Wendy Powers2 and
Roselina Angel3
1Department of Animal Sciences, Purdue University, U.S.A.
2Department of Animal Science, Iowa State University, U.S.A.
3Department of Animal Sciences, University of Maryland, U.S.A.
Introduction
Public concerns with air quality from livestock and poultry operations have dramatically increased in recent years. Previous “nuisance” concerns with pollutants such as particulate matter, volatile organic compounds, ammonia (NH3), methane, hydrogen sulfide, and odors have not necessarily subsided, but have been refocused on human health implications. Outside of private lawsuits, state and federal regulatory agencies are seriously focusing on further determination of emissions to see if they fit current regulations –or- if new regulations should be drafted.
Part of the issue at hand, is the paucity of data on emissions from different livestock and poultry operations. Inherently, an emission is the product of the concentration of the pollutant in question multiplied by a flow rate. Each of those factors has a unique set of measurement challenges associated with precision, especially with naturally ventilated buildings, outdoor manure storage structures, or with animals that are not reared in confinement. In the absence of data, the regulatory community often will utilize the only information available – which can often produce skewed results.
For example, regulatory agencies in California used an emission factor for VOC from dairy cows in order to calculate poultry emissions assuming that it should be same per unit of body weight (Mitloehner, 2005). Unfortunately, this is not the first time that policy preceded science, as we are entering an era where funding and timing of that funding relative to policy development does not go hand-in-hand. A summary of literature that could be used to derive emission thresholds for different livestock and poultry operations under federal Emergency Planning and Community Right-to Know Act (EPCRA) and Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) can be found at http://ag.ansc.purdue.edu/anscext/AirEmissions/ID-325W.pdf.
By the same token, development of regulations is quite complex. Simply put, any environmental regulation has to ultimately achieve a particular performance or outcome standard. Achieving that standard, however, becomes relatively difficult when having to address:
a) who is regulated,
b) what threshold of operation is selected (is this arbitrary or to encompass a certain proportion of regulated industry(ies),
c) does picking a threshold shift land base or operational ownership without achieving the performance standard,
d) economic burdens of compliance,
e) current technology for achieving the performance standard,
f) process-based regulation that shifts the site of environmental impact.
Moreover, regulations that are enacted and enforced separately can have non-synergistic impacts on environmental performance standards. Case in point, federal application of separate water and air rules to livestock and poultry operations. According to Aillery et al. (2005), separate air and water policy largely will encompass different proportions of the industry and possibly cause geographical shifts in livestock and poultry operations. For example, if a 10% reduction in atmospheric NH3 regulation were put in place with the current CAFO rule, the industry would bear additional storage, handling, treatment, hauling, and application costs of $208 million (CAFO alone = $534.5M; 10% atmospheric ammonia reduction alone = $42.2M; combined = $742.4M). Environmental benefits would be estimated at a reduction in ammonia emissions by 10% but at the cost of an additional 2 to 3% of field N runoff. Therefore, Aillery et al. (2005) surmised uncoordinated policies could ultimately increase discharge of N into surface and ground water.
How much do turkeys emit?
Very limited research has been conducted to determine air emissions from turkey barns. One of the only studies to date in the published literature was conducted by Slobodzian-Ksenicz and Kuczynski (2002) in Poland with turkey hens. Notably, the numbers of birds to achieve the CERCLA/EPRA reporting requirements (100 lb of NH3 / day) are well below that of typical operations in the US (Table 1). Some of the limitations, however, in interpreting this data are the lack of performance (body weight) information and diet nutrient content, as well as only studying one flock of hens on litter without any litter build-up.
Another two-week emission study was conducted by Schmidt et al. (2002) from turkey barns in Minnesota. The study was conducted for two-10 day periods during January/February and June.
Results of this study’s emissions are reported in Table 2. Extrapolation of this data for regulatory consideration would suggest that the CERCLA/EPRA reporting threshold of 100 lb of NH3 per day would be obtained for turkey facilities which house 69,000 birds. However, limitations of this data set include the short duration of sampling and no performance criterion as was also the case in the Slobodzian-Ksenicz and Kuczynski (2002) study.
Given the limitations of the information from the previous studies, a study was conducted by Purdue University, Iowa State University, and the University of Maryland to provide better estimates of nutrient excretions from turkeys that reflected the influence that diet has on nutrient excretions (Adedokun et al., 2006).
Two diets were fed to tom turkeys at three ages and a mass balance approach, based on nutrient composition and weight of consumed feed, carcass and litter, was used to develop new estimates of excreta composition and mass. The two diets consisted of an industry diet and a low phosphorus diet with added phytase. The non-phytin P concentration of the diet was 0.70, 0.65, 0.60, 0.55, 0.45, or 0.40% for the industry diets and 0.52, 0.42, 0.34, 0.30, 0.24, or 0.20 %, for the low phosphorus diets from 0-3, 3-6, 6-9, 9-12, 12-15, and 15-18 wk of age, respectively. Birds from four pens per diet were weighed and killed at 12, 15, and 18 wk of age for nutrient retention (in whole carcasses) and excretion (in litter) determinations.
Diet did not affect body weight or feed intake from 12 to 18 weeks of age (18-week body weight = 35.3 lb; FCR = 2.49). Total excreta production at 18 weeks was not different between treatments and averaged 26.8 lb/bird. Litter moisture at the end of the 18-week flock was 22%. Mass of phosphorus excreted at 12, 15, and 18 weeks of age was 37, 46, and 40% more for birds fed the industry diets when compared to that of birds fed the low phosphorus diets. While there were no differences in body weight, nitrogen excretion or solids excretion, phosphorus excretion (intake – carcass) was 40% less and sulfur excretion was 11% less for birds fed the low phosphorus diets compared to birds fed the industry diets.
As both the nutrient retained in the carcass as well as that in the litter were determined, the amount of nutrient that was potentially volatilized could be determined. The calculated N volatilization at 18 weeks was similar for birds fed both diets and was estimated to be 39 % of excreted nitrogen (4.58 g N per bird per day to 18 weeks). If all of the nitrogen emitted was in the form of NH3, then 11,324 birds would reach the EPCRA/CERCLA reporting threshold of 100 lb of NH3 per day.
In the case of the above study, the emphasis was on dietary phosphorus effects on litter phosphorus content. Similar work is needed to evaluate the dietary nitrogen and sulfur effects on excretion of those compounds, and perhaps more importantly on the volatilization of nitrogen- and sulfur-containing gases. Clearly, as one puts together a nutrient management plan for a turkey operation, a mass balance approach is needed to make site-specific estimates of the nutrients generated.
Diet modification
Diet modification is an approach to the air emissions issue that has been studied and shown to have promise. The good news is we do have a relatively good literature base for reducing compounds and nutrients within manure. The bad news is that these reductions do not always correspond to reduced emission factors (which we have an insufficient literature base at present), and the extent of improvements made may not be sufficient to meet compliance needs.
Generally, there are two general types of diet modification:
• Nutrient input mass reduction
• Nutrient form modification
The first, nutrient input mass reduction, changes the concentrations of the nutrient being fed such as decreasing the protein content of the diet while supplying the amino acid needs for animal performance using purified amino acids. The second, nutrient form modification, changes the chemical form of the nutrients being excreted through diet manipulation (i.e. diet acidification, dietary inclusion of additives such as urease inhibitors, or feedstuff selection to shift the site of nitrogen (N) excretion). Those strategies that reduce nutrient input mass must, by mass balance definition, decrease nutrient mass output, yet those that only change the form may initially reduce nutrient emissions to air because they “trap” nutrient in chemical forms that are not volatilized. The important question that needs answering is - for how long are these nutrients trapped in a solid form? Thus, the extent to which any reductions observed in the animal housing area through dietary strategies that change the excretion form are preserved during manure/litter storage is currently unknown and must be defined before successful implementation can occur and before mass flow models can be established.
To reduce nutrient input mass through dietary changes two approaches can be used. The first is to reduce dietary nutrient concentrations to more closely meet the need of the animal. In the case of nitrogen (N), one way this can be done is to reduce the concentration of protein and thus N in the diet but meet amino acid requirements by supplementing the diet with purified amino acids. Most of the research conducted to date on reducing nitrogenous air emissions through dietary nutrient input mass reduction has relied on this strategy. Within this approach, it is imperative that the changing nutrient needs of the animal are considered as well as strategies that allow for more frequent feed changes. If not, we may have periods of time with sufficient over feeding of a nutrient (for e.g. with lysine as exemplified in Figure 1).
In the case of turkeys, phase feeding also can have a dramatic impact on overall N consumption. For example, Waldroup et al. (1997) reported that when turkeys were fed 105% NRC (1994) recommended amino acid formulation, they maximized performance and breast yield when fed in 3-week phases (Figure 2). When diets were fed in 4-week phases, only 100% of NRC (1994) recommended concentrations of amino acids were needed. Using the Nicholas 700 predicted intake, one is able to translate those formulations into actual N fed (Figure 3). Due to the timing of when the 3-week phases were fed, however, the turkeys fed 105% of NRC (1994) amino acid concentrations in 3-week phases consumed 8% less N over the life of the flock (Figure 3). Given current (2/6/06) pricing for corn, SBM, poultry oil, Lys, Met, and Thr the feed cost would be $0.142 cheaper per marketed tom for the life of the flock for the 3-week versus 4-week phases.
A novel strategy that has the potential to further reduce air emissions is the second approach. This approach focuses on dietary modifications that shift the chemical form of nutrients being excreted thereby reducing the amount of nutrient volatilized. Because these two approaches work through different mechanisms they could successfully be coupled resulting in much greater reductions in the mass input and thus mass output of nutrients. Although dietary studies focusing on crude protein reductions have demonstrated reductions in N excreted, little work has also looked at what will be volatilized under specific storage conditions and what form that N will be in once excreted (NH3, NOx, N2O, N2).
Variation in nutrient utilization
Present commercial livestock and poultry breeds/strains are more efficient in utilizing nutrients and the present commercial feeds are better formulated to meet the requirements of the rapidly growing animal (Havenstein et al., 1994). For example, N and phosphorus (P) excretion per kg live weight was 55 and 69% less, respectively from a 1991 commercial broiler strain versus a 1957 commercial broiler strain when fed the same diet. Considerable variation exists within the literature, however, for utilization of different nutrients. Much of the variation can be attributed to feeding of different ingredients, ages, or health status. Nutrient retention values for N, P, and dry matter (DM) as summarized from 84 peer-reviewed articles from 1985 to 2003 are presented in Table 3 (Applegate et al., 2003). Notably, substantial variation existed within and between specie. For example, for the turkey, N, P, and DM retention each had a range of 27, 22, and 15 %, respectively.
Process uncertainty and feeding safety margins
Knowledge of N reduction strategies by the industry is imperative, but often difficult to implement. For example, process variation in sampling, creating diets, ingredient N content as well as N utilization by the animal still limits reductions by the industry in order that they guarantee their animals never become deficient. To illustrate this, process uncertainty can be calculated for feed formulation for a turkey (square root of the sum of squared coefficients of variations, Funk et al., 2003) from the variation listed for these processes (Table 4). Even if the lesser of the variation is assumed, the overall process uncertainty is 17.4% (or 22.3% at the worst). Even if exact ingredient analysis is known, due to bird utilization and diet manufacturing limitations, the process uncertainty could be no better than 15.8 to 18.0%. The industry, however, has been feeding at considerably lower safety margins than at these levels of uncertainty. Processes that reduce variation in individual nutrient retention, greatly improve N digestibility, or consistent nutrients within ingredients may hold the most promise in reducing the N excretion by all livestock and poultry.
Reducing dietary protein
Minimizing N excreted is the most obvious method to curb NH3 emissions. By reducing the available substrate, less NH3 will be formed and volatilized. Unfortunately, there is a widespread belief that whenever crude protein (CP) concentrations are lowered, performance would be negatively affected. Burnham (2005) speculates this belief stems from researchers (Neto et al., 2002; Bregendhahl et al., 2002) who have lowered crude protein concentrations beyond practical formulation and then did not supplement back with sufficient amounts of limiting amino acids other than methionine and lysine. Reductions in the non-essential amino acid pool, coupled with supplying a more “ideal” amino acid profile in the diet can substantially increase the efficacy of overall N retention by the bird. On a practical basis, however, bird performance can be hindered by excessive lowering of CP in diets due to a number of factors. According to Waldroup (2000), these factors can include: reduced potassium levels, altered ionic balance, lack of nonessential amino acids, imbalances among certain amino acids (e.g. branched chain amino acids), and/or potential toxic concentrations of certain amino acids.
Broilers. Reducing CP content of broiler diets by less than two percentage units resulted in decreased litter N content but no significant differences in NH3 concentration in the house (Ferguson et al., 1998). The 13.3% decrease in N intake did correspond to 18.2% reduction in litter N content. Elwinger and Svensson (1996) fed broilers diets containing 18%, 20% or 22% CP and measured NH3 emissions from the litter bed. Total N losses in the houses averaged 18% to 20% of total N input.
Turkeys. Reducing CP content (particularly by formulating to essential amino acid needs rather than setting of a crude protein minimum) of turkey diets can have considerable economic benefits. For the strains of turkeys when the studies were conducted, several researchers have noted that when essential amino acid requirements are met, crude protein recommendations of NRC (1994) are not warranted (Sell and Jeffrey, 1994; Waibel et al., 1995; Boling and Firman, 1997; Kidd et al., 1997; Waldroup et al., 1997). Depending on phase feeding programs, these studies indicate that 100 to 107% of NRC (1994) recommendations for essential amino acids were needed to maximize growth and breast meat yield. Little if any work has been done with turkeys, however, with consideration to loss of N to the environment.
Generally, as a guide, for each 1% reduction in dietary CP estimated NH3 losses are reduced by 10% in swine and poultry (Sutton et al., 1997; Kay and Lee, 1997; Blair et al., 1995; Jacob et al., 1994; Aarnink et al., 1993). As animals are fed closer to true N requirements, further reductions in dietary CP may result in less pronounced reduction in N excretion and NH3 losses.
Formulation on a digestible amino acid basis will also a) reduce the total amount of crude protein fed, and b) limit the excessive amount of non-essential amino acids fed – particularly if higher digestible crude protein feedstuffs are available.
Formulation on a digestible amino acid basis
Digestible amino acid values are considered by many to be the best measure of the amino acid value of ingredients. Long-term, reductions in protein formulation with adoption of the digestible amino acid concept should greatly reduce feed cost and N emissions. Further benefits of formulating on a digestible amino acid basis include decreasing safety margins, increasing the accuracy of predicting performance, and increasing the uniformity of product after processing. Unfortunately, knowledge of what the causes of variation in amino acid digestibility within and between ingredients is not substantial. Additionally, inconsistent methodologies make it difficult to make the switch to using digestible amino acid values, especially for non-traditional feed ingredients. Notably, most of the grow-out poultry studies focusing on use of digestible amino acid formulations have only focused on performance and economic considerations and not necessarily on N excretion or emission reduction (Fernandez et al., 1995; Rostagno et al., 1995; Dari et al., 2005). Formulation on a digestible basis can have large economical and environmental benefits, particularly when formulating with ingredients known to have lower digestibility.
Ingredient selection
Selection of feedstuffs with relatively high digestibility can help with overall reductions in amino acid formulation. Table 5 presents data for protein sources and their respective true and apparent digestibility. Notably, sources such as feather-meal are not typically considered due to their amino acid profile, but also their digestibility. Similarly, formulation for emission reduction should also consider the protein quality as exemplified in the range of apparent digestibility where processing temperatures could cause Maillard reactions as well as other conditions that would reduce amino acid availability.
Meat and bone meals (MBM) containing large proportions of hair or collagen can greatly impact total amino acid availability. For example, Lilburn (2004) reported that the digestibility of Lys, Met, Thr, and Try were substantially lower in a MBM sample containing a high amount of hydroxy-proline (indicative of high collagen content). Additionally, apparent amino acid digestibilities for ingredients should not be assumed to be the same for turkeys as they are for broilers. For example, Lilburn et al. (2004) noted that the apparent digestibility of the Thr in corn, wheat, and SBM by 4-wk-old poults was 16.4, 20, and 16.2% lower than that for the same ingredients tested in cecectomized roosters. In those comparisons, a MBM sample tested was not different between roosters and poults. While looking at 15 feed ingredients, Firman and Remus (1993) noted differences in apparent amino acid digestibilities from a few ingredients between cecectomized turkeys (12 wk-old) and roosters. For example apparent digestibility of lysine from corn was 16% less in the turkey versus roosters; several ingredients, however, were not substantially different between turkeys and roosters.
pH manipulation of diet
Ammonia is a primary byproduct of uric acid (poultry) and urea (swine) degradation as well as microbial degradation of undigested protein (Figure 4). Although uric acid contains two additional metabolic steps for conversion, both uric acid and urea are very quickly converted to NH3. This conversion for uric acid occurs via the primary uricolytic bacteria, Bacillus pasterii (Schefferle, 1965) which has an optimal growth pH of 8.5 (Elliott and Collins, 1982). Therefore, the primary factors driving growth and NH3 conversion in poultry manure/litter would be temperature (75+ F), moisture (40 to 60 percent), and pH (<7.0; i.e. higher pH increases ammonia/ammonium ratio) (Groot Koerkamp, 1994). This conversion also is reliant upon two enzymes: uricase (uric acid conversion to allantoin) and urease (urea to NH3). Once NH3 is formed, it can either be volatilized, or remain in the non-volatile state as ammonium (NH4+). In order to stay in the non-volatile state, NH3 must be protonated, which essentially means litter pH must be 7.0 or less. Several researchers have investigated how this can be accomplished through post-excretion amendments, but the more efficient conversion would ultimately be through dietary manipulation.
Pigs. Sutton et al. (1997) observed a reduction in swine manure pH when 5% cellulose was added to the diet. Canh et al. (1997) observed reduced NH3 emissions of 26% to 53% by including Ca-salts up to dietary Ca levels of 7 g/kg to 10 g/kg resulting in reductions in urinary pH ranging from 1.6 to 1.8 units. Hendricks et al. (1997) observed a 37% reduction in NH3 emissions following feeding calcium benzoate to grow-finishing pigs. Kim et al. (2000) observed a 30% reduction in NH3 emissions associated with growing pig diets containing a combination of phosphoric acid and calcium sulfate and lesser reduction in emission (17%) when diets contained a combination of monocalcium phosphate, calcium sulfate, and calcium chloride. Relative to the control diets, no NH3 emission reductions were observed when diets contained a combination of monocalcium phosphate and calcium sulfate despite significantly reduced urinary pH in these animals. This may relate to the buffering capacity in animals fed these diets. In all of these studies, measures were made after less than 24 h post-excretion. Long-term impacts, however, remain in question.
Laying hens. Little research with laying hens has been reported with the strategy of trying to reduce litter/excreta pH with dietary manipulation. The primary strategy attempted to date, includes that of replacing a portion of limestone with calcium sulfate (gypsum; up to one third can be replaced without affecting bird performance or shell characteristics; Keshavarz, 1991). Hale (2005) noted that replacement of 45 percent of dietary limestone with calcium sulfate resulted in a 17 percent reduction in NH3 emissions over a 7-day manure incubation. Reductions up to 95 percent were obtained when 35 percent dietary limestone was replaced with calcium sulfate, in combination with a 0.3 percentage unit reduction in dietary crude protein and 1.25 % added zeolite into the diet. Unpublished data by Powers et al. (2005) suggest that this combination will reduce NH3 emissions by 30-40 percent, but at the expense of increased H2S emissions.
Ammonia binding – urease inhibition
Several feed additives claim to reduce N excretion and NH3 emission potential by binding NH3 or inhibiting urease. Yucca plant extracts have been evaluated by Powers et al. (1999) who reported that NH3 concentrations in stored dairy manure were reduced. Amon et al. (1995) fed a De-Odorase, a yucca extract, to fattening pigs and observed reduced NH3 concentrations in the feeding rooms over a 7-wk period. Ammonia concentration was reduced, on average, 26% in rooms where the extract was fed. Similarly, NH3 emission was reduced 26% in the study. Dietary inclusion of clinoptilolite and other clay minerals to reduce NH3 emissions has resulted in variable findings. Kithome et al. (1999) observed that application of a layer of 38% zeolite placed on the surface of the composting poultry manure reduced NH3 losses by 44%. Amon et al. (1995) observed greater NH3 concentration and emission when clinoptilolite was used in broiler houses. However, there is no documented effectiveness following long-term manure storage using any of the above additives.
Sulfur emissions reduction
Poultry. Methionine is the reactionary substrate from microbial decomposition for production of odorous compounds such as methyl mercaptan, hydrogen sulfide, dimethyl sulfide, dimethyl disulfide, dimethyl trisulfide, and carbonyl sulfide (Kadota and Ishida, 1972). Powers et al. (unpublished data, 2005) demonstrated the magnitude of impact that a minimal change in diet formulation can have on H2S emissions from 21-wk old laying hens. The H2S emissions were 42% less from birds fed 0.1% less DL methionine versus a control diet containing 0.2% supplemental DL methionine (total dietary sulfur was reduced by 0.01 percentage units). Chavez et al. (2004) noted that in fresh broiler manure, the production of H2S, and other sulfide gases were fairly similar between birds fed either liquid methionine hydroxy-analog and DL methionine, but were much less than birds fed dry methionine hydroxy-analog and all three dietary treatments were less than those fed sodium methioninate aqueous solution. The odour detection threshold (as determined by a trained odour panel), however, was not detectable between treatment groups.
Pigs. Beyond minimizing sulfur amino acid concentrations in the diet, another considerable source of sulfur in the diet is through sulfated mineral sources. For example, when Kendall et al. (2000) replaced mineral sulfate sources (Zn, Fe, Mn, and Cu) in diets for grow-finish pigs with carbonate, oxide, and chloride sources, H2S concentrations in room and exhaust air from confinement buildings were numerically reduced by 39 and 30 percent, respectively.
Conclusions
Taken alone each of the following strategies for N emission reductions: a) dietary crude protein reductions, b) manure acidulation, and c) NH3 binding have demonstrated measurable reductions in NH3 emissions. Arguably, these reductions (when considered alone) may not be large enough, nor have the duration needed for practical implementation by the swine and poultry industries. Therefore, studies encompassing combinations of these strategies are imperative such that the industry will have documented magnitudes of reduction when EPA rolls out emission regulations in 2009.
As public scrutiny of animal agriculture increases, having holistic (integrated) lists of pollutants emitted from different livestock and poultry operations will make the regulatory decision-making more insightful. Getting this accomplished, however, will be difficult without sufficient shifts in research funding focus.
To illustrate the relative uncertainty that we have with certain classes of compounds, it was only recently that research began focusing on aerial emission of antibiotics and their residues. For example, Zhan et al. (2001) provides us with one of the few reports on tylosin concentrations being exhausted from swine finishing barns.
Assuming their estimate of 8.1+ 5.3 ng/L of exhausted air is within range, Powers et al. (2005) estimates that with a feed dosage of 20 g of tylosin per ton of feed, up to 30 percent of the administered dosage could be exhausted from the building in the active form.
Transmission of antibiotic residues, along with pathogens, disease, endotoxins, and greenhouse gases are just a few of the other issues that have been raised by the public.
References are available on request.
From Proceedings of the “Midwest Poultry Federation Convention”, St. Paul, Minnesota, U.S.A.
