Final Report for SW02-004
Nine commercial Idaho dairies (12,700 cows) were involved in all phases of this project. Nutrient flow data at farm level were collected and feeds, manure, milk, and soil samples were analyzed for various nutrients. The efficiency of whole-farm nutrient use varied greatly among participating dairies. Exporting nutrients as manure or crops, in addition to milk, was the key for achieving whole-farm balance. Levels of dietary phosphorus (P) exceeded current recommendations for high-producing dairy cows. Following our intervention, seven of the participating dairies adopted reduced P feeding, which resulted in reduced fecal P concentrations and improved the estimated farm P balance.
1) Collect data on the nutritional (N and P) status of representative dairies in Gooding/Jerome counties. These case studies will provide the needed database on the current nutritional practices in relation to nutrient balance in an area having a high density of dairy cows.
2) Based on current research, propose to the cooperator dairies modifications of existing nutrition practices that will result in more efficient on-farm nutrient utilization and overall reduction in N and P excretions into the environment.
3) Monitor the results of implementing the nutritional changes on the efficiency of nutrient utilization and N and P balance in the participating dairies.
4) Analyze the effect of the nutritional changes on the chemical properties of the manure and the economics of using manure on croplands.
5) Estimate the radius to economically transfer accumulated manure by size of the dairy operation.
7) Estimate whole-farm balance of K for the participating dairies.
Increased milk yield and herd size are the basis of intensive dairying in the United States. Public pressure, governmental regulations, and the motivation to be a “good neighbor” require dairy farmers to pay close attention to nutrient inputs and outputs and to consider the environmental impact of their operation in the context of the whole production system. In areas with intensive animal agriculture, where animal units are highly concentrated, the land base in usually not sufficient to maintain a sustainable production of milk (or meat), which inevitably necessitates import of nutrients from outside the production system. As a result, if nutrients are not effectively removed from the system, accumulation occurs endangering the quality of soil and ground and surface water resources. The type of livestock production in South Central Idaho has changed in the last ten years: from 106,000 mother beef cows and 94,000 milking dairy cows in 1991 to 107,500 beef cows and 310,000 dairy cows today (with Gooding, 131,500 dairy cows on 220,362 acres and Jerome, 64,000 cows on 193,921 acres being the most densely populated counties). The increase in animal population in confined feeding operations in the region has significantly increased manure accumulations, the cost associated with alternative disposal methods, and environmental pollution. This trend is one of great economic and environmental concern to the public and dairy producers in the region.
Management practices at smaller scales (individual farms, for example) can have a profound effect on nutrient flows on regional scale. Therefore, research efforts have been directed towards evaluating whole-farm nitrogen (N) and phosphorus (P) balances on dairy farms of various sizes (Bacon et al., 1990; Spears et al., 2003a,b; Cerosaletti et al., 2004). These studies suggested that large proportions of imported N and P were unaccounted in products leaving the farm. Predictably, large on-farm surpluses would result in increased soil N and P levels, but soil mineral levels were not published in these reports. Only one comprehensive study investigated whole-farm N and P balances on large western dairies (Spears et al., 2003a,b), emphasizing the need for more nutrient flow data on these distinctly different production systems. Except the report by Cerosaletti et al. (2004), no other studies have examined whole-farm balance of potassium (K) and none have been carried out on dairies in the western United States.
Nutrition can significantly reduce nutrient surpluses from dairy operations. Research in Europe (Valk et al., 2000; Valk, 2002) and the United States (Wu and Satter, 2000b; Wu et al., 2000; Knowlton and Herbein. 2002) has reevaluated the dietary P requirements of lactating dairy cows and demonstrated the potential of reducing nutrient losses from dairy operations through dietary manipulation (Satter et al., 2002; Klopfenstein et al., 2002; Cerosaletti et al., 2004). Such dietary manipulations can have a major impact on the overall balance of nutrients at individual farm level and the entire production system (Cerosaletti et al., 2004).
Participating dairies: Initially, ten dairies located in south-central Idaho were involved in the study. Eight dairies with six owners participated in all stages of the project. The dairies varied in size, milk yield per cow, and arable land, but had similar manure management systems. Three of the dairies had less than 1,000 lactating cows and, except for one dairy, had an average milk production during the study period of >9,000 kg. Seven dairies were entirely Holstein herds and one had a mixed Holstein/Jersey cows herd. The facilities were either freestall or open lot. In all dairies manure was accumulated in the dry lots and was removed twice a year. One dairy used all manure on its own fields while the others exported various amounts of manure. All dairies had lagoons of different sizes and solid separators. Lagoon water was used for irrigation within the farm. Dairies had various land bases. One dairy had only 34 acres of arable land and one was a mixed crop/animal operation, completely satisfying its forage needs (1,246 acres). This latter dairy would be referred to as DairyF in this report. All dairies produced some forages on their land (corn silage, alfalfa hay and haylage, triticale silage), but only four were exporting forages out of the farm. All dairies were purchasing concentrate feeds and mineral/vitamin supplements.
Sampling and data collection: Data on composition of diets, milk, manure, separator solids, feces, urine, and soil were collected throughout the duration of the project. Data regarding whole-farm nutrient balance were collected within a calendar year (January 1st through December 31st, Year 1 of the project). Nutrient inflows and outflows were obtained from farm records during monthly visits. During each visit, data on purchases (feed, fertilizer, animals) and sales (forages, animals) were collected from computer files or written records. Feed purchases data were corrected for feed inventories present on the farm on January 1st and on December 31st. Data on milk sales were obtained from the processing plants with permission of the farm owner. Amount of manure shipped was estimated based on the net weight of an average truckload and the number of loads that were shipped out of the farm. Forages, manure, and lagoon water produced and used on the farm were considered not leaving the system and were not included in the nutrient balance.
Samples of complete diets (all dairies fed total mixed rations, TMR), forages, concentrate and by-products, and mineral/vitamin supplements were collected on four separate visits from May through August during the first year of the study. At each visit, three random samples (approximately 200 g each) were taken from the concentrate feeds and by-products. These samples were combined per visit and stored refrigerated (4°C) until analyzed. At each visit, six random samples were taken from all forages present on the farm. Hay was sampled (100 g/sample) using a forage sampler (Nasko, Fort Atkinson, WI). Silage samples (300 g each) were taken from the silo after removing the top 30 cm of the silage. Hay samples were stored at 4°C and silage samples were kept frozen (-20°C). Hay and silage samples were composited (per visit) before analysis. Mineral/vitamin supplements were either samples and later analyzed, or (where available) composition was recorded from the bag label. All lactating cow diets fed at the dairy were sampled separately. Five random samples (500 g each) were collected at each visit, combined, and the composite sample was stored frozen (-20°C) for further analyses. Thus, four separate feed samples were analyzed for each dairy.
Five random dry manure and separator solids samples (500 g each) were collected at each visit. Dry manure samples were taken from each group of cows fed a separate diet after removing the top 15 cm of the manure pile. Samples from each visit were combined and stored frozen (-20°C) for further analyses.
Fecal samples (200 g each) were taken from 15 randomly selected lactating cows receiving the same diet. Fresh fecal samples were obtained from the rectum or from the ground. One composite fecal sample per diet fed was stored frozen (-20°C) for further analyses.
Urine samples (300 ml each) were taken by massaging the vulva from 10 randomly selected lactating cows receiving the same diet. One composite urine sample per diet fed was acidified with 2M H2SO4 (to pH < 3.0), diluted 1:10 with distilled water, and stored frozen (-20°C) for further analyses. Two fields for soil sampling from each of the participating dairies were selected for monitoring of soil mineral composition. The fields were selected based on consistency of dairy waste application over the past few years. The fields received either lagoon water via sprinkler irrigation or dry manure via spreader trucks in the fall and spring. Each field was sampled in two areas, each area representing the soil and landscape features of a large portion of the field. Samples were taken in the spring over a 0.5 ha area in each location to a depth of 60 cm at 30 cm intervals. The positions of the sampling points were recorded with a GPS (Global Positioning System) and resampled in the fall. Soils were analyzed for N, P, and K at a soil analysis laboratory (Harris Laboratory, Lincoln, NE) using accepted agronomic methods (Gavlak et al., 2003). Other analyses: Dry matter was determined by oven-drying at 65°C. Total mixed ration samples were analyzed for N, P, and K by Dairyland Laboratories, Inc. (Arcadia, WI). Forages, concentrate and by-product feeds, feces, urine, manure, and separator solids were analyzed at Oklahoma State University (Stillwater, OK; Jones and Case, 1990; Gavlak et al., 2003; Wolf et al., 2003). Calculation of whole-farm nutrient balance: The following data were used to calculate whole-farm balance of N, P, and K: (1) tons of individual feeds purchased during the calendar year, DM content of the feed, and concentration of N (crude protein ¸ 6.25), P, and K in feed DM; (2) tons of fertilizer purchased and N, P, K content; (3) number of animals purchased, live weight, and N, P, and K composition of the whole animal or weight gain; (4) estimated atmospheric N fixation by legumes grown on the farm; (5) amount of straw (for bedding) imported and N, P, and K concentration in straw; (6) amount of milk shipped from the farm and N, P, and K content of milk; (7) tons of feed sold off the farm, DM content, and N, P, and K concentrations; (8) number of animals sold, live weight, and N, P, and K composition of the whole animal or weight gain; and (9) tons of dry manure removed from the farm, DM content, and N, P, and K concentrations. In all cases, lagoon water was used for irrigation on the farm and was not considered in the nutrient balance analysis. The amount of feed purchased and sold was obtained from farm records. Dry matter content was determined by oven-drying. Forages, by-product feeds (canola meal, hominy feed, beet pulp, citrus pulp, distiller’s grains, wheat middlings, etc.), and manure were analyzed for N, P, and K content. Chemical composition of grains (corn, barley) was taken from NRC (2001). Mineral/vitamin supplements were analyzed for N, P, and K content, or composition was taken from labels, where available. Fertilizer composition was as specified by the manufacturer. Atmospheric N fixation by legumes grown on the farm was based on alfalfa hay and was assumed to be 60% of the N content of the forage, as analyzed. This figure was based on published fixation rates for alfalfa from 128 to 208 kg N/ha per year (Burns and Hardy, 1975). Thus, at the current average alfalfa hay yield of 8.75 t/ha (Idaho Agricultural Statistics Service, 2005; www.nass.usda.gov/id) and average N content of the alfalfa hays produced on the participating farms of 3.2 ± 0.12, 60% fixation rate would amount to an average of 168 kg of atmospheric N fixed/ha per year. All dairies imported some straw for bedding. Published composition of wheat straw (NRC, 2001) was used to calculate the amount of N, P, and K imported to the farm with bedding. Milk samples were analyzed for fat, protein, lactose, milk urea N, and solid non-fat residues (Washington DHIA, Burlington, WA). Nitrogen content was found as: protein ¸ 6.38. Concentrations of P (0.09%) and K (0.14%) in milk were taken from NRC (2001). Whole body N and P content of growing and adult animals (2.53 and 2.88% and 0.72%, N and P, respectively) was taken from the Maryland Nutrient Balancer, v. 1.25 (Kohn, 2004). Whole-body K content was assumed to be equal to K requirements for growth (NRC, 2001). Thus, at 1.6 g absorbed K requirement per kg average daily gain and 90% absorption efficiency, the K content of the whole animal was assumed to be 0.18% for all categories of animals. Whole-farm N, P, and K balances were estimated using Microsoft Excel 2000 (Microsoft Corp., Redmond, WA). Four of the participating eight dairies were owned by two families and for the purpose of this analysis each of these two ownerships was considered one entity. Thus, the nutrient balance data presented in this report are for six separate farms. Statistical analyses: Feed, manure, feces, urine, and milk samples from the four farm visits were analyzed separately and data were averaged per farm. The mean values were used in the statistical analysis. Descriptive statistics (chemical composition and nutrient balance data) and simple (Pearson) correlations among nutrient balance variables were carried out using PROC MEANS and PROC CORR procedures of the SAS software system (SAS, 2004). Comparison of soil N, P, and K levels between spring and fall samples and dietary and fecal P concentrations before and after recommended reduction in dietary P were done using PROC MIXED procedure of SAS with farm as a random effect.
Chemical composition of the diets fed at the participating dairies (Objective 1): In Year 1 of the project, average crude protein (CP) content of the diets from the participating dairies was 17.8%, which is within the range of CP routinely fed to high producing dairy cows, but somewhat above current recommendations (NRC, 2001). Reduction in the efficiency of utilization of dietary N for milk protein synthesis with increasing CP content of the diet has been well documented (Wu and Satter, 2000a; Broderick, 2003; Hristov et al., 2004c). Thus, despite of the increased milk yield, feeding more CP will inevitably result in greater losses of N with excreta, primarily urine. Crude protein concentration in the diets from some of the dairies was high (max of 20.2%). Another important observation was that in some diets solubility of dietary CP was reaching 50%. Soluble protein is rapidly degraded in the rumen and contributes significantly to urinary N excretion. Diets were adequate in fiber (NDF and ADF, neutral- and acid-detergent fiber, respectively) and contained starch and total non-fiber carbohydrates (NFC) within the range accepted for high-producing dairy cows. Concentration of P was on average 0.48%, but some of the diets had P concentration of above 0.60%. Phosphorus concentration in lactating dairy cow diets should not exceed 0.38% in order to minimize the environmental impact of the industry. Total digestible nutrients (TDN), net energy of lactation (NEL), and digestibility of dietary DM, determined experimentally, were within normal ranges. Marker methods are usually a compromise between accuracy and practicality and in this study some of the diets tested had unrealistically low digestibility of DM. In all participating dairies, cows were fed separate diets in the early (fresh or high diets) and late (lactating cow diets) stages of lactation. We did not observe any significant differences between the two categories; crude and soluble protein, fiber, carbohydrate fractions, and P concentrations were similar. Information on dry cow diets is not included in this report.
Fat and protein content of milk samples from dairies participating in the 1st stage of the project was within the range typically reported for high-producing dairy cows. Milk urea N is an indirect indicator of the efficiency of utilization of dietary N and has been used as a tool to monitor protein feeding of dairy cows (Carlsson and Pehrson, 1994; Schepers and Meijer, 1998; Jonker et al., 1999). Milk from some of the dairies had high MUN concentrations most likely indicating overfeeding of ruminally degradable protein. The relationships between dietary CP concentration and solubility and MUN, however, did not exist in this dataset. Concentration of MUN can vary greatly between cows and time of sampling (Hristov and Ropp, 2003) and our data suggest that, in commercial dairies, the relationship between MUN and protein feeding of the cows may not be as strong as in controlled research trials.
Data on N, P, and K content of lactating cow diets, the predominant forages, feces, urine, and manure from the dairies participating in all stages of the project are shown in Table 6 of the printed report. Forages purchased and fed to the lactating cows were typical for the northwest U.S. (Mowrey and Spain, 1999; except corn silage, which was fed in all participating dairies) and similar in composition to published values (NRC, 2001). Diets contained on average 17.6% CP, which is typical for high-producing dairy cow diets (Hristov et al., 2004a). Dietary P concentrations were higher than NRC (2001) recommendations. Other studies have also reported higher than recommended levels of P in dairy diets (Dou et al., 2003; Cerosaletti et al., 2004). The average concentration of K in the lactating cow diets from this study exceeded NRC (2001) recommendations (for a 90 DIM, 680 kg BW, 40 kg milk yield cow). The average concentration of K in the forages fed was similar to published values (NRC, 2001), but some alfalfa hay samples analyzed extremely high in K.
Increasing dietary P results in increased concentration of acid-digest total P and water-soluble inorganic P in dairy feces (Chapuis-Lardy et al., 2004). In intensive agricultural systems, P:N ratios in manure are twice as high as P:N required for plant growth, and excessive application of manure may exceed the assimilatory capacity of the soil and planted crops leading to eutrophication of aquatic systems (Satter and Wu, 1999), identified by the U.S. Geological Survey (1999) and Environmental Protection Agency (1996) to be the most ubiquitous water quality impairment in the U.S. (Sharpley et al., 2000). Phosphorus in soil is associated with solid components and is most often carried to surface water through erosional processes. However, Lentz and Westermann (2001), Leytem et al. (2003) and Turner et al. (2003) have found significant movement of insoluble P in some Idaho soils. Thus, P export to surface waters can be a significant problem in manure-amended soils. Since manure typically has low N:P ratios, application of manure based on N availability and plant requirements results in overloading of P, and the potential for impairment of surface water. As data from this project indicate, P levels in manure-amended soils from the participating dairies were exceeding the state’s recommended P threshold concentrations of 40 mg/kg for soils where groundwater is >2.5 m from the surface and 20 mg/kg when groundwater is < 2.5 m (NRCS, 1999). The excessive P concentration in the 30-cm sample is a clear indication of over application of animal manure on these fields. Fertilizer guides in the region do not recommend fertilizer application of P for any crops on soils >30 mg/kg. No significant differences in P levels were observed between the spring and fall samples (P = 0.922 and 0.853; 30- and 60-cm samples, respectively). Nitrate-N was >40 mg/kg for five of the eight dairies and two were over 80 mg/kg. These high concentrations are more than the crop needs for optimal growth and represent an environmental hazard. Although three of the dairies had optimal levels of soil N, they were excessively high in P. Potassium concentrations were above optimal, but not toxic. These soil concentrations illustrate the need to improve whole farm N and P efficiency and maintain better control over the whole-farm nutrient balance.
Whole-farm nitrogen balance (Objective 1): Table 7 of the printed report depicts N imports, exports, and whole-farm N balance for the dairies participating in all stage of the project. Consistent with previous reports (Spears et al., 2003a), N imported with feedstuffs was the major import to the farm; on average 90% of all N imports. As most dairies grew alfalfa forage, the second largest N import item was N fixation. The proportion of N imported with fertilizer was low (or zero, considering soil test values) for most of the dairies, except the dairy that grew large amount of forages (DairyF, 13% of the total N imports). This dairy imported only 43% of its total imported N as feed N and N fixation was a major N import comprising 38% of all N imports. The average N imported with purchased animals was low (1.7% of the total) and some dairies did not import any animals during the project period. All dairies imported straw for bedding and N imports with this item represented on average 1.7% of all N imports.
The average proportion of N exported as milk from the participating dairies was 53% of all N exported. The estimated milk N efficiency (milk N exports ¸ feed N imports) was 22% and was similar to the 23% reported by Spears et al. (2003a) (which included all animal product exports) and to European data (25%; Børsting et al., 2003). These figures are close to observations from a larger meta-analysis where average efficiency of transfer of feed N into milk protein N was estimated at 24.7 ± 0.14%, with minimum and maximum of 13.7 and 39.8%, respectively (Hristov et al., 2004a). The average proportion of N exported with animals (sold or culled) was 8%. The combined figure of 61% of N exported as animal products is similar to the one reported by Spears et al. (2003a). For most dairies, manure was the second largest N export, on average 27% of the total. DairyF, however, was utilizing all of the manure produced for its own crop production. Spears et al. (2003a) reported for Western dairies that N exported with manure was on average 34% of all N leaving the farm. All dairies participating in this study produced various amounts of forages and in most cases these forages did not leave the farm. The average proportion of N exported as forages was 12% (of the total N export), but varied significantly among the dairies. Thus, DairyF exported 77% of its N as forages and only 19% as milk.
The average whole-farm N balance was 317 t/year, but varied significantly among the dairies. The average efficiency (N output/N input) of use of N imported to the farms was 40% and was similar to the 36% reported by Spears et al. (2003a). The most efficient dairy in this study was DairyF with N efficiency of 64%. This dairy was satisfying all of its forage needs, was a relatively small dairy (550 cows), had relatively high milk yield per cow (11,887 kg/year), and was exporting large amount of N with farm-grown forages. Remarkably, this dairy was not exporting any manure. The dairy with the lowest N efficiency (25%) was a large 2,800-cow dairy, had slightly lower compared to DairyF per cow milk yield (10,748 kg/year), and had the lowest proportion of manure N exported from the farm of the total N exports (16% compared with 27% on average for all farms).
Nitrogen lost from manure as ammonia was not measured in this study and consequently was not included in the balance and efficiency estimations. Ammonia losses from manure are large and occur rapidly after feces and urine are mixed. Nitrogen in fecal matter is predominantly undigested feed N, microbial N, and N from endogenous origin. The main form of N in urine, however, is urea (Bristow et al., 1992) and ammonia emitted from livestock facilities is mainly a product of urinary urea breakdown (Rom and Dahl, 1997). If mixed with feces, urea is quickly converted into ammonia by the abundant urease activity present in fecal matter. Depending on factors such as pH, air velocity, temperature, and concentration, a large proportion of ammonia can be rapidly volatilized and lost to the environment (Monteny and Erisman, 1998; Ni, 1999; NRC, 2003). Moreira and Satter (2002) suggested that losses of ammonia from manure can be estimated based on the N:P ratio in fresh urine/feces and aged manure. This approach is based on the assumption that there are no volatilization losses of manure P. Thus, any changes in N:P ratios will be a quantitative indication of the amount of N lost through volatilization. We used this approach to estimate ammonia N losses from manure in this study and the effect of these losses on the overall whole-farm N balance. Average proportion of urinary to fecal N as excreted by the cow (g/d) was assumed to be 1.57 ± 0.11 based on feeding trials conducted in our laboratory, in which total fecal and urinary N excretions were measured (Hristov and Ropp, 2003; Hristov et al., 2004b,c; Foley et al., 2004; and Hristov et al., 2005). This ratio and the concentration of N in feces were used to estimate N concentration and consequently N:P ratio in freshly excreted manure. We assumed no P excretion with urine. This assumption was based on our own measurements (urinary P was below 0.01% in four samples and was not detected in the remaining urine samples) and on data from Van Horn et al. (1994), James et al. (1999), Wu et al. (2000), and Knowlton and Herbein (2002) who found very low urinary P concentrations and negligible contribution of urinary P to total P excretion in cattle. Fecal and manure N and P concentrations were as measured in this study. Thus, N:P ratio in fresh manure was estimated to be 8.09 ± 0.62 and that of dry manure samples was 2.62 ± 0.13. Based on these two values, we estimated the proportion of manure N lost as ammonia N for each dairy [found as: (1 – 2.62 ¸ 8.09) ´ 100]. The average proportion of manure N lost as ammonia was estimated at 68 ± 1.9% (min and max of 58 and 75%, respectively). Assuming 0.306 kg N excreted per cow (680 kg BW) per day (USDA, 1992; EPA, 2004), the average amount of N excreted annually from the dairies participating in this study would be 240 t (SD = 220 t; min and max of 61 and 633 t, respectively). Based on this value and the average manure ammonia N loss of 68%, the average annual ammonia N losses would be 159 t (SD = 151 t; min and max of 36 and 437 t, respectively), or 74 kg/lactating cow (SD = 7.3 kg; min and max of 65 and 84 kg, respectively). This would represent 28% of the total N imported to the farm (SD = 11%; min and max of 14 and 47%, respectively). When the estimated ammonia N lost from manure was added to the N exported from the farms, the average whole-farm balance of N was decreased to 157 t (SD = 135 t; min and max of 55 and 397 t, respectively) and the overall efficiency of use of imported N was increased to 68% (SD = 15%; min and max of 51 and 91%, respectively). This figure is significantly higher compared with the 40% efficiency when ammonia losses were not accounted in the whole-farm N balance. Apparently, N lost to the atmosphere as ammonia can hardly be considered an efficient use of feed resources and its emissions are currently being regulated in the United States (http://www.eh.doe.gov/oepa/guidance/cercla/rqs-gen.htm). The ammonia N losses estimated in this study are higher than those reported by Demmers et al. (1998) and Koerkamp et al. (1998) and published by EPA (2004). However, summarized data by Rotz (2004) suggest that volatile N losses from cattle facilities can be as high as 40 to 90% of the total N excreted. Using component prediction models and a whole-farm simulation model, Rotz and Oenema (2005) estimated total ammonia N losses from northeastern U.S. dairies at 47 to 87 kg/cow, depending on the type of housing and manure management practices. A case-study by Rumburg et al. (2004) reported annual ammonia emissions from a dairy as high as 170 kg/cow.
Whole-farm phosphorus balance (Objective 1, 2 & 3): Similar to the N balance data and reports by Spears et al. (2003b) and Cerosaletti et al. (2004), P imported with feedstuffs was the major P import to the farm; on average 95% of all P imports. The proportion of P imported with fertilizer was low (or zero) on most of the dairies, including DairyF (1.2% of the total P import). The average P imported with purchased animals was 3.2% of the total P import. Phosphorus imports with bedding straw represented on average 2.0% of all P imports.
The average proportion of P exported as milk from the participating dairies was 43% of all P exports. The estimated milk P efficiency (milk P exports ¸ feed P imports) was 27.6% and was within the range (23 to 47%) reported by Spears et al. (2003b) for western dairies and somewhat lower that the 35 to 40% reported for New York dairies by Cerosaletti et al. (2004). These figures are close to observations from a larger meta- analysis where average efficiency of transfer of feed P into milk was estimated at 30.5 ± 0.21%, with minimum and maximum of 12.4 and 50.0%, respectively (unpublished data from Hristov et al., 2004a). The average proportion of P exported with animals (sold or culled) was 8.6%. For most dairies, manure was equally important to milk as P export item; the average P exported with manure was 41% of the total P exports. As with N, DairyF was utilizing all of the manure produced on the farm for its own crop production. Spears et al. (2003b) reported that P exported with manure was on average 31% of all P leaving the farms. All dairies in the current study produced various amounts of forages and in most cases these forages did not leave the farm. The average proportion of P exported as forages was 7% (of the total P exports), which was significantly less that the N export with forages, and varied significantly among the dairies. DairyF, for example, exported 68% of its P as forages and only 25% as milk.
The average whole-farm P balance was 29 t/year and varied significantly among the dairies. The average efficiency (P output/ Pinput) of P imported to the farms was 66% and was similar to the 62% reported by Spears et al. (2003b). As with N, the most efficient dairy in this study was DairyF with P efficiency of 90%. This dairy had a P balance of only 2.7 t/year, or 10% of the total P imports. The major factor for this low balance was the large export of P with forages produced on the farm (15.9 t P/year). The dairy with the lowest P efficiency (48%) was the same dairy, which had the lowest N efficiency. This dairy was exporting only 1.9 t P/year with forages produced on the farm and, similar to N, had significantly lower than the average proportion of manure P exported from the farm of the total P exports (25% compared with 41% on average for all dairies).
One of the objectives of this project was to reduce manure N and P excretions through dietary means. Despite of our efforts, we were unable to convince the participating dairymen or their consulting nutritionists that reducing dietary N would not affect their production. Following our recommendations, however, and with the active involvement of the consulting nutritionists, seven of the eight participating dairies (five owners) reduced their dietary P to NRC (2001) recommended levels in the second year of the study: from 0.49% (Year 1) to 0.38% (Year 2), SE = 0.017. In all cases this reduction was achieved through removing the inorganic P supplements from the diets. Samples from the modified diets and random fecal samples were taken during four farm visits, three months after the dietary P reduction took place (second year of the project). Sampling protocols were as described earlier in this report. Dietary P was reduced (by 24%; P = 0.007) following removal of mineral P supplements from the diet, but this resulted in only numerical decrease (by 16%; P = 0.167) in fecal P concentrations. Despite of the lack of a significant reduction in fecal P, however, the decreased import of feed P would have a significant impact on the overall P balance on these dairies. We estimated that due to the reduced P levels of the diet, the net reduction in P imports would range from 5.7 to 61.4 t/year (average of 26.0 t/year, SD = 21.6 t/year). The estimated average reduction in P imports per lactating cow would be 11.9 kg/year (SD = 5.58 kg), ranging from 7.6 to 21.7 kg/year. This reduction would have an equivalent effect on the overall P balance and respectively, the efficiency of imported P use on the farm. In a similar attempt, Cerosaletti et al. (2004; 25% reduction in dietary P) reported an estimated 33%-reduction in fecal P concentrations and a 49%-reduction in the whole-farm P balance on a New York dairy.
Whole-farm potassium balance (Additional Objective): As with N and P, K imported with feedstuffs was the major K import to the farms; on average 92% of all K imports. Except DairyF (2.6% of the total K imports), none of the participating dairies imported any K as fertilizer to their farms. Potassium imported with purchased animals was insignificant and imports of K with bedding straw were on average 7.8% of all K imports.
The average proportion of K exported as milk from the participating dairies was 25% of all K exports. The estimated average milk K efficiency (milk K exports ¸ feed K imports) was 11%. The average proportion of K exported with animals was insignificant for all dairies. For most farms, manure was the main K export item; the average K exported with manure was 55% of the total K exports. As with N and P, DairyF was utilizing all of the manure produced on the farm for its own crop production, but was exporting a large amount of K with the forages produced on the farm: 92% of all K exports. All dairies grew alfalfa forages and, due to its high K content, exports with forages sold off the farm were relatively high for some farms, 5.6 to16.3% (except DairyF) of the total K exports.
The average whole-farm K balance was 182 t/year and as with N and P varied significantly among the dairies. DairyF, with its large export of K with forages was in a negative K balance (-44.5 t/year), i.e. this dairy was a net exporter of K during the project year. The average efficiency (K output/ K input) of K imported to the farms was 58%. When DairyF was excluded from the analysis, the average whole-farm K balance for the remaining dairies increased to 227 t/year and the average efficiency decreased to 38%. Apparently, as with N and P, DairyF had the most efficient use of imported K in this study. Again, the major factor for this low balance/high efficiency was the large export of K with forages produced on the farm (113.5 t K/year). The dairy with the lowest K efficiency (18%) exported 17.9 t K/year with forages produced on the farm and only 48.5 t K/year with manure (compared to 106 and 140 t K/year, respectively, for the other two dairies of similar size). Studies investigating whole-farm K balance are scarce. Cerosaletti et al. (2004) estimated 20 to 76 kg/cow K balance and efficiency of imported K use of 18 to 45% for two smaller New York dairy farms. The average balance of K per cow per year and efficiency of imported K use in this study (excluding DairyF) were 108 kg (SD = 50 kg) and 38% (SD = 12.5%), respectively.
Correlations (Objective 1): Tables 10, 11, and 12 of the printed report depict correlations matrixes for the three nutrients investigated. Within its limitations, this analysis is useful in identifying relations between import/export items and overall farm balance and efficiency of use of imported nutrients. Nitrogen import with feed correlated (P < 0.05) positively to N exports with milk and animals and import of N with fertilizer correlated (P < 0.05) positively to N exported from the farm with forages produced on the farm. As expected, the whole-farm N balance correlated (P < 0.05) positively to feed N imports, but also to milk N exports (P = 0.053), the latter resulting from the positive correlation (P < 0.05) between milk N exports and feed N imports. Efficiency of farm use of imported N was correlated (P < 0.05) positively only to feed N exports. Similar trends were observed for P; P exports with milk correlated positively (P < 0.05) to feed exports and P produced on the farm with forages correlated (P < 0.05) positively to fertilizer P imports and feed P exports. Manure P (and N) exports correlated (P < 0.05) positively to imported animal P (and N). Similar to N, whole farm P balance correlated (P < 0.05) positively to feed P imports and milk P exports, but the correlation to animal exports was also significant (P < 0.05). The efficiency of use of imported P was significantly correlated (P < 0.05) only to P imports with fertilizer, which resulted from the high correlation between the latter and feed P exports. Correlations among K import/export variables and whole-farm K balance and efficiency of imported K use were similar to those for N and P. However, milk K export did not correlate (P = 0.299) to overall farm K balance. The efficiency of use of imported N correlated (P < 0.05) positively to fertilizer K imports and K exports with feed and the amount of forage K produced on the farm. Similar to N and P, these latter correlations emphasize the importance of producing and selling forages off the farm, in addition to milk and manure exports, for maintaining low nutrient farm balance. This is also evident from the greater N, P, and K efficiency of DairyF compared to the other participating dairies, which did not produce and sell nearly as much forage as did DairyF. Economic analysis (Objectives 4 & 5): We estimated that as a result of the reduced P content of the lactating cow diets, the producers participating in the second phase of the project saved 2 cents/100 lbs of milk produced. At average milk produced per year of 27,518 metric tons, the dairymen participating in the project would realize cost savings of approximately $12,108/year. This reduction in feed cost would be achieved without affecting milk yield or reproductive efficiency. It is estimated that using dairy and beef cow manure on soils in South Central Idaho and a potato-grain-grain rotation has the potential to reduce fertilizer costs, including application costs, from the present $320/acre to $96/acre for one mile distance between the feedlot and the receiving field ($253/acre for 12 miles distance). Optimal quantity of manure for this rotation will stabilize at 30 tons the first year, 11 tons the second year, and 10 tons the third year. In this scenario, synthetic N use will be reduced by 500 lb/acre, and synthetic P use by 180 lb/acre. Nitrate leaches and P run-off are expected to decline by over 50%, significantly improving ground and surface water qualities. Depending on the properties of the manure, and the properties of the soil, it estimated that manure could be hauled between one to nineteen miles from its source to be spread on cropland before its hauling and spreading costs would equate the cost of using synthetic fertilizer. However, these estimates are based on organic N mineralization data generated in a controlled laboratory environment, ignoring the P content in the manure and its rate of mineralization in soil amended manure.
The results from this project showed that N, P, and K purchased with feedstuffs was the major import item and milk, manure, and forages sold off the farm were the major export items on commercial dairy farms in South Central Idaho. The whole-farm import/export analysis indicated net accumulation of N, P, and K on the farm. Exports of nutrients with milk and manure were not sufficient to achieve whole-farm balance and sustainability of the dairy operations and greater than recommended levels of dietary P and K were partially responsible for the observed low efficiency of use of imported nutrients. As a result, soil P levels were unacceptably high and above state threshold standards in all samples from manure-amended soils. In all dairies, dietary P concentration was above current NRC recommendations for high-producing dairy cows. Reduced dietary P levels adopted by some of the participating dairies resulted in significant reduction of P imports and showed potential for reducing whole-farm P balance, which will serve as an example for other dairies in this intensive livestock area. The expected long-term effect of these measures is reduction in soil P levels. The project demonstrated that in addition to milk and manure exports, export of nutrients with forages produced on the farm is a major factor in achieving whole-farm N, P, and K balance. This important information has been communicated to Idaho dairy producers, has been used in several education materials, and has already generated a significant interest among dairymen, consulting nutritionists, and extension educators.
Education and Outreach
The results from this project have been communicated to producers, consulting nutritionists, and extension educators at several regional meetings: Ag and Water Quality Conference, Boise, ID, 2004; Pacific Northwest Animal Nutrition Conference, Seattle, WA, 2004 and Boise, ID, 2005; and United Dairymen of Idaho Annual Meeting, Pocatello, ID, 2004.
Results from the project have been used to prepare the following extension and educational materials:
1. Hristov, A. N. 2005. Idaho mineral case study. Feed Management Workshops in Twin Falls, ID and Puyallup, WA, August 3-4, 2005.
2. Hristov, A. N. 2004. Nutrient Management Study. United Dairymen of Idaho Annual Meeting, Pocatello, Idaho, November 1-2, 2004.
3. Hristov, A. N. 2004. Nitrogen and Phosphorus Nutrition of Dairy Cows and Reproduction. WIN2ME Workshop, Ag and Water Quality Conference, Boise, Idaho, October 18, 2004.
4. Hristov, A. N. 2004. Nitrogen case studies in Idaho dairies. Nitrogen Management in a Whole Farm Nutrient Management Context, Advanced Feed Management Workshop. Pacific Northwest Animal Nutrition Conference, Seattle, October 5-7, 2004.
The following scientific reports and manuscripts have been published or submitted for publication:
1. Hristov, A. N., R. P. Etter, A. Melgar, J. I. Szasz, K. L. Grandeen, S. Abedi, J. K. Ropp, D. Falk, W. Hazen, and R. Ohlensehlen. 2003. Nitrogen, phosphorus, and other minerals in Idaho dairy diets. J. Dairy Sci. 86 (Suppl. 1):227.
2. Hristov, A. N. 2005. Whole-farm nutrient balances on Idaho dairies. Proceedings, Pacific Northwest Animal Nutrition Conference, Boise, ID, October 18-20, 2005, pp. 161-172.
3. Hristov, A. N., W. Hazen, and J. W. Ellsworth. 2005. Nitrogen, phosphorus, and potassium balance and potentials for reducing phosphorus imports in Idaho dairy farms. J. Dairy Sci. (submitted, JSD-25-0840).
Education and Outreach Outcomes
Areas needing additional study
We have identified a problem of a significant importance for South Central Idaho, which is common for areas with intensive livestock production, i.e. accumulation on the farm of environmentally important nutrients such as N, P, and K. This study indicated that export of N, P, and K with milk and manure is not sufficient to maintain sustainable milk production when the land base of the farm is limited. As the results from this project indicated that P levels are high in manure-amended soils from large dairies in the region, nutrients must be exported out of the production system until soil P levels are below state threshold standards. One method of achieving nutrient balance on large dairy and beef operations is to sell manure to crop producers. Therefore, generating scientific information on the economic utilization of manure, market for manure, and the environmental benefits of exporting nutrients off the livestock production system is critically important. A method to estimate the optimum quantity of manure per acre, acreage needed, hauling radius, and cost to utilize manure with different chemical properties on different soils is needed. Field data on the mineralization of organic N and organic P in manure-amended soils is needed to realistically estimate the economic potential of utilizing manure on cropland under Idaho’s climate and soils conditions. The Araji econometric model (Araji et al., 2001) needs to be modified to estimate the application of optimal quantity of manure subject to P constrain.