CORRELATION BETWEEN CLAY CONTENT AND SEED DEPTH IN NO-TILLAGE CORN (Zea mays L.) CORRELAÇÃO ENTRE O CONTEÚDO DE ARGILA E A PROFUNDIDADE DE DE- POSIÇÃO DAS SEMENTES MILHO (Zea mays L.) EM PLANTIO DIRETO

The region of Campos Gerais, in the State of Paraná, is leader in grain yields and state-of-the-art no-tillage farming. The widely adopted no-tillage system tends to increment soil variability. There are several studies about spatial variability of soil characteristics affecting grain yield, but not enough attention has been given to the variability resulting of human actions. The objective of this study was to evaluate the correlation of soil clay content and the depth of placement of corn seeds on areas under no-till systems management. The four selected cornfields areas for the study are property of local farm cooperatives associates. The points for evaluation were defined considering the local mapping of soil texture. Clay content varied from 94 to 489 g kg-1 on Plot 1; from 222 to 414 g kg-1 on Plot 2; from 269 to 509 of rows to be considered for plant distribution and seed depth analyses. The coefficient of variation (CV) of plant distribution was between 23 and 56%. For seed depth, the CV was between 18 and 34%. The regression analysis showed high coefficients of determination (r2) for plots 1 and 2 (r2=0,85 and 0,83). The clay content was generally higher on plots 3 and 4. In this case, the analysis of variance was not significant, and the coefficients of determination were low (r2=0,22 and 0,01). Results indicate that clay content values may be used to delimit g kg-1 on plot 3; and from 368 to 698 g kg-1 on plot 4. The type of planter determined the number management zones on the field, where the depth of seed placing in the planting process can be regulated in different ways.


INTRODUCTION
The region of Campos Gerais, in the State of Paraná, is leader in grain yields and state-of-the-art farming technology.No-tillage system is widely adopted, and precision agriculture techniques are gradually being established.Precision Agriculture (PA) is defined as a production management system; it is based on process optimization, and considers spatial variability and correlated factors (MOLIN, 1997).
The division of the crop field in smaller areas, or management zones, based on variables that affect grain yield is an alternative to minimize the effects of variability.Some studies show a correlation between grain yield and soil physical characteristics (SWAN et al., 1987;BAKHSH et al., 2000).
The planting process is a source of variability, and the control over this process is a form of PA accessible to most farmers.Accuracy and precision during the planting process increase production rates and reduce the input Palavras-chaves: Textura do Solo, Processo de Semeadura, Zona de Manejo.
of seeds and fertilizers.
The structure of the plant is important in the interception of solar radiation, and is a direct result of plant population and distribution on the field (MEROTTO JUNIOR et al., 1998, andFANCELLI &DOURADO NETO, 2000).SILVA & FREITAS (1994) observed that the distance measured in number of plants was more important than in metric units for pollination effectiveness.
In a study of the planting process, LIU et al. ( 2004) observed better correlation of corn yield to emergence variability, than to plant distribution variability, do not forgetting that problems in plant emergence can provoke a decrease on plant population.Several studies describe some form of correlation between corn yield and heterogeneity of emergence (NAFZIGER et al., 1991;MEROTTO JUNIOR et al., 1998).The temporal dissimilarity of plantlet emergence may cause losses up to 3.0 Mg per hectare (MEROTTO JUNIOR et al., 1998).
The higher the depth of seed placement, the higher the consumption of energy for emergence, implying negative influences by the lower temperatures of the soil and lower levels of oxygen; conversely, the lower the depth of placing implies higher susceptibility of the seeds to water deficits.Plants emerging at distinct speeds, in addition to the initial stress, are likely to face lower photosynthetic rates due to shading or having its pollination process affected (NAFZIGER et al., 1991).
Positive linear correlation between corn seed depth and emergence time in favorable temperatures was observed by GUPTA et al. (1988).PRADO et al. (2001) did not find significant differences in corn emergence speed at varying seed depths, with water supply.YORI-NORI et al. (1996) found inverse proportionality between seed depth and emergence speed of popcorn seeds.
The literature reports several interesting planting depths; FANCELLI & DOURADO NETO (2000) state it is 3 to 5 cm for clay soils, and 4 to 6 cm for sandy soils.WEIRICH NETO (2004) found the "mathematical optimum" seed depth; seeds placed above or below the optimum took longer to emerge.
Seed depth is an important variable, though very difficult to control.SATTLER (1992) has developed a device to control seed depth at planting.In no-tillage, there are common impediments such as residues, irregular surface and large variability of soil resistance (JANKE & ERBACH, 1985;MORISON JUNIOR & GERIC, 1985).
Different seed depths and ridge-opening depths for different water contents in the soil were observed by FEY (2000).BATEMAN (1972) found higher standard deviations for deeper projected depths.
The objective of this study was to evaluate the correlation of soil clay content and the depth of corn seed placing on areas under no-till systems management.
The points for evaluation were defined considering the local mapping of soil texture.Seed depth and plant distribution was evaluated at each point.To evaluate seed depth placement, the seedlings were cut close to the soil ten days after emergence, and after carefully taking out the roots, the distance from the cut to the seed was measured, employing ten replications by each sowing row.The number of sowing rows of the planter determined the number of rows to be considered for plant distribution analyses, and the length was the minimum enough.
Four distinct planters were employed; in the area 1 a John Deere planter, and in the areas 3, 4, and 5 a Semeato machine.On each area, the seeding was performed in the same day, employing the same machine, same operator, and the same adjustment.It was therefore expected the same depth of seed placement, not taking in consideration the likely spatial variability, which may have soils texture, in particular clay content, as the main reason.
The target adjustment for depth of placement was 5, 5, 4, and 5 cm, respectively to areas 1, 2, 3, and 4. All machines were adjusted to a final population of 65.000 pl ha -1 , and were executed by the mangers of each area.
The variability on depth of placement can reduce plant population, induce errors in the plant management associated to plant growing stages, and affect the pollination.Plant distribution was represented by population pressure, or the area occupied by a single plant, in m 2 or plants per hectare (SCHIMANDEIRO et al. 2006).
The data for seed depth and population pressure was analyzed with descriptive statistics.The analysis of variance for seed depth considered each point as one treatment.The relation between soil texture and seed depth was analyzed with linear regression analyses.The statistics for this study was calculated with the software MINITAB 12.2.
The properties selected for this study belong to the project "Investigation on Precision Agriculture Practices" of Fundação ABC.The project was launched in 1998/1999 and includes properties in the region of Campos Gerais, in Paraná state.

RESULTS AND DISCUSSION
Table 1 displays the descriptive statistics of clay content in the soil on the study plots.
In soil texture studies in Campos Gerais, SÁ (1995) found clay content varying from 208 to 592 g kg -1 , and WEIRICH NETO ( 2004) observed an interval between 320 and 560 g kg -1 , with CV of 13%.It is clear that values on table 1 are typical to the Campos Gerais region.Table 1 -Descriptive statistics of clay and sand content (depth 0-200 mm) Table 2 displays the descriptive statistics of seed depths found on field.There was a higher variability of planting depth on plots 2 and 3, as the CV was higher (see Table 2).Although a CV of 19% is statistically considered average, when referring to planting depth of corn seeds this value is considered low.Accordingly to CASÃO JUNIOR et al. (1998) and CASÃO JUNIOR et al. (2000) found seed depth CVs varying from 22.9 to 41.6%, and from 26.3 to 32.7%, respectively.Average seed depth in plot 3 was 37 mm, which lies within limits considering the soil texture variation from medium to clay (Table 2).
Table 3 displays the descriptive statistics of population pressure.The CVs of plots 1 and 4 are lower than in other plots.In plot 4, standard deviation was low, and there was an average of 73,850 plants per hectare.In this case, such numbers are a result of crop management, as the farmer aimed for a higher number of plants.
In plot 2, the CV was 55%, characterizing a very high variability in plant distribution (Table 3).
An assessment of plant population pressure in Campos Gerais shows CVs higher than 55% in 3 out of 48 no-tillage cornfields (SCHIMANDEIRO et al., 2006); such variation is extremely high when maximum yields are expected.The source of missing or double plants may be irregular planting speed, machine adjustments, or biometric variations on the seed.Except for the area 1, population pressure were above the expected 65,000 pl ha -1 .
Table 3 -Descriptive statistics of population pressure (pl ha -1 ) The correlation of soil texture and seed depth for plots 1 and 2 (Figures 1 and 2) shows a close association between the two variables.The analyses of variation of both regressions were extremely significant, with coefficients of determination (R-squared) of 0.85 and 0.83%, respectively.The relation is inversely proportional: the higher the clay content in the soil, the more superficial the seed was planted.In plots 3 and 4 the analysis of variation was not significant, and the coefficients of regression were low (Figures 3 and 4).
A tendency can be established consider ing the values on variability (Table 1) and correlation (Figures 1, 2, 3 and 4): the higher the CV, the stronger the correlation between seed depth and clay content in the soil.The minimum values of clay content on plots 1 and 2 were lower than on plots 3 and 4; a slight increase of clay content to the lower values would result in more severe variations to the soil structure, soil resistance to penetration, water content.
On plots 3 and 4, the average clay content values were higher; in this case, a variation of the clay content would not have such strong influence to the soil structure.The adjustments of the planter would have more influence on the process; even though there is no correlation between seed depth and clay content on Plots 3 and 4 (Figures 3 and 4), there was an intense variation of seed depth values, especially on plot 3.
Soil clay content is an indicator for reviewing the adjustment of the depth of placement of corn seeds, as well as it can be an interesting variable as it concerns crop productivity and its tendency to influence on the depth of seed placement.

CONCLUSION
Two out of four study plots showed significant positive correlation between clay content in the soil and seed depth.The plots where the correlation was established had lower average values and higher variability of clay content.
On an area, where significant correlation was observed, there was a lower population pressure.
P.H.WEIRICH NETO, P.W. GARBUIO, L.M.GIMENEZ, M.J. COLET.Initial climatic stress combined with varying seed depth may delay plantlet emergence, causing qualitative and quantitative variability in plant distribution.