Soil Fertility Affects Establishment and Persistence of Red Clover in Grazed Pastures

J.R. Gerrish1 

Interseeded legumes play an integral role in pasture improvement programs in the humid-temperate region of the US. Soil fertility may play a role in success of establishing and maintaining legumes in grass dominant swards. Impact of soil pH, P, and K on red clover establishment and persistence in grazed pastures was evaluated. Spring red clover (Trifolium pretense L.) plant population and estimated canopy cover were not significantly affected by soil fertility parameters. Autumn measurements of red clover plant population were highly correlated with both soil P and K levels. Mature red clover plant numbers were most affected by soil P in both the 0-3 in. and 3-6 in. layer. Both grass cover and red clover population and canopy cover increased with increasing levels of P. Soil pH had minimal effect on any sward parameters, but the range of observed pH levels was fairly narrow with minimum level of 5.8.

Introduction: Success of overseeding legumes in pastures has been very variable. Soil fertility may be one factor affecting establishment and subsequent persistence of overseeded red clover. Numerous small plot studies have addressed questions of legume response to soil pH, phosphorus, and potassium levels, but less information is available regarding plant response in grazed situations. This study evaluated red clover population in the pasture environment relative to estimates of several soil fertility parameters.

Material and Methods: The study was conducted at the University of Missouri-Forage Systems Research Center located in north-central Missouri to evaluate the effect of soil pH, Bray 1 P and exchangeable K on the establishment and persistence of oversown red clover. Eight pastures consisting of either smooth bromegrass (Bromus inermis Leyss.) or orchardgrass (Dactylis glomerata L.) sods were subdivided into three paddocks for rotational grazing. One 5.33-acre paddock from each base pasture was used in this study, providing four replicates of each base grass sward.

Within each sample paddock, five transects were established originating at the water source. Permanent sampling sites, approximately 2 yd square, were established along each transect at 100-ft intervals. Due to variance in paddock shape, number of sample sites in each paddock ranged from 26 to 33. At each sample site, soil samples were collected from a one square-yard quadrant each September for four consecutive years. A different quadrant of the 2 yd square sample site was used each year to avoid any effect the sample probe holes may have had on water or nutrient movement to deeper soil strata. Samples were taken to 6-in. depth and were divided into the 0-3 in. layer and the 3-6 in. layer. At the same time soil samples were collected, red clover stand was assessed by two methods. Visual estimates of ground cover percentage of red clover, base grass, and bareground were made when the clover had regrown to approximately 4 in. during the rest period. Individual mature plants and red clover seedlings were also counted within a 10.76 ft2 quadrat. The same two methods of red clover stand evaluation were used in mid-April at all sample sites.

Differences in red clover population among individual paddocks and base grass swards were analyzed using analysis of variance. Regression analysis was used to determine relationships between soil fertility measurements and red clover population. Both linear and quadratic functions were evaluated. Where non-linear response functions provided significantly better fit of the data, second order equations are presented.

Results and Discussion: While individual paddocks differed in mean soil fertility levels and red clover population, smooth bromegrass and orchardgrass paddocks did not differ in any of the measured parameters. Data were pooled across all eight paddocks for regression analysis of soil and sward characteristics. Significant responses of sward parameters to soil variables are indicated in Table 1.

Spring sward measurements were not highly correlated with any soil parameters. Spring grass cover increased slightly in response to higher exchangeable K levels in the 3-6 in. layer but neither red clover plant population or estimated ground cover was significantly affected by any soil variable. Spring clover populations appear to be more affected by severity of winter weather and grazing pressure during the previous fall and winter rather than by soil fertility.

Red clover seedling plants, mature plants and estimated ground cover were all significantly affected by soil variables at the September observation date. Clover seedling number increased significantly as pH in the 0-3 in. layer increased above 6.0 indicating that the soil environment for seedling establishment may be improved by surface lime application. Observed bare ground percentage decreased linearly with increasing pH in the 0-3 in. layer. The range of observed soil pH in this study was only from approximately 5.6 to 6.8. This range in pH may not have been great enough for a measurable increase in red clover population due to increasing soil pH.

Soil P had the greatest impact of the measured soil variables on most sward parameters (Fig. 1 and 2). Both grass cover and mature red clover plants measured in September increased linearly as P in both the 0-3 and 3-6 in. layers increased. Estimated red clover ground cover increased in response to higher P levels in the 3-6 in. soil layer. All sward parameters increased at a more rapid rate as 3-6 in. P increased compared to 0-3 in. P. This response may be due to more extensive rooting at deeper depths as 3-6 in. P increased. Two of the four years of this study had below normal rainfall during the July to September period. Deeper roots due to higher P levels may have offset some drought stress. Baker (1980) reported increased presence of volunteer red clover on grass plots receiving P treatment compared to untreated plots. He also indicated greater response of white clover (T. repens L.) establishment to applied P compared to lime on a limestone based soil while lime was the most beneficial factor on a sandstone based soil.

Visual estimate of red clover canopy cover was not highly correlated to soil fertility parameters. Number of mature plants in the sward appears to be a better indicator of pasture fertility status than does visual estimate of canopy cover. The sward height at which visual estimates were made may also have been too short for red clover to have expressed the greater growth potential that would have been expected on a higher fertility site.

A surprising result was a highly significant quadratic response of mature red clover plants to increasing exchangeable K levels which indicated declining plant numbers at higher levels of K. The higher levels of K are far below any potential toxicity level so the result was initially confusing. However, as this data is sample site specific, the cause of this response is easily explained. In this same grazing study, soil nutrient redistribution by grazing livestock has already been reported (Gerrish et al., 1993). Almost all of the sample sites with K soil tests in excess of 400 lb/acre exchangeable K were within 150 ft of watering sites. These sites also had the highest bare ground estimates and lowest grass canopy cover estimates, probably due to overgrazing and soil compaction in these areas (Fig. 3). Thus while red clover plant population initially increased in the general grazing areas as K levels increased, the declining plant population at higher fertility levels is in response to grazing factors, not soil fertility. The red clover population response to soil K described above is one of the reasons why it is important to study forage fertility responses in the pasture environment, not only in small plot settings.

In summary, soil P appears to be a critical factor in establishment and maintenance of red clover in grazed pastures. Red clover plant population increased linearly as soil P increased throughout the range of Bray P1 values measured in this study. Even though red clover plant population increased at higher P levels, dry matter yield has been shown to peak at much lower soil P levels.

Literature Cited: 

Baker, Barton S. 1980. Yield, legume introduction, and persistence in permanent pastures. Agron. J. 72:5:776-780.

Gerrish, J.R., J.R. Brown, and P.R. Peterson. 1993. Impact of grazing cattle on distribution of soil minerals. pp 66-72. IN: Proc. Amer. Forage Grassld. Conf., March 29-31, 1993, Des Moines, IA. Amer. Forage Grassld. Council, Georgetown TX.

Table 1. Levels of significance for regression coefficients with
soil parameters as independent variables.
                     ---------------Soil Parameters---------------
Sward Parameter      pH 0-3a  pH 3-6  P 0-3   P 3-6  K 0-3   K 3-6

  RC Cover, %          .04     n.s.   n.s.    n.s.    n.s.     n.s.
  RC mature plantsb    n.s.    n.s.   n.s.    n.s.    n.s.     n.s.
  RC seedlingsb        n.s.    n.s.   n.s.    n.s.    n.s.     n.s.
  Grass Cover, %       n.s.    n.s.   n.s.    n.s.    n.s.    .001
  Bare Ground, %       .02     n.s.   n.s.    n.s.    n.s.     n.s.


  RC Cover, %          n.s.    n.s.     .1     .01    n.s.     n.s.
  RC mature plants     n.s.    n.s.   .001    .001    .001     .001
  RC seedlings         .001    n.s    n.s.    n.s.    n.s.      .03
  Grass Cover, %       n.s.    n.s.   .001    .001    n.s.      .08
  Bare Ground, %       n.s.    n.s.   .001    .001     .05     .001
a soil depth, 0-3 in. or 3-6 in.
b plants per ft2

1Research Assistant Professor, University of Missouri- Forage Systems Research Center, Linneus, MO 64653

Figure 1. Impact of soil P level on red clover plant population in grazed pastures.

Figure 2. Impact of soil P level on percent bare ground and grass canopy cover in grazed pastures.

Figure 3. Impact of soil exchangeable K on percent bare ground and grass canopy cover in grazed past

This paper was published in the Proceedings 1996 AFGC Annual Conference Vancouver BC, June 12-16, 1996.