Agro-Ecosystem Indicators of Sustainability

as Affected by Cattle Density in Ranch Management Systems

Investigators:

K.L. Campbell, J.C. Capece, J.J. Mullahey, G.T. Bancroft, M. Fanning, D.A. Graetz, J. Holt, R. McSorley, M. Mozaffari, R.M. Muchovej, K. Portier, F.M. Roka, A.D. Steinman, and G.W. Tanner

(written and assembled by J.C. Capece and M. Mozaffari)

INTRODUCTION

University of Florida Institute of Food and Agricultural Sciences (UF/IFAS), in close cooperation with Archbold Biological Station (ABS), South Florida Water Management District (SFWMD), and Florida Cattlemen's Association (FCA), has launched a new field experiment and decision support system research project to develop sustainable environmentally sensitive cattle ranch management practices for central and south Florida. The imminent start of the field experiment presents a unique opportunity to implement a full suite of well-planned component studies designed for the holistic examination of the ranchland agro-ecosystem.

Project partners have acquired adequate resources to implement the project's core components: (1) creation of a 16-plot, 1040-acre pasture array system instrumented and staffed to evaluate the effects of cattle stocking rate on runoff water quality, and (2) development of a hydrology/water quality/land use decision support system. However, the project lacks sufficient resources to address the two remaining critical system components: (3) biological effects of cattle stocking rate on agro-ecosystem indicators including animal performance, vegetation/forage quality, soil nutrient dynamics, nematode biodiversity, and avian utilization, and (4) economic effects of cattle stocking rates on ranch sustainability as predicted by simulation models that integrate cow-calf performance and other financial considerations. This proposal focuses on the rationale and funding request for implementation of components 3 and 4, since funding for components 1 and 2 is assured through existing resources. To properly examine the interrelationships of the animal-ranch-watershed-wildlife system, components 3 and 4 must be implemented concurrently with components 1 and 2. USDA support would allow researchers to implement all research components simultaneously, thus capitalizing on the window of opportunity offered by the commencement of the stocking rate experiment.

During the past five years, a variety of agricultural, hydrological and ecological research has been conducted within the project's collaborative framework (Capece and Mozaffari, 1996). Given this sustained progress, the program is now at the stage where larger experimental efforts are feasible. The newest project is a threeyear field experiment to evaluate the effect of cattle stocking rate (acres/cow) on water quality (nutrient runoff), and the potential role of grazing wetland prairies in nutrient assimilation. The site of the experiment is the 10,300-acre MacArthur Agro-ecology Research Center (MAERC), located on Buck Island Ranch near Lake Placid, Florida (Figure 1). The central theme of the experiment is the concurrent management of ranching systems for cattle production, water quality improvement, and wildlife habitat (Figure 2).

The component investigations which comprise this proposal have been under consideration for several years, as documented in the most recent project annual report (Capece and Mozaffari, 1997). Their implementation awaits introduction of cattle into the experimental arrays and funding support. The most recent meeting/workshop convened to discuss and refine experimental plans was held on February 6, 1997 at Buck Island Ranch. The agenda is provided as Attachment 1. Detailed descriptions of past, current and planned research/education activities associated with the MAERC project are available on the project homepage at http://www.imok.ufl.edu/buck/.

RATIONALE AND SIGNIFICANCE

National Significance

Florida is a major beef-producing state. East of the Mississippi River, Florida and Kentucky have the largest number of beef cattle. Even when western states are included, Florida still ranks among the top ten beef producing states. In 1994, Florida maintained 1.13 million beef cattle (FDACS, 1994). Cowcalf operations dominate the Florida cattle production industry, making the beef production of other states linked to Florida's production.

The vast majority of Florida's cattle is located in south and central Florida, south of a line between Daytona Beach, Orlando and Tampa (Figure 1). Much of what was once native subtropical wet prairie ecosystem in this region is now managed for grazing. Land use changes within these ecosystems have resulted in dramatic changes in the wildlife habitat characteristics and the patterns of nutrient flow for upland, marsh and lake ecosystems. For example, total P concentration in Lake Okeechobee has almost doubled since 1970's and chlorophyl a level significantly increased between early 1970's and 1990 (James et al., 1995). Coincidental with this general area of south Florida is one of the nation's fastest growing urban populations and one of the nation's most sensitive ecosystems. This region of Florida is home to many endangered plants and animals making it a "national hotspot" for endangered species (Cox et al., 1994). Also, water from this cattle production region feeds into Lake Okeechobee and the Florida Everglades. Preservation and restoration of this unique ecosystem ranks at the top of our national environmental priority list.

Regional Significance

In an effort to restore the Everglades/Lake Okeechobee ecosystem, the South Florida Water Management District (SFWMD) developed a Surface Water Improvement and Management (SWIM) program for the Kissimmee River Basin and Lake Okeechobee watershed. The SWIM rule was intended to obtain phosphorus load reductions beyond what had already been realized from the Rural Clean Water Protection Program (1980 1990), the dairy buy-out program and the other related programs. Nondairy sources of P are primarily from beef cattle pasture (improved pasture and native range). Although the SWIM plan is now fully operational, additional phosphorus load reduction due to its implementation has not occurred. While present management practices have not been successful in achieving the current SWIM standards, the present regulatory limits (0.35 mg/L for improved pasture and 0.18 mg/L for unimproved pasture/range) may actually be lowered to more stringent levels to serve the additional goal of Everglades restoration. If further reductions in off-site nutrient loads are to be realized, then improved management practices must be developed for cattle ranch systems.

While water quality is the primary driving force for BMP development and implementation, wildlife habitat and ranch economics are equally important to the overall goal of sustainable ecosystem restoration. Thus, the issues of nutrient dynamics, water quality, wildlife habitat, and ranch profitability must be treated as a single system (Figure 2).

The practice of grazing wetland prairies is a particularly interesting aspect of the system since those wetlands may be important areas for additional phosphorus assimilation from adjacent pastures. In order to meet phosphorus load reduction goals it will be necessary to increase the wetland area available for nutrient assimilation. Grazing of native range may be compatible with this need or it may act to negate the benefits of wetland assimilation of phosphorus. For example, if grazing pressure is low to moderate, grazing may increase nutrient assimilation by stimulating new growth (Steinman, 1996). This can be the result of a change in species composition, the formation of new tissue with higher phosphorus quotas, or over compensatory growth (Paige and Whitham, 1987). However, grazing may reduce P assimilation if the grazing pressure is too intense (Steinman et al., 1991). If grazer density is very high, then the overall biomass will be reduced and less P will be taken up. Grazing may also promote the growth of species with less P requirement. Thus, the role of native range (wetland prairies) and improved pasture (introduced grasses) is a target of attention in the drive to accomplish additional nutrient load reduction. Ranchers may be required to adjust their management (grazing density, fertilization, etc.) to help reach the P reduction goals. Additionally, grazing of wetlands is being scrutinized by the water management districts with respect to water quality issues and habitat destruction. Therefore, it becomes a regional priority to conduct scientific investigations and establish the relationship between range management practices and the ecosystem impacts.

Rationale for the Cattle Stocking Rate Experiment

Utilization of wetland prairies for cattle grazing is critical to beef cattle production in Florida. The wetland prairies are grazed during the winter months when forage productivity is low on upland improved pastures. If it were concluded that grazing these wetland prairies is incompatible with efforts to meet the water quality and wildlife habitat goals, grazing in these areas could be further regulated or completely banned. However, a total ban would constitute a significant financial hardship to Florida cattlemen. The information obtained in this study will be used to determine if grazing degrades water quality and habitat function and if there is a problem, how ranch management practices can be modified to minimize such effects while maintaining ranch profitability.

Data collected during 1996 at grazing areas within Buck Island Ranch exemplify the problem. None of the pasture sites meet the current SWIM guidelines for phosphorus concentration (Table 1). While a variety of potential BMPs can be implemented (fencing, drainage, rotational grazing, feed/water location, and fertilization), stocking rate is a relatively easy management variable for the rancher to manipulate. Thus, the strategy pursued by the MAERC research team is to focus the first field experiment on grazing density by manipulating cattle stocking rate. Upon completion of this first experiment, researchers will examine other management variables and potential BMPs.

Several waterrelated studies were conducted in Florida to quantify both runoff and nutrient loading rates from dairy and beef pastures (Capece, 1994; McCaffery, 1976; Goldstein, 1986; Capece, 1984). Table 2 summarizes nutrient loading rates measured on several pastures in the Kissimmee River Basin and the corresponding cattle stocking rates. While the phosphorus discharge appears to be generally sensitive to land use intensity, no clear relationship can be discerned due to the similarity of stocking rates for these sites. Establishing this relationship is the fundamental goal of the controlled field experiment soon to begin at Buck Island Ranch.

Rationale for the Component Studies

The suite of component studies designed to accompany the stocking rate experiment was selected on the basis of five criteria:

(1) contribution to the holistic understanding of the ranchland ecosystem,

(2) contribution to transferability of water quality results to other locations,

(3) contribution to extending the experimental results to wildlife implications,

(4) contribution to assessing the financial feasibility of new ranching practices, and

(5) compatibility with the scale and other constraints of the stocking rate experiment.

The ranchland agro-ecosystem is affected by many more external and internal factors than are shown in Figure 2. The diagram presents only those system components which are affected by cattle stocking rate and impact the stated sustainable system goals (water quality, landscape, wildlife, and economics). However, the broad scope and complexity of each component necessitates selection of specific studies which embody the essence of the component and thus yield sufficient representation of the system. Based on this general rationale and the above stated specific criteria, six component studies were selected for inclusion in the stocking rate experiment:

(1) cattle performance (animal science)

(2) forage growth and quality (range science)

(3) financial assessment and simulation model (economics)

(4) nutrient assimilation and sorption (soil science)

(5) nematode community structure (ecological indicator), and

(6) avian community response (wildlife).

These six studies complement the core water quality assessment, hydrologic modeling, and decision support system projects already in motion. One other component study (not included in this proposal) is also under development. It will document the response of additional ecological/wildlife indicators within the pasture drainage systems (algae, macroinvertebrate, amphibian, and fish communities). Together this set of core projects and complementary studies constitute a holistic analysis of the ranch agro-ecosystem as affected by cattle stocking density.

LITERATURE REVIEW

Developing a sustainable cattle ranch management system requires information on impact of cattle stocking rates on indicators of ecosystem sustainability. Livestock operations in Florida are being targeted for non-point pollution of lakes, rivers and ground-water aquifers. More specifically beef operations have been considered as one of the major sources of P that is contributing to the degradation of water quality in lakes and reservoirs (Allen et al., 1982; Boggess et al., 1995). Losses of soluble-P fertilizers applied to beef cattle pastures as well as beef cattle manure were believed to be a significant source (23%) of P loading into Lake Okeechobee (Allen et al., 1976; Allen et al., 1982). Reducing the rate of P fertilization application in recent years has reduced off site P loading from pastureland.

However, additional P load reductions are needed as previously discussed elsewhere. Effects of cattle grazing on water quality have not been fully examined to formulate a conclusive opinion. It has been stated that water quality is inversely correlated with stocking rate of grazing animals (Foy and Kirk, 1995). Therefore, reducing P loading by modification of cattle stocking rates should be evaluated.

Impact of cattle stocking rate on off-site water quality and decision support system model components of the sustainable cattle production systems are being investigated through separate grants.

A holistic approach toward system development requires concurrent study of effect of cattle stocking rates on other system components (Figure 2). If cattle management systems are to be developed to improve water quality, all aspects of cattle production must be fully examined, and then management practices should be developed that are not detrimental to the economic viability of the ranching system. The effect of cattle stocking rate on animal performance needs to be quantified and compared among cattle stocking rates to eventually develop management practices that would improve water quality without loss of animal performance.

A sustainable forage-livestock system involves integrating forage and livestock management options that meet production and economic goals (Matches, 1989). However, choices based on forages (quantity and quality), animal performance, and desired water quality from runoff are seen as wiser decisions. Forage growth and quality is another important component of developing a sustainable ranching system through modification of cattle stocking rate.

The function and value of ranchland as wildlife habitat for both game and non-game species is an important goal of a sustainable system. Wildlife improvement and biodiversity enhances the overall quality of the rangeland ecosystem. It also increases the potential ranch income through eco-tourism or hunting ground. Livestock pastures in southcentral Florida represent a mixture of open grasslands, wetlands and scattered islands of trees, resulting in a heterogeneous assemblage of habitat types and, therefore, bird species (Tanner and Christman, 1996). Grasslands typically support much lower avian species diversity than forests due to the more simplistic habitat structure (Bolen and Crawford, 1996). Avian species commonly occurring in central Florida pasture wetlands include an assortment of wading birds, shore birds, rails, and ducks (Johnson et al., 1991; Gray, 1993; Tanner and Christman, 1996). The occurrence of wetlands in grasslands has long been recognized as a factor increasing the number of bird species using the open plains (Hawkins, 1945).

Habitat use by birds is strongly associated with vegetation structure (MacArthur and MacArthur, 1961). Cattle, being primarily grazers, will influence the structure and composition of the herbaceous components of the pastures as they forage across the landscape (Westoby et al., 1989; Guthery et al., 1990; Drawe, 1991). Holechek et al. (1989) pointed out that livestock stocking rate will have the greatest impact on wildlife, beef production, and vegetation. Past research on the effects of cattle grazing on birds has been primarily directed towards game species. However, Guthery (1996) observed that "literature on the use of grazing in management of upland game bird habitat is a potpourri of unsupported conjecture, mixed research results, and conflicting recommendations". Guthery speculated that the reason for this situation is an absence of management theory that would explain ecological processes and serve as a basis for predicting outcomes of grazing programs.

It is the intent of this aspect of this multidisciplinary project to relate seasonal composition and abundance of birds to measurements of vegetation structure in the 16 experimental pastures on Buck Island Ranch. We hypothesize, following Petraitis et al. (1989), that avian species diversity will be greater in those pastures receiving the moderate stocking rate (intermediate disturbance) than the low and high stocking rates.

Long-term sustainability of an ecosystem also depends on the ability of soils in the ecosystem landscape to recycle nutrients and reduce potential off-site water quality problems. Phosphorus sorption capacity by soil plays a pivotal role in nutrient recycling in agro-ecosystems and in reducing the potential for transport of P to surface waters. This is reflected in increases in soil test P and total P in the surface and subsurface horizons of soils receiving animal manure. Vivenkandan and Fixen (1990) measured soil test P in an Egan silt loam (Udic Haplostolls) 8 years after one single application of beef cattle manure. Despite P removal by eight crops of corn, available P in the surface horizon as estimated by Bray P1 (0.03M NH4F + 0.025M HCl) was 169 and 320 mg/kg in plots that received 90 and 280 Mg/ha of manure relative to 45 mg/kg P in the check plot. Application of beef feedlot manure for 8 years at the rate of 67 Mg/ha/yr to a Pullman clay loam (Torretic Paleustolls) used for continuous grain sorghum production increased total P in the surface horizon (0 to 30 cm) from 353 mg/kg in an untreated check plot to 996 mg/kg in manured plots. Available P, as measured by Bray P1 (0.03M NH4F + 0.025M HCl), increased from 15 to 230 mg/kg (Sharpley et al., 1984).

In addition to increasing soil test P and total P, addition of animal manure has been shown by some researchers to reduce the capacity of soil to sorb additional P. Application of beef cattle manure to an Egan silty clay loam (Udic Haplustoll) in southeast South Dakota increased soil soluble P to 27.5, 50 and 71.6 mg/L for treatments that received manure at the rates of 90, 199 and 280 Mg/ha (dry weight basis) respectively, relative to 3.2 mg/L in the check plot (Vivenkandan and Fixen, 1990). However, addition of animal manure has not decreased P sorption capacity in all soil (Field et al., 1985). The reduction in P sorption capacity of some manure amended soils has been attributed to the complexation of organic acids and anions (produced as manure decomposition end products) with Fe and Al and subsequent blocking of P retention sites by these complexes (Reddy et al., 1980; Sharpley and Halverson, 1993; Sims and Wolf, 1994; Singh and Jones, 1976).

Accumulation of P in soil can be estimated by soil testing and the P sorption capacity of soils can be determined from sorption isotherms. However, sorption isotherms are timeconsuming (Sample et al., 1980). Bache and Williams (1971) proposed the use of a P sorption index, obtained from equilibration of a soil for 17 h with a single P solution at a ratio of 1.5 g P/kg soil, as a fast and simple means to estimate P sorption maxima for soils. This was recently confirmed by Mozaffari and Sims (1994) for surface and subsurface horizons of soils from a watershed dominated by animalbased agriculture.

From an environmental standpoint, continuous addition of P from animal manure at excessive rates particularly in soils low in Fe, Al and clay may lead to the saturation of soil P sorption capacity. Once this occurs, any additional input of P may result in its loss into shallow water tables or drainage water. In the Netherlands, longterm application of P in manures and fertilizers in quantities exceeding crop removal during the past two decades has resulted in saturation of more than 42% of the soils in grasslands with respect to P. As a result, soluble P concentrations as high as 1 mg/L, relative to a critical level of 0.1 mg/L for eutrophication of P limiting surface waters, have been reported in these areas (Breeuwsma and Silva, 1994). Sims (1991) pointed out that excess P accumulated in soils treated with organic waste (such as animal manure) has the potential to leach into the soil and enter the surface water through lateral transport.

These findings have significant implications for soils under cattle production in Florida. Much of Florida's cattle industry is located in south and central Florida. Soils in this area are primarily Spodosols, Entisols and Histosols (Flaig and Reddy, 1995). Surface horizon have low P sorption capacity due to low clay, iron, and aluminum. Additionally, many of these soils are artificially drained. These conditions may enhance the lateral movement of water and nutrients toward the drainage ditch and ultimately into the nearby streams and lakes (Campbell et al., 1995; Graetz and Nair, 1995).

An understanding of the effect of cattle stocking rates on P accumulation in soils, its effect on the ability of the soil to sorb additional amount of P is essential in developing sustainable cattle production systems. This information combined with data on nutrient removal by forage will provide a more complete picture of nutrient cycling within the cattle ranching ecosystem.

While the soil science component of this proposal evaluates the effect of stocking rate on P cycling and dynamics, the nematology research focuses on the effect of cattle stocking rate on the biological indicators of sustainability. One means of assessing the effect of cattle stocking rates on ecosystem sustainability is through measurement of the animal communities which make up the ecosystem resource base. Soil nematodes seem to be particularly suitable as indicators of soil ecosystem conditions (Neher and Campbell, 1994). Nematodes are abundant members of soil fauna in all ecosystems. They depend on a broad spectrum of foods and can be classified to several trophic groups: bacterial feeders, fungal feeders, plant feeders, algal feeders, omnivores, and predators (Yeates et al., 1993). This implies interactions among nematodes and other soil biota, and often results in enhanced nutrient circulation and higher productivity (Yeates and Coleman, 1982). The abundance of nematodes, particularly bacterivores and fungivores, is related to the abundance of bacterial and fungal decomposers, and the pathways and rates of decomposition and mineralization (Wasilewska, 1979; Freckman and Ettema, 1993).

Ecological indices of nematode community structure such as richness (Magurran, 1988), trophic diversity (Heip et al, 1988), evenness (Pielou, 1975), as well as the ShannonWeaver (1949) and Simpson (1949) indexes are shown to be promising for their sensitivity to management induced changes such as cattle stocking density. The nematode maturity index, based on the composition of the nematode community in terms of their life strategies (r vs Kselected) (Bongers, 1990), is also reported to be useful in monitoring soil conditions (Freckman and Ettema, 1993; Wasilewska, 1994; Neher and Campbell, 1994). There is, however, little information relating nematode community structure to sustainability and to practices which may be useful in sustainable agriculture. The proposed research will utilize indices of nematode community structure in a perennial agro-ecosystem to evaluate the impact of cattle stocking rates on this component of the natural resource base.

Since the proposal of the maturity index by Bongers (1990), there has been much interest in using indices of nematode community structure as indicators of disturbance to soil ecosystems (Ettema and Bongers, 1993; Freckman and Ettema, 1993; Neher and Campbell, 1994; Wasilewska, 1994; Yeates and Bird, 1994). While the maturity index has been useful in studies of succession (Ettema and Bongers, 1993; Wasilewska, 1994), other indices have been more responsive to disturbances from routine agricultural practices (Neher and Campbell, 1994; Yeates and Bird, 1994).

Soil nematode communities in Florida, as in other locations, contain a wide range of genera representing all trophic groups (McSorley, 1993; Powers and McSorley, 1994). Many individual genera respond to management practices (McSorley, 1993), and percent composition of the nematode community changes with moisture and rainfall patterns (Powers and McSorley, 1994). Very recent work (McSorley and Frederick, 1996) indicates that various indices of nematode community structure are descriptive of seasonal stratification of the nematode community along rows of a soybean field. Of several indices of nematode community structure, richness (Margalef, 1958), evenness (Pielou, 1975), and the ShannonWeaver (1949) and Simpson (1949) indices were most effective in distinguishing differences in community structure in rows and between plant rows. These same indices were more effective than the maturity index or fungivore: bacterivore ratios in distinguishing among degrees of disturbance to agricultural sites in South Australia (Yeates and Bird, 1994). Several of these indices should be useful in distinguishing effects from the stocking rate treatments in the proposed experiments.

Research on effect of cattle stocking rate on biological indicators of ranchland system sustainability should be complemented by developing economic simulation models that can evaluate ranch production information and assist the ranchers and planners to develop alternative management strategies. Paul B. Thompson (1995) posed the key question about sustainable agriculture systems: "How are imperatives of production and imperatives of preservation balanced?"

This important issue can be addressed by providing an improved way to test the economic impact of various recommendations before they become regulations. If water quality deficiencies are found to be directly linked to specific farming practices, decisions must be made about modifying those management practices. While improving water quality of Lake Okeechobee and the Florida Everglades is a public policy objective, all stakeholders--cattlemen, educators, politicians, and regulators--appreciate the importance of a private land-owners' economic survivability as they adopt management practices designed to maintain/improve the environment. Experience with Lake Okeechobee dairies showed that the transition toward more sustainable systems can be difficult.

Cattlemen, and indeed all the stakeholders, will need a decision aid which can help analyze the economic impact of BMP's designed to cure those problems. The primary objective of this part of the project is to develop a framework by which to estimate the impact of changes in pasture and cattle management decisions on the financial performance of the Buck Island Ranch cattle operation.

A decision aid tool that accounts for economic considerations requires two components. First, financial performance must be estimated from a given set of physical performance indicators. Second, changes in physical performance must be estimated from a change in management decision(s).

A variety of analytical methods are available to develop a decision aid tool that addresses how economic performance would be affected by a management change. A basic tool is the partial budget analysis. Partial budgets focus attention on two competing choices (Westberry, 1970). Variables not under consideration are assumed to be constant. Consequently, it is difficult to capture the effects on an entire system from a change in operation procedures.

A whole farm analysis, as the name suggests, can offer a more complete picture of how the economic system of a farm or ranch changes when operating rules change. One example is the Standardized Performance Analysis (SPA) which estimates a ranch's production and financial performance on an annual basis (McGrann, 1995). For a cow-calf operation, calving percentage and average calf weight per exposed female are two important physical performance measures. As its name suggests, SPA calculates physical performance measures in a standardized manner, allowing a rancher to compare his or her operation over time and with other operations analyzed with SPA. However, SPA lacks the operational detail needed to estimate the economic efficacy of management decisions such as fertilizer levels, burning practices, pasture configurations, and stocking rates.

Linear programming (LP) is an optimization technique that specifies the best combination of resources for a given set of defined constraints (Heady et. al; 1965). An LP model can be constructed to describe any system and can predict how a farm could reorganize its resources to comply with new regulations. However, LP presupposes behavioral assumptions, such as profit maximization, which limit the range of outcomes that could be considered. Simulation models are not bound to produce optimal solutions. A wider range of conditions can be considered to study the possible outcomes.

A simulation model describes the operation of a system in terms of the relationship among individual components of a system (Hillier and Lieberman, 1974). In economic analysis, simulation models anticipate changes to key economic variables. For example, input-output models simulate changes in regional employment and income levels from changes in economic activity from one sector. In agricultural economics, simulation models have been used to address firm level decisions. Roka and Hoag (1995) used simulation techniques to test the sensitivity of hog market weights to changes in manure value. Holt et al. (1990) addressed the question of raising or purchasing replacement heifers under stochastic production and price conditions.

While important work has been done in each of the objective areas of this multidisciplinary proposal as demonstrated in this literature review, very little information is available regarding the interrelationships and interactions of the overall agro-ecosystem and its sustainability. The goal of this research proposal is to provide information and tools for use in evaluating the performance of the overall system.

STATEMENT OF OBJECTIVES

The objectives of this multidisciplinary proposal are :

I. To investigate the effect of cattle stocking rate on the following indicators of cattle ranching system sustainability:

1. Animal performance,

2. Seasonal forage biomass production and quality in summer and winter pastures,

3. Changes in soil test P, soil P sorption capacity over time, and their implications for P transport,

4. Avian community structure and biodiversity, and

5. Nematode community structure and biodiversity.

II. To develop an economic simulation model that will:

1. Track the economic performance of cow-calf operations in South Florida for a given set of agronomic and animal performances relationships, and

2. Improve our understanding of ranching systems by identifying the agronomic and animal management decisions that have important economic consequences.

RESEARCH METHODS

The typical cattle ranch in Florida grazes its livestock on improved grass pastures during the summer months, then moves the cattle to native range (wetlands) for the winter months. To test the effects of grazing intensity on water quality and nutrient assimilation, this study will impose four cattle stocking rates on both an improved pasture site and a native range site. Data collected will be analyzed using standard statistical tools to test the hypotheses that stocking rate has no effect on runoff water quality or nutrient assimilation. The construction and other work required to prepare experimental pastures may cause temporary changes in nutrients in runoff and soil chemistry of the site. The project will be carried out in two phases in order to separate the effects due to site disturbances from those due to stocking rate treatments. The first phase of the project will be an equilibration period lasting up to one year. In phase two, the test herds will be introduced to the grazing plots at the specified treatment stocking densities. Water quality and nutrient assimilation data will be collected continuously throughout both phases.

Stocking Rate Treatments

J.J. Mullahey, M.D. Fanning, R.M. Muchovej, G.T. Bancroft, and A.D. Steinman

The experimental design for the improved pasture study is a completely randomized block employing four (4) stocking rate treatments on eight pastures as described in Table 3. Stocking rate treatments on the improved pasture plots will be 0, 1.4, 2.5, and 3.3 acres/cowcalf unit. Experimental design for the native rangeland study is also a completely randomized design employing four (4) stocking rate treatments on eight plots, with the stocking rates being different than those used on the improved pasture plots (Table 1). Native rangeland stocking rate treatments will be 0, 2.3, 4.0, and 5.3 acres/cowcalf unit. The difference in animal densities in the summer and winter array is necessitated by differences in potential biomass production between these areas. Each study animal will be assigned to a stocking rate at the beginning of the study and remain at this same stocking rate for the life of the project.

These grazing areas reflect the two principal pasturing regimes of a typical central Florida ranch. One array site is located on a wetter range area containing a mixture of native grasses, along with some bahiagrass. This range area is used for winter and spring (dry season) grazing by cows immediately after calving and during breeding. The other array site is on welldrained and improved pasture with bahiagrass, which is used for summertime (wet season) grazing of cowcalf pairs. The two arrays will be similar in design and instrumentation. The winter range array consists of a 700acre area. Within this array eight 80acre range plots are delineated. The winter range plots are 30 acres larger than the summer pasture plots because, in general, cattle are kept on winter range in lower densities than on summer pastures. The 80acre plot size allows the number of cows within a grazing herd to be kept at a level that provides greater statistical significance when evaluating animal characteristics. The 500acre summer array consists of eight 50acre plots.

Surface Water Measurements and Water Quality Investigations

K.L. Campbell, J.C. Capece, D.A. Graetz, and A.D. Steinman

Surface drainage water leaving each pasture plot will be directed to a trapezoidal flume. These flumes are hydrologically unobtrusive because they do not significantly alter water table levels or surface runoff. Peak capacity of both the winter and summer array flumes will be 7 cfs. This design specification was established based on prior regional research conducted by University of Florida. In addition, automated meteorological stations will be installed at each pasture array. Each pasture plot will also be equipped with water table wells to assess water table status.

Each flume will be equipped with an automatic water sampler. Programmable data loggers will trigger the samplers based upon flow volume and hydrograph geometry. Flow data from the flumes will be combined with nutrient concentration data to determine loading rates for both phosphorus and nitrogen. Water budgets for each array element will be determined from flow, meteorological, and water table data. This will permit assessment of the water budget and its influence on nutrient runoff loads.

This component of the stocking rate experiment is funded by an DEP/EPA 319 grant awarded to South Florida Water Management District (Title: Optimization of Best Management Practices for Beef Cattle Ranching in the Lake Okeechobee Basin Phase 1). Copy of that proposal is available by request. Currently contract arrangement are being made and Quality Assurance Project Plans (QAPP) are being developed to meet the Florida Department of Environmental Protection's Quality Assurance/Quality Control (QA/QC) requirements.

Effect of Cattle Stocking Rate on Cow-Calf Production

M.D. Fanning

The breeding females to be utilized for the experiment will be randomly chosen from one of the three breeding herds (570 head) on the Buck Island ranch. The animals are Brahman crossbred cows ranging in age from 4 to 9 years old. Animal selection will be based upon age [ability to fulfill experiment duration (3 yr)], pregnancy status at time of starting the experiment, health, conformation, and disposition. Animals selected will be identified with a number tagging system. One hundred forty breeding females will be selected and stratified by age, stage of pregnancy, and frame size then randomly assigned to a stocking rate (Table 3). Remaining breeding females will be used as replacement females for the experiment. Females on the experiment will be replaced with the same type, age and production status of cows from the replacement herd whenever stocking rate changes due to loss of an animal. Open females will be replaced with 4 year old pregnant cows from the replacement herd once a year at weaning time.

Animals will be maintained on winter pastures from November through May and on summer pastures from June through October. Animals will continuous-graze while on pastures and nutritional supplementation will be provided throughout the year as standard management practices require. Each stocking rate herd will be maintained as a unit when moved between winter and summer pastures.

Cows will be exposed to bulls (2 bulls per group) starting midJanuary for 120 days and bulls will be rotated among groups of cows every 21 days. Cows will be palpated at the end of the breeding season for pregnancy status and to determine date of pregnancy, and again at weaning time to determine final pregnancy rate. Calves will be weighed and tagged within 24 hours of birth.

Cows will be weighed and body condition scored (BCS) and calves weighed 3 times per year: prior to the breeding season; prior to the calving season; and when cattle are moved from winter to summer pastures. At weaning, calves will be weighed and marketed. The sale price of the calves will be recorded. All cows will receive the same health care as the other herds at the Buck Island Ranch.

Cow body weight, BCS, calf birth date, calf birth weight, calf weaning weight, calf average daily gain, pounds of beef produced per acre, and interval to pregnancy will be analyzed using ANOVA techniques (Steel and Torrie, 1980) based on a randomized block design. Pregnancy rates will be analyzed using chisquare analysis and contingency tables (Grizzle et al., 1969). All calculations will be performed using GLM and Catmod procedures of Statistical Analysis Systems (SAS, 1988). A return per animal in each stocking rate will be calculated and compared among stocking rates.

Effect of Cattle Stocking Rate on Forage Growth and Quality

J.J. Mullahey, R.M. Muchovej, and G.W. Tanner

Species composition and above ground biomass production and utilization will be estimated using the quadratlist method and sequential quadrat clip method, respectively. Within each of the 16 summer and winter pastures, ten 1.25 X 1.25 meter wire exclosures will be placed in a systematicrandom fashion. A paired plot of similar vegetation (based on composition and structure) will be identified and its distance and angle from the center of the exclosure measured and recorded. At monthly and trimonthly intervals in the summer and winter pastures, the occurrence and above-ground fresh weight of each plant species will be measured from within a 0.5m2 quadrat placed in the center of each exclosure and at the designated paired plot. Summer (bahiagrass) pastures will be clipped more often than the winter pastures (semiimproved) because of the stimulating effects of normal fertilization practices and due to the timing of cattle use of these pastures during the most active growing seasons of the year.

Forage production will be estimated by summation of the incremental changes in weight during the growing season, and utilization will be estimated by comparing plant weights measured inside and outside the exclosures during the period of the year when cattle are grazing the pastures. Species composition inside and outside the exclosures will be estimated by determining the frequency of occurrence of each species among respective quadrants. After each clipping, the exclosure will be relocated to a similar nearby location and paired plot reestablished.

Dry weight conversions will be obtained by ovendrying a subsample of each species or species group at 600 C for a minimum of 48 hours. The dried samples will be analyzed for N, P, K, and Total Digestible Nutrients content.

Expected results include seasonal biomass production curves, nutrients outputs (N,P,K) from both pasture systems, utilization functions, and compositional changes in vegetation as a function of stocking rate. Data will be analyzed using regression to develop predictive equations for biomass production and nutrient outputs. Planned contrasts will be used to test differences between each rate of stocking and the check.

Statistical Analysis

K.M. Portier

Because pasture sizes are established at commercial levels, stocking rate will be replicated only twice. This limit on the number of replication is somewhat mitigated by increased measurement of cattle, soil, vegetation and runoff parameters over time. This will require more sophisticated statistical analysis including the use of repeated measures/mixed effects linear models to produce fair comparisons among treatments and to accommodate expected correlations in sequential measurements. Since the number of replications is low for standard comparisons within a year, statistical analyses have been designed to incorporate repeated observations over time into these comparisons, acknowledging in the analysis that temporal correlations in these data will exist. Both annual totals and temporal patterns will be examined and accounted for as part of this analysis. SAS (1988) will be used for data management and statistical analysis.

Effect of Cattle Stocking Rate on Soil Test P and P Sorption Capacity

D.A. Graetz, M. Mozaffari, and J. Capece

Soil samples will be collected from the summer and winter pasture arrays, prior to the initiation of the stocking rate experiment (base line samples) and at regular intervals thereafter. Soil sampling interval will be four months or six months depending on the results of analysis of the samples collected after the first sampling. Soil samples will be collected from 05 and 520 cm depth. Samples will be collected using a 2 X 2 acre grid pattern. This sampling density can be modified if the results of chemical analysis of the initial samples suggest the need for more (or less) intensive sampling. Base-line soil samples will be characterized for basic physical (texture) and chemical (pH, organic matter, CEC) properties according to standard procedures. Available P (soil test P as measured by Mehlich 1 extraction solution) will be measured for base-line samples and those collected thereafter. Analysis of variance and regression analysis will be used to investigate the effect of cattle stocking rate on the available P. Phosphorus sorption capacity of the soil samples will be determined by a single point sorption isotherm (Mozaffari and Sims, 1994). This will assist in determination of the impact of cattle stocking rates on the long term capacity of soil to assimilate P from animal manure. In addition, soils and standing water in localized areas will be sampled periodically to determine if nutrients are being accumulated in swales and ditches.

Effect of Cattle Stocking Rate on Nematode Community and Structure

R. McSorley

Nematode data will be collected from the common experimental design, so that these data can be easily integrated with all other data collected during the project. Soil samples for nematode analysis will be collected three times per year (Jan/Feb, May, Nov.) from each of the 16 pasture plots. Because nematode population densities within 50 to 80 acres may be variable (McSorley, 1987), two subplots of 2 acres each will be selected within each plot and sampled separately. These subplots will beselected systematically, but will be representative of the dominant flora within each main plot. A total of 32 nematode samples will be collected on one sampling date. Each soil sample will consist of 10 soil cores (2.5 cm diam x 20 cm deep) collected over the 2acre area. The cores comprising a sample will be mixed in a plastic bag and placed in a cooler for transport to the Nematode Ecology Laboratory in Gainesville. In the laboratory, nematodes will be extracted from 100cm3 soil subsamples by wet sieving followed by centrifugation (Jenkins, 1964). Extracted nematodes will be classified, identified to genus, and counted on an inverted microscope.

The density of each nematode genus (numbers/100 cm3), the total number of genera, and the total number of nematodes in each sample will be determined. Nematode genera will be assigned to one of five trophic groups (bacterivores, fungivores, plant parasites, predators, omnivores), based on the classification system of Yeates et al. (1993). The total number and percentage of nematodes in each trophic group will then be determined for each sample. Data will be obtained on many different (30+) nematode genera, as well as on nematode communities. Trends in nematode population levels with stocking rate may be observed with several of these groups.

Several indices of nematodes community structure or composition will be calculated from data on density of the nematode genera. These include diversity indices (Freckman and Ettema, 1993; Simpson, 1949), the maturity index (Bongers, 1990), the ratio of bacterivores and fungivores to plant parasites (Wasilewska, 1994), and the ratio of fungivores to bacterivores as used by several authors (Freckman and Ettema, 1993; Yeates and Bird, 1994). The diversity indices will be calculated across trophic groups as well as across genera to obtain measures of the relative diversity or dominance among trophic groups and among taxa (Neher and Campbell, 1994).

Avian Community Response to Cattle Stocking Rate

G.W. Tanner and G.T. Bancroft

The avian community in the 16 experimental pastures will be sampled using the same methodology developed for this site during the base-line description period (Tanner and Christman, 1996). Bird surveys will be conducted within a 800m x 200m strip transect located in the center of each pasture in July (summer), October (fall), January (winter) and March (spring) of each year. Surveys will begin within the first 30 minutes of daylight and will be discontinued by 3 hours past daylight. The investigator will walk at a slow speed with frequent stops, spending about 3040 minutes per transect, recording all birds seen or heard within the strip as well as the birds' position relative to the center line of the transect. Each pasture will be surveyed four mornings in each of the four seasons. For those species having adequate abundance and distribution, Emlen's (1977) plot mapping method will be used to estimate actual density of breeding territories within the pastures.

Areal cover and distribution of wetlands within each pasture will be mapped from existing vegetation cover maps of the study site. Proportion of these habitat types will be used as a correlate with avian species composition and abundance. Habitat structure will measure two parameters: standing crop biomass and foliage height. Standing crop biomass data will be derived from the clipping study described elsewhere in this proposal. Foliage height will be estimated using a visibility board (Nudds 1977) during each seasonal bird survey. Readings will be made at 30 m distances in four random directions from each of 9 stations positioned at 100m intervals along the bird transects. Habitat parameters will be correlated with avian census data for each season.

Economic Simulation Model for a South Florida Cow-Calf Operation

J. Holt and F. M. Roka

Cattlemen, and indeed all the stakeholders, will need a decision aid which can help analyze the economic impact of BMPs designed to cure environmental problems. A cow-calf operation is a system linking forage and herd management decisions. The primary objective of this part of the project is to develop a framework by which to estimate the impact of changes in pasture and cattle management decisions on the financial performance of a south Florida cow-calf operation.

A generic simulation model will be developed to track the economic performance of cow-calf operations in South Florida for a given set of agronomic and animal performance relationships. Such a model will be an ex ante tool to predict outcomes from prescribed management changes. Further, it will improve our understanding of ranching systems by identifying the agronomic and animal management decisions that have important economic consequences.

A decision aid tool that accounts for economic considerations requires two components. First, financial performance must be estimated from a given set of physical performance indicators. Second, changes in physical performance must be estimated from a change in management decision(s). Building this tool will require three sources of data: 1) data ranchers take as given, such as market prices; 2) biological relationships which, for example, link the effect of pasture fertilization to cow herd performance; and 3) rancher management strategies, including grazing practices and animal stocking rates.

Validation of the simulation model will start with financial data from Buck Island Ranch. The Ranch participates in the Standardize Performance Analysis (SPA) program. SPA results provide accrual based economic performance measures that are comparable with outcomes generated by an economic simulation model.

Physical parameters of the model will be refined with evidence developed from the stocking rate experiments at Buck Island Ranch. We will further broaden our base of empirical knowledge by using the model to track economic performance on other south Florida ranches, correlating resource constraints and different sets of management practices.

DISSEMINATION OF RESULTS

Results of this research will be prepared for publication in appropriate refereed journals. In addition to publication of component study results in disciplinary journals, papers will be prepared addressing the agro-ecosystem interactions and interrelationships observed and measured by this multi-disciplinary research program. In addition, progress and findings reports will be presented at state and national commodity group meetings, scientific association conferences, and university seminars.

Information will also be provided to the grower community and the public. The project team currently publishes bi-monthly articles in the Florida Cattleman magazine. The Buck Island Ranch site has become a regular component of state and regional training programs related to beef cattle, forage production, wildlife management, and water resources. Meeting and workshop tour groups visit the ranch on average once a quarter.