher dominance of such weed species are common, among short fallow land (Akobundu, 1993) and their subsequent transfer to cultivated fields after fallow removal (Smith and Akinade, 2000). However, with their higher relative abundance in the conventional cover crop plots compared to creeping vegetable cover crop plot, it could suggest that certain modified micro environment like reduced light intensity, temperature, etc. may have occurred due to the growth of the creeping vegetable crops. This may have affected the emergence of the grasses and sedges. This is line with the observations of Nwagwu et al. (2000) and Blackshow et al. (2001). Nwagwu et al. (2000), observed the potency of cucurbita pepo in reducing the population of certain grass weed species relative to melon in an intercrop system in Southwestern Nigeria. This they attributed to the wide spreading habit and vigour of the Cucurbita pepo.      Nyctaginaceae  Teasdale (1993) reported that thick cover crop stands are capable of reducing light transmittance to weed seeds, thus, reducing their germination. This could have been responsible for the reduced weed density; spectrum and dry matter (Tables 3, 4

and 5) recorded in
Cucurbita/Rubber and Cowpea/Rubber plots, relative to other treatments. These two creeping vegetable cover crops had the highest vigour, rapid establishment and area of spread (cover) compared to the other creeping vegetable crops evaluated. *Rubber + Centrosema pubescens The inability of the Telfairia occidentalis to perform like other cover crops could be ascribed to its sensitivity to grasses, which may have affected its growth vigour at the early stage (two MAP).The most predominant weed species under Telfairia/Rubber were grasses like Panicum maximum and Brachiaria lata. The sensitivity of this creeping perennial plant to grasses has been reported by Esiaba (1982), more especially when allowed to spread on the ground without being staked, as was the case in this study.
The lower values in most weed species relative abundance, spectrum, density and biomass recorded in the Cowpea plot relative to others with exception of Cucurbita pepo plots, is similar to the observation made by Okoh et al. (2004). They observed that cowpea varieties with vigorous spreading habit, as is the case with the sub sp. sesquepidallis are capable of reducing weed spectrum and density. In addition, White et al (1989) reported, that the potentials of annual legumes like Cowpea in weed control does not only depend on the spreading ability but, also on some allelophatic chemical contained within the plant that suppresses weed germination and growth vigour.
The consistently lower relative abundance values in individual weed species, spectrum and density recorded during the fourth and 12 th MAP sampling period, across the treatments could be ascribed to the shading effect of the cover crops. The period coincided with the peak of vegetable growth and spread of the cover cops in each year. This reflected on the weed dry matter (biomass) weight recorded across the treatments ( Table 5). Majority of the weed species observed in the conventional cover crop plots fall within the group Kuan (1993) classified as noxious weed species under immature rubber. Thus, suggesting that with the use of creeping vegetable cover crops with vigorous spreading habit like the vegetable Cowpea (sub sp. sesquepidallis) or Cucurbita pepo, most of the noxious weeds could be controlled.
As shown in Table 6, the effect of the creeping vegetable on plant height, and stem girth within the first 10 months of the plant growth did not differ significantly from those conventional cover crops. The significant difference noticed from the 12 MAP for height and 15 MAP for girth with Cucurbita pepo given higher values is in conformity with other reports. This could be as a result of prolonged growth of some of this vegetable cover crop, which ensures higher soil moisture retention due to its massive mat-layer formation on the soil surface (Ayeni, 1993). Any crop left in the field before and after the drying up of the cover crop, benefits from the conserved soil moisture (Nwagwu et al., 2000).