|
|
| Bildiri Özetleri | |
| Ana Sayfaya Dönüş |
| ISD Ana Sayfası |
|
Stefka Alexieva, Ivanka Stoimenova "N. Poushkarov" Institute of Soil Science and Agroecology, 7 Shosse Bankya Str., 1080 Sofia, Bulgaria ABSTRACT Three years (1994-1996) field experiment with maize and natural infestation with different biological weed species was conducted on Pelic Vertisols in West Bulgaria (near Sofia). Three variants of N, P fertilization were realized with non-treated and treated by atrazine + nicosulfurone. The differences in the weed density and dry weed biomass accumulated at the non-treated and treated variants were used in the suggested by us mathematical model to determine the competition between maize and mixed type of weed infestation. The model allowed to determine with satisfactory accuracy the yield losses due to the weeds density or dry weed biomass accumulated. INTRODUCTION Crop responses to weed competition are difficult to predict, particularly in non-irrigated crop production systems (Hahn et al., 1995; Lindguist et al., 1996). In recent years, much emphasis has been placed on the development of simple, empirically derived regression models that equate eventual crop yield loss with some measurable characteristic of a weed population early in the growing season. Typically, descriptive models of crop) weed competition are valid only for the conditions from which they were developed (Weaver, 1996). Even in the basic case of one weed species competing with a particular crop, the instability of empirical yield loss models between sites and among years at the same site have confounded the usefulness of this approach (Kropff et al., 1984) Cousens et al., 1988; Lindguist et al., 1996). In addition to the impact of growing season weather factors that influence the interactions between specific crop and weed combinations include crop and weed densitities, relative times of emergence and edaphic characteristics (Firbank et al., 1990; Kropff & Spitters, 1991; Wilson et al., 1995). Such models can be used to integrate and to assess the relative importance of the multiple influences on competition. They may also give insight as to why specific responses are evident in the field only in certain years and locations, while providing a powerful tool for quantifying the longterm behaviour of specific crop weed competition models are often based on existing ecocrophysiological models for plant growth (Weaver, 1996). Kiniry et al., (1992) developed the ALMANAC model to provide a practical, easily adopted tool for simulatig competition in mixed plant communities. ALMANAC requires a relatively small number of species - specific plant parameters and is considered of intermediate complexity (Debake et al., 1997). When early maize growth was not compromised by water stress, the maize canony overtopped the A. theophrasti canopy, relegating the latter to a subordinate position. Generally, when early season precipitation exceeded 100 mm, predicted loss was low (McDonaldt & Riha, 1999). In this paper, the dynamics of competition between maize and A. retroflexus, C. album, E. crus-galli, S. glauca, and C. bursa pastoris are studied in order to determine which are the critical phases of maize development as affected by the weeds in relation to N, P application. The following aspects were studied? (1) Does mixture of weeds lead to yield losses of maize in the different regime of N, P application? (2) Which aspect of climate is responsible for structuring the maize response to weeds competition? (3) If climate date can be used to improve our capacity to rationalize pre-and post-emergence weed control decisions. MATERIAL AND METHODS A field experiment was carried out during the period 1994-1996, in Sofia district on pelic vertisoils. The content of humus was in soil depth 0-28 cm - 3.43% and 3.65% in depth 28-60 cm. The available total N and P in the layer 0-28 cm were 0.23% and 0.082%, in the layer 28-60 cm the total N and P are 0.145% and 0.082% respectively. The previous crop was Triticum astivum L. harvested for grain. Three years monoculture of maize randomized in four replications with plot size 200 m2. The following scheme was used: Factor A - treated with N, P fertilizers; a1 - N0P0; a2 - N50P40; a3 - N70P60; Factor B - weed control; b1 - hand - weeding; b2 - atrazine 1500 g a.i. ha -1 + nicosulfurone 480g a.i. ha -1. The atrazine was applied pre-emergence of the maize for selective control of broad-leaved weeds and nicosulfurone was applied of 3-5 leaves of maize and 1-3 leaves of grass weeds. The hand-weeding of the control and herbicide treatments was made in the growth stage "7-9 leaf". In all trial treatments according to the factors A and B, weed spp. and weed dry biomass was determined as number m-2 and g m-2. The dry weight was obtained after 54 hours drying at 80 °C. The maize field was not irrigated; phosphorus fertilize was applied during the main soil ploughing in the autumn and nitrogen fertilizer was applied with the last presowing cultivation. The crop was sown between May 5 and 10. Maize density was 47000 plants ha-1 spaced 70 cm between the rows. The results from the experiment that we made during the period 1994-1996 are presented in table 1. The real yield loss  are calculated as follows: = ( 1 - Y1 / Y2 ) 100 (%) (1) Description of the mathematical model : The mathematical interpretation of the dependence between and  A (respectively D) is based on the following premises:1. The functional relationship between the depended quantity is analyzed for years with approximately equal climate conditions for a culture (in our case the period 1994 and 1996); during 1995 when the sum of the precipitation's is above 100 mm/m2, then it is necessary to introduce a correction coefficient, definition of limit values of that climate factor. 2. The dependence between and A (or D) according to the date from table 1, it is possible to approximate precisely using a parabola of the type Y = a.X2 (the branch in area +x, +y is valid), here the depended quantity Y is identified with A (or D); X - with and the independent parameter in this function "a" is the quantity - 1/A12 (or 1/D12).3. The graphic description of the parabola is presented in fig.1.   There is an analogy between the geometrical place of parabola points and the experiment data for , A (or D); and A1 (D1) so with increasing of A (or D) and A1 (D1) the parabola branch is opening and the values of X increasing, respectively the losses increase. These three premises define the following relation between the losses A and the quantity dry weed's biomass A (2)and the degree on weed density D  (3)The coefficient M reflects the climate conditions and has the following values: - When the sum the precipitation for the period < 100 mm/m2 - M=1; - When the sum the precipitation for the period > 100 mm/m2 - M=0.1. The expression (2) and (3) give the mathematical interpretation of the dependence between the yield losses in [%] from the degree of weed density or the dry weed's biomass. The results from the calculated values of A and D compared with the real yield losses are presented in table 2.
DISCUSSION The degree of infestation and the quantities of the weed dry biomass in the control variant depended on the level of mineral fertilization (Table 1). During the three year period of investigation these two parameters were highest in the variants with the highest norms of fertilization and their values were lower in the variants without fertilization. During the period "sowing - 7-9 leaf" the highest degree of infestation showed Amaranthus retroflexus L. in hand weeding variant. The obtained participation part (in percent) of A. retroflexus in the total degree of weed infestation varied from 43.1 to 65.2, depending on the norms of mineral fertilization and precipitation. The weed seeds quantity in the soil bank, soil type and precipitation in the first 20-30 days after the pre-emergence treatment influenced generally on the soil herbicides effect. The treatment with atrazine 1500 g a.i. ha-1 pre-emergence +nicosulfurone 480 g a.i. ha-1 in phase "3-5 leaf" of maize combination killed all annual weeds in the applied 3 variants of N, P fertilization in 1996.  This effect was due to the optimal water content (75-80% from FC) in the period "sowing -7-9th leaf" of the maize (36 days) and relatively low precipitation (49,2 mm) in the same period. The larger precipitation (122.8 mm) in the same period of 1995 significantly decreased the herbicides mix effect. As a result the herbicides moved on and under the soil surface, weeds survived (in different percentage) in all of the variants of treatments. The differences of the soil herbicides effect in the beginning of the crop vegetation allowed us to determine the competition between the maize and weeds as a difference between the non-treated and treated variants. The best weed control was observed in treatments with the highest application of mineral fertilization and that control was less in variants without mineral fertilization (Table 1). In 1996 when the precipitation was 49.2 m-3 in combination with high effective temperature sum (ETS), was observed the highest weed control in all of the treatments. The herbicide application influenced strongly of the dry weed biomass quantity in comparison to the infestation degree. The differences in the weed density and the dry weed biomes accumulated in the non-treated and treated variants were used in the suggested by us mathematical model to determine the competition between maize and mixed weed interactions. The object of the model was the mixed type of weed infestation, because usually the crop infestation was with different weed species. The data of the non-treated and herbicide-treated 3 variants of N, P fertilization were the base of the model. The model allowed to determine with satisfactory accuracy the yield reduction due to the weeds density or the dry weed biomes. The lower differences between real measured and predicted yield losses may be due to the decreased crop: weed competition for water in the year with more precipitation (1995) .In favorable early season weather allowed maize to dominate the weed mixed canopy and maize crop did not incur substantial yield losses regardless the weed density or dry biomes. This prediction in consistence with other studies that have documented the importance of water in early season growth for structuring the competitive relationship between crop and weed population. When early maize growth was not compromised by water stress, the maize canopy overtopped the weed mix infestation, relegating a latter subordinate position. These analyze support the field observation that maize yield response to A. theophrasti density is inherently variable in rained environments (Hahn et al., 1995; Lindguist et al.,1996; McDonald & Riha, 1999). Excluding the precipitation in early maize growth season, mathematical models of yield losses (e.g. density or dry biomass dependence) are unable to capture the variation in maize response to mixed weed type competition. In this study the mathematical models predicted lower differences between real founded and predicted yield losses ±1%. The predicted yield losses differed more from the real once in the dryer 1996, when the competition for water between crop and weeds probably increased. Conversely, when early maize growth was compromised by water stress, the mixed weed population was highly competitive with maize and correlated with maize yield losses. In unfavorable early season weather, the mathematical models allowed to determine with satisfactory accuracy the yield reduction. The real and predicted yield losses reached ±7% when was used weeds density and ±3.5% when was used dry weeds biomass. REFERENCES Cousens R, Firbank LG, Mortimer AM & Smith RGR (1988) Variability in the relationship between crop yield and weed density for winter wheat and Bromus sterilis. Journal of Applied Ecology 25: 1033-1044. Debaeke P, Caussanel JP, Kiniry JR, Kafiz B & Mondragon G (1997) Modelling crop, weed interactions in wheat with ALMANAC. Weed Research 37: 325-341. Firbank LG, Cousens R, Mortimer AM & Smith RGR (1990) Effects of soil type on crop yield-weed density relationships between winter wheat and Bromus sterilis. Journal of Applied Ecology 27: 308-318. Hahn RR, Degni JG & Mt Pleasant J (1995) Leaf cover measurement for velvetleaf thresholds in field corn. In: Proceedings Northeast Weed Science Society, 49: 94. BOSTON. Kiniry JR, Williams JR, Gaddman PW & Debaeke P (1992) A general, process-oriented model for two competing plant species. Transactions of the ASAE 35: 801-810. Kropff MJ & Spitters CJT (1993) A simple model of crop loss by weed competition from early observations on relative leaf area of the weeds. Weed Research 31: 97-105. Kropff MJ, Vossen FJH, Spitters CJT & De Groot W (1984) Competition between a maize crop and a natural population of Echinochloa crus-galli (L.). Netherlands Jornal of Agricultural Science 32: 324-327. Lindquist JL, Mortensen DA, Clay SA et al. (1996) Stability of corn (Zea mays) - velvetleaf (Abutilon theophrasti) interference relationships. Weed Science 44: 309-313. Mc Donald AJ & Riha SJ (1999) Model of crop: competition applied to maize: Abutilon theophrasti interections. II. Assessing the impact of climate: implications for economic thresholds. Weed Research 39: 371-381. Weaver SE (1996) Simulation of crop-weed competition: models and their applications. Phytoprotection 77: 3-11. Wilson BJ, Wright KJ, Brian P, Clements M & Stephens E (1996) Predicting the competitiove effects of weed and crop density on weed biomass, weed seed production and crop yield in wheat. Weed Research 35: 265-278. |