| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Animal Sciences, The Ohio State University, Columbus, OH 43210
Department of Agricultural and Consumer Economics, University of Illinois, Urbana, IL 61801
Corresponding author: N. R. St-Pierre; e-mail: st-pierre.8{at}osu.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: heat stress • temperature-humidity index • livestock economics • livestock production
Abbreviation key: DMILoss = the reduction in DMI from heat stress (kg per animal or per 1000 birds per day), DOLoss = the change in the average number of days open from heat stress, 
THI = the change in apparent THI from a heat abatement system, EGGLoss = the loss in egg production from heat stress (kg per hen per day), GainLoss = the loss in body weight gain (kilogram per animal or per 1000 birds per day), H = relative humidity (%), PDeath = the change in monthly death rate from heat stress, PR = monthly pregnancy rate, RCullRate = the change in monthly reproductive cull rate due to heat stress, T = temperature (°C), THI = temperature-humidity index, THILoad = integral of the daily THI sine curve above THIthreshold, THILoadm = the average monthly THILoad, THImax = daily maximum THI, THImin = daily minimum THI, THIthreshold = THI threshold above which heat stress occurs in a given animal class, ZTC = zone of thermal comfort
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Much work has been done to identify the physiological effects of heat stress and the mechanisms by which animal productivity is reduced. In dairy, heat stress consistently result in reduced DMI (West, 1994) and this effect is generally greater in pluriparous than in primiparous cows (Holter et al., 1996, 1997). The extent of production loss is often difficult to estimate because heat stress effects are typically hidden among high natural and managerial sources of variation (du Preez et al., 1990c; Linvill and Pardue, 1992), plus other confounding factors, such as stage of lactation, breed, and age (Ray et al., 1992; Ravagnolo and Misztal, 2000; Ravagnolo et al., 2000), and carryover effects (Collier et al., 1982a).
Heat stress reduces the expression of estrous behavior (Hansen et al., 2001), alters follicular development (Wise, et al., 1988; Wolfenson et al., 1995) and the growth and function of the dominant follicle (Wilson et al., 1998a, 1998b), compromises oocyte competence (Collier et al., 1982b; Wolfenson et al., 2000), and inhibits embryonic development (Drost et al., 1999). The quantification of the effect of heat stress is further complicated because it has both a concurrent and delayed effect on the reproductive system (Wolfenson et al., 1997; Rotz et al., 2000, 2001). Consequently, heat stress reduces fertility of female (Folman et al., 1983) and male cattle (Ax et al., 1987), resulting in reduced reproductive performance (Monty and Wolf, 1974; Salah and Mogawer, 1990).
The incidence of new udder infections and frequency of mastitis increases during hot summer months because the udder’s defense mechanisms become deficient (Giesecke, 1985). Cow mortality increases during periods of heat stress (Hahn, 1985), but the quantitative relationship between mortality risk and magnitude of heat stress remains to be defined. The quantification of the effects of heat stress on dairy cattle is further complicated because cattle have the ability to acclimate to changes in the environment (Wolfenson et al., 1988; du Preez et al., 1990c), genetics plays a role in tolerance to heat stress (du Preez, 2000; McDowell et al., 1996), current selection for production reduces heat tolerance in the United States (Ravagnolo and Mitsztal, 2000), and nutrition and management strategies can reduce its effect (Coppock et al., 1982; Schneider et al., 1984; Knapp and Grummer, 1991).
Most of the effects of heat stress identified in dairy cattle are also present in beef cattle, albeit to a lesser extent due to the overall lower body heat production (lower plane of production) of beef cows combined with a traditional breeding season during which the incidence of heat stress is low. In growing cattle, heat stress has decreased DMI, increased DM digestibility (Lippke, 1975), decreased rate of gain (Ray, 1989; Mitlohner et al., 2001) partially negated by compensatory gain (Mader et al., 1999), and reduced fertility of males (Meyerhoeffer et al., 1985) and females (Biggers et al., 1987). Quantification of these effects is complicated by acclimation of animals (Robinson et al., 1986) and breed differences in their susceptibility to heat stress (Hammond et al., 1998; Gaugham et al., 1999).
In sows, heat stress has consistently been associated with decreased DMI, milk yield, and increased sow lactation BW loss while reducing the weight gain of the litter preweaning (McGlone et al., 1988b; Johnston et al., 1999; Renaudeau and Noblet, 2001; Renaudeau et al., 2001). Litter size, however, is either unaffected (Johnston et al., 1999) or is increased by heat stress (McGlone et al., 1988b) due to decreased piglet mortality. Additionally, piglets from sows under heat stress exhibit strong compensatory weight gains postweaning, essentially negating most of the heat stress effect while suckling by 2 wk postweaning (Renaudeau and Noblet, 2001; Renaudeau et al., 2001). The sow reproductive system is sensitive to heat stress pre- and postmating. Heat stress affects fertility of both male and female pigs for up to 5 wk after a stressful event (Wettemann and Bazer, 1985). Embryo development is compromised with heat stress (Kojima et al., 1996), and the proportion of sows showing delayed return or failure to return to estrus after mating is increased noticeably (Hennessy and Williamson, 1984; Gross et al., 1989; Liao and Veum, 1994). Sow mortality also has been associated with heat stress (D’Allaire et al., 1996). Nutrition can mitigate some of the effects of heat stress in sows. Fiber addition to the diet increases, but fat addition decreases, the impact of heat stress on sows (Schoenherr et al., 1989). During growth, young gilts are not affected much by heat stress until breeding time, at which heat stress has the same depressive effect on reproduction as in older animals (Flowers et al., 1989). Severe heat stress can also affect the growth of market pigs, although acclimation is a factor (Collin et al., 2001). During periods of heat stress, growing pigs reduce fasting heat production by 18%, daily heat production by 22%, and thermic effect of feed by 35% (Collin et al., 2001). Social stressors (regrouping) magnify growth and intake depression resulting from heat stress (McClone et al., 1987).
Prolonged, severe heat stress affects DMI and daily gain of broiler chickens, especially after 28 d of age (Cooper and Washburn, 1998; Yalcin et al., 2001a). The ZTC in broiler chickens, especially under 4 wk of age, is substantially greater than that of most other commercial farm animals (NRC, 1981). Additionally, acclimation to high thermal conditions at an early age (4 to 7 d) noticeably reduces the effect of heat stress at a later age (Yahav and Plavnik, 1999; Altan et al., 2000; Yalcin et al., 2001a). Acclimation reduced heat production by 11.4% and evaporative heat loss by 14.8% (Wiernusz and Teeter, 1996), and lowers heat stress mortality (May et al., 1987). Thyroid size is reduced in birds grown under heat stress, especially if heat stress is cyclic (Dale and Fuller, 1980). Heat stress during rapid growth has also been associated with undesirable meat characteristics (Sandercock et al., 2001). Male broiler breeders are affected more by heat stress than females (McDaniel et al., 1995). Bird mortality increases during heat stress (Bogin et al., 1996; De Basilio et al., 2001) and is greater near marketing time and in the presence of some anticoccidial drugs (McDouglad and McQuistion, 1980; Arjona et al., 1998), as well as during transportation to central processing plants (Mitchell and Kettlewell, 1998).
Research on heat stress in laying hens is not entirely consistent regarding its effects on percent hen-day production, but results show a consistent decrease in egg weight and shell thickness (Wolfenson et al., 1979; Emery et al., 1984; Muiruri and Harrison, 1991; Wolfenson et al., 2001). Acclimation to heat stress in layers is pronounced (Sykes and Fataftak, 1985, 1986; Sykes and Salih, 1986). Dietary parameters can modulate the effect of diet stress (Bollengier-Lee et al., 1998; Bollengier-Lee et al., 1999; Sahin et al., 2002) as well as management factors (Kassim and Sykes, 1982; Sahin and Kucuk, 2001).
Literature on heat stress in turkeys relates primarily to mortality (Evans et al., 2000) and the association between heat stress and the incidence of pale, exudative meat (McKee and Sams, 1997; Owens et al., 2000).
In all, research has identified many of the mechanisms by which heat stress affects the different classes of farm livestock. Recommendations regarding housing, ventilation, and cooling systems are now issues that are probably applicable on a regional basis (Flamenbaum et al., 1985; Lin et al., 1998; Armstrong et al., 1999). Some economic analyses have been done, but they failed to recognize that capital costs of cooling systems are incurred even during periods when heat stress is absent (Igono et al., 1987). Efforts are under way to quantify livestock responses for heat stress management (Mayer et al., 1999; Nienaber et al., 1999), although these efforts are not inclusive of all farm animals of economic importance. Currently, there are no known estimates of the total economic losses to US livestock industries that are attributable to heat stress. An estimation of such losses would serve in assessing the need for public research investments in heat stress abatement and could be used as a quantitative platform to issue regional recommendations for the various classes of food producing animals. The objectives of the present study are to provide estimates of national and regional economic losses from heat stress by major US food-producing animal industries and to identify areas for which information is lacking to adequately quantify important processes.
|
RESEARCH AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
To account for the extent and cumulative severity of heat stress within days, two additional variables were calculated (Figure 1
). The temperature-humidity index was assumed to follow a perfect sine function with a period of 24 h. This assumption underestimates duration of heat stress at higher latitudes in summer time, but gains in accuracy with more complex models (e.g., Linvill and Pardue, 1992) are overall small. A THIthreshold was identified for each class of animal (Table 1) and is defined as the THI level at which heat stress begins. Using THImin, THImax, and THIthreshold, duration (D) of heat stress and time summation of THI in excess of the threshold (THILoad) were calculated. Details regarding the calculation of D and THILoad are provided in Appendix in the form of a computer code.
|
|
. Births of animals were assumed uniform throughout the year with the exception of beef cattle from which 75% of the breedings were modeled to occur during the spring season.
|
. For dairy cows, the following set of equations was used:
|
 | ([1]) |
where
| DMILoss | = | is the reduction in DMI from heat stress (kilogram per animal per day),
| THImax | = | is the maximum THI during a day,
| THIthreshold | = | is the THI threshold, above which heat stress occurs for dairy cows,
| D | = | is the proportion of a day where THI > THIthreshold (e.g., 0.33),
| MilkLoss | = | is the reduction in milk production (kilogram per cow per day),
| PR | = | is the monthly pregnancy rate (e.g. 0.15),
| THILoadm | = | is the monthly average THILoad,
| DOLoss | = | is the change in the average number of days open,
| RCullRate | = | is the change in monthly cull rate (e.g., 0.01),
| PDeath | = | is the change in monthly death rate from heat stress, and
| EXP | = | is the exponentiation function (i.e., e exponent the expression in parentheses).
|
The relationships between DOLoss, RcullRate, and PR were derived using a Markov chain Monte Carlo procedure (St-Pierre and Jones, 2001).
Dairy Replacement Model
Insufficient data were available to develop a model specific to growing dairy animals. We used the finishing beef cattle model and adjusted the parameters to reasonable targets of daily gain and DMI. Replacement animals under 1 yr of age were modeled according to the following equations:
 | ([2]) |
where
| GainLoss | = | is the loss in BW gain (kilogram per animal per day).
|
Equations for replacement heifers over one year of age were:
 | ([3]) |
THILoad and THILoadm are functions of THIthreshold, which was set at 77 for animals under a year of age and 72 for older replacement animals.
Beef Cow Model
Studies used to develop response functions in beef are reported in Table 3
. Equations used to model the response of beef cows to heat stress were:
 | ([4]) |
Although it is probable that DMI of range cattle drops when animals are heat stressed, published observations are lacking to quantify the process. Thus, we assumed this loss to be negligible.
Finishing Cattle Model
The following set of equations were developed for this class of animals:
 | ([5]) |
Sow Model
Studies used to develop equations for sows and grow-finish hogs are reported in Table 3. For sows, the following set of equations resulted:
 | ([6]) |
where
| ARate | = | is the abortion rate.
|
Although sows reduce feed intake when heat-stressed, this is done at the expense of BW loss that must be replenished later. Thus, there are no realized net savings in feed over a full reproductive cycle, which is why we set the value of DMILoss to 0. From a reproduction standpoint, we assumed that sows are not culled for reproductive failures due to heat stress. The cost of a prostaglandin injection to resume reproduction was added to each reproductive failure.
Grow-Finish Hog Model
Equations used for grow-finish hogs were:
 | ([7]) |
Poultry-Broilers Model
Studies used to develop response functions for all three poultry species are reported in Table 3. For broiler chickens, the following equations were developed.
 | ([8]) |
Poultry-Layers Model
The following equations were used to model physical losses of laying hens:
 | ([9]) |
where
| EGGLoss | = | is the loss in egg production (kilogram per hen per day).
|
Note that the equation for EGGLoss incorporates the negative effects of heat stress on both the percent hen-day production and egg size. Production losses are converted to dozen egg equivalents assuming that a standard dozen of eggs weighs 0.72 kg (i.e., 1 egg = 0.06 kg).
Poultry-Turkeys Model
Data on the effect of heat stress in growing turkeys are scarce. We used the model developed for broilers, substituting parameters in line with normal growth of turkeys at an average 4.5 kg of BW.
 | ([10]) |
Physical and Economic Inputs
Table 1 reports THIthreshold assumptions used for each of the 10 animal classes. Because current selection for production reduces heat tolerance in dairy (Ravagnolo et al., 2000), we lowered the THIthreshold of dairy cows from the traditional value of 72 established many years ago to a value of 70. Other values of THIthreshold were as reported or calculated from literature data.
Unit values for each of the five categories of losses are given for each animal class in Table 1. Values were chosen to represent average US costs over the last 5 yr. The price of some animal commodities (e.g., milk) varies appreciably over US regions and over time. The variation in output unit values was not factored in our model.
Cooling Systems
Equations presented so far are applicable to animals maintained in a system of minimal cooling. In confinement, such a system would rely on natural ventilation or mechanized ventilation where air exchange is limited to providing animals with adequate air exchange to maintain its chemical quality but without creating sufficient air movement around the animals to result in significant cooling effects. In dry lots, the equations implicitly assume that animals have access to shade because solar radiation is not factored in the response model.
Moderate heat abatement.
The first intensity of heat abatement modeled was conceptualized as a system of fans or forced ventilation and was classified as "moderate". In dairy cows, literature data (Berman et al., 1985; Flamenbaum et al., 1986; Strickland et al., 1989; Means et al., 1992; Turner et al., 1992; Lin et al., 1998) were used to derive the effectiveness of moderate heat abatement, which was expressed as the decrease in apparent THI experienced by the animals. In our model, the actual THI is replaced by the apparent THI when one of the three levels of heat abatement is used. Figure 2a depicts the effect of moderate heat abatement intensity on apparent THI as a function of temperature and relative humidity according to the following equation:
|
 | ([11]) |
where
| THI | = | is the change in apparent THI
| T | = | is ambient temperature (°C), and
| H | = | is ambient relative humidity (%).
|
This equation was used across all animal types to estimate the physical effectiveness of a moderate heat abatement system. From a cost standpoint, one cooling unit was used per 50 m2 of housing or per 3800 kg of BW. The purchase cost per cooling unit was set at $250, which was annualized at a rate of 15% to cover maintenance, depreciation, and interest costs. The sum of all fixed costs associated with the additional investments was labeled capital cost. Operating costs assumed an electrical consumption of 0.65 kW/h of operation, and $0.09/kW•h of electricity.
High heat abatement.
Conceptually, this intensity of heat abatement has the effectiveness of a combination of fans and sprinklers in dairy. For dairy cows, published data (Flamenbaum et al., 1986; Igono et al., 1987; Strickland et al., 1989; Means et al., 1992; Turner et al., 1992; Lin et al., 1998) were used to quantify the decline in apparent THI using the following equation:
 | ([12]) |
Figure 2b shows the drop in apparent THI for a high heat abatement system. Capital costs for this system were calculated as those of the moderate system plus $60 of additional investments per 50 m2 of housing or per 3800 kg of BW; these costs were annualized at a rate of 25% to cover depreciation, interest, and the additional maintenance. Operating costs were the same as those for the moderate system augmented by $0.01/h of operation.
Intense heat abatement.
Conceptually, this intensity of heat abatement has the cooling properties of a high-pressure evaporative cooling system in dairy. Field data from a commercial manufacturer (Korral Kool, Inc., Mesa, AZ) were used to quantify the cooling effect of an intense heat abatement system. Evaporative cooling is the only commercially available system that actually decreases the actual THI as opposed to changing the apparent THI. The drop in apparent THI at various combinations of T and H is shown in Figure 2c based on the following equation:
 | ([13]) |
Capital costs were calculated based on additional investments of $6000 per 120 m2 or per 8865 kg of BW, annualized at a rate of 15%. Operating costs were calculated using a rate of $0.23/h of operation per unit.
Simulation
Monte Carlo techniques were used to simulate the variation of weather data across time. A variance-covariance matrix and a vector of means of minimum and maximum T and H were calculated for each month within each state. These were used to generate 30 d of weather data per month, assuming a multivariate normal distribution of all four variables using the algorithm of Fishman (1978). This process was iterated 1000 times for each month within each state and for each of the 10 animal classes and four heat abatement intensities.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
. The aggregation of weather data to the state level distorts the heat stress picture for a few states. In Texas, for example, the weather in July is typically hot and dry in the northwest panhandle but hot and humid in the area along the Gulf of Mexico. Although this aggregation may impact our assessment of the optimal cooling system for a given animal class in a few states, it probably has minor impact on the overall economic impact on a national basis.
|
demonstrates the need to account for T, H, and THI patterns beyond their simple daily averages. For example, Ohio and Montana have the same average maximum T, but minimum T is 5.5°C less in MT. The average maximum THI in Idaho and Rhode Island are identical, but the average minimum THI is 10 units less in Idaho. High humidity compounds the effects of high temperatures. For example, although Utah and South Carolina have nearly the same average maximum temperature (31.5°C), the THImax and the THImin are 7.1 and 13.8 units lower, respectively, in Utah. The difference between the average minimum and maximum THI varies considerably across states. In general, the THIspread is small in southeastern states and large in western states. This has a substantial impact on the magnitude and duration of heat stress on a given day. During an average July day in Florida, for example, a dairy cow would be constantly under heat stress conditions, whereas a cow in Arizona (the state with the highest mean maximum temperature in July) would be exposed to THI conditions under her THIthreshold for approximately 8 h/d.
Impact of Heat Stress on Productivity Without Heat Abatement Systems
Dairy cows.
The impact of heat stress on the productivity of dairy cows in the absence of heat abatement is presented in Table 5. Reduction in milk production ranges between 68 and 2072 kg/cow per year in Wyoming and Louisiana, respectively. The effect on reproduction varies considerably across states, with a low of 4.3 and 2.7 in Wyoming and a high of 57.7 and 88.0 in Louisiana for DOLoss (days) and RCullRate (animals/1000 animals), respectively. Annual heat stress is summarized in terms of duration (hours per year) and extent (as a sum of THILoad per year). The THILoad per hour of heat stress varies across states to a low of 4.4 (2558 ÷ 581) and a high 8.0 (25,597 ÷ 3185) units/h in Idaho and Texas, respectively, averaging 6.4 units/h across all states. Clearly, cows in Alabama, Florida, Louisiana, Mississippi, and Texas are severely affected both in duration and extent of heat stress in the absence of heat abatement. In Florida, for example, close to 50% of all annual hours are under temperature and humidity conditions resulting in heat stress. Nationally, the average dairy cow is exposed 14.1% of all annual hours to conditions of heat stress.
|
and 7 present the impact of heat stress on productivity of dairy replacements in the absence of heat abatement. The reduction in annual growth of young heifers varies across states with a low of 0.2 and a high of 7.9 kg/heifer per year in Wyoming and Texas, respectively. In older heifers, reduction in annual growth is least in Idaho, Maine, Montana, and Wyoming and greatest in Louisiana at 1.0 and 17.4 kg/heifer per year, respectively. Overall, replacement heifers are much less impacted by heat stress than dairy cows. Younger heifers have a higher THIthreshold, resulting in considerably fewer excess THILoad (2588 vs. 9337) than dairy cows. Similar results are obtained with yearlings, although the differences with dairy cows are of lesser magnitude.
|
|
. Overall, the magnitude of production losses is relatively small across all states. This is due to 1) the relatively high THIthreshold of beef cows, which is a consequence of their lower metabolic rate than dairy cows, and 2) breeding of beef cattle in the United States occurs primarily during the spring, a season of lesser heat stress.
|
. Most of US beef production occurs in the western part of the central plains (Table 2). Over 70% of all cattle finished in the United States are fed in Texas, Kansas, and Nebraska, which are three states with THILoad values above the average of other beef-producing states. With the exception of Texas and Oklahoma, the estimated annual GainLoss is less than 10 kg/yr, which is equivalent to seven additional days in the feedlot assuming a daily gain of 1.6 kg/animal.
Swine.
Without any heat abatement, sow productivity is severely affected by heat stress in many states, some of these states being important in pork production (Table 9). In Texas, for example, an estimated 18.8 additional days open per sow would result from unabated heat stress on a yearly basis. The two states with the greatest number of farrowings per year, North Carolina and Iowa, would incur losses of 7.2 and 5.2 additional days open per sow on a yearly basis.
|
). The two largest hog-producing states, North Carolina and Iowa, have heat stress durations and extents that are somewhat close to the national average, resulting in GainLoss of 2.9 and 2.0 kg/animal per year.
Poultry.
Broiler performance is not affected markedly across all states even in the absence of heat abatement (Table 10). The GainLoss per 1000 birds is in all instances less than 0.5% of the total weight of bird produced. This is simply because the duration and extent of heat stress in broilers is relatively low across all states due to a high THIthreshold in broilers.
|
).) Layers produce approximately 25,000 dozen of standard eggs (60 g) per 1000 birds per year. Thus, the EGGLoss in Florida, for example, amounts to 7.3% of total potential yearly production. The range in loss of productivity is predictably large, with the least being 118 and the greatest 1807 dozen of standard eggs lost per 1000 birds per year in New York and Florida, respectively.
|
). Growth loss is minimum in Vermont and maximum in Texas, at 6 and 153 kg of GainLoss per 1000 birds per year, respectively. Relative to total growth, however, GainLoss from heat stress represents less than 1.5% of annual turkey production of approximately 10,000 kg per 1000 birds.
|
. Optimality of heat abatement was defined as minimum total economic losses, i.e., the greatest gain in revenues from heat abatement after subtracting the costs in that heat abatement system. Specifically, it is the least sum of DMILoss, MilkLoss, GainLoss, EGGLoss, DOLoss, RCullRate, and PDeath summed over all animals within an animal class in a given state and converted to dollar losses, plus the sum of capital and operating costs of a given heat abatement system for that given animal class in that given state. This optimality criterion is not to be confused with maximum reduction in production losses, which, in most instances, would result from the intensive heat abatement. For example, an intensive heat abatement system would reduce California MilkLoss more than a high abatement system (5 vs. 154 kg/cow per year), but the total economic value of this additional reduction plus the net effect on DMILoss, DOLoss, RcullRate, and PDeath is less than the additional $ 86.7 million of annual capital costs and $8.0 million of annual operating costs required by the intensive system (data not shown).
|
In beef production, losses in productivity do not justify any heat abatement in any of the states for both beef cows and finishing cattle (Table 14). These results are not surprising, considering the extensive nature of beef cow production. On a national basis, heat stress results in $87.0 million in total losses to the beef breeding herd, which translates to a small $2.60/cow per year. Even in Texas, a state with significant heat stress and $33.2 million in annual losses, the amount of loss per cow is estimated at $6.07/cow per year or less than 1.5% of annual gross income per cow (data not shown). The failure of any heat abatement intensity to be justified economically in finishing cattle is more surprising, considering the large economic cost estimated at $282 million per year nationally. This figure translates to $12/animal per year on a national basis, or approximately 1.5% of gross income per animal (data not shown). Other advantages associated with the current major beef-producing states, such as lower feed costs, probably far outweigh the economic loss from heat stress. Additionally, beef producers can practice low input cooling strategies, such as ground wetting, that are very low cost and have been shown to be effective at reducing heat stress (Mader, 2002).
|
). Intensive heat abatement is optimal for the two largest sow-producing states, North Carolina and Iowa. Although optimal heat abatement does improve animal performance, the economic loss due to heat stress is not reduced considerably: $97 vs. $113 million per year nationally. Our model of losses in sows only accounted for losses in the form of additional days open in sows. The effect of heat stress on litter weight is not well defined, and young piglets seem to exhibit considerable compensatory gains in the 2 wk postweaning (Renaudeau and Noblet, 2001; Renaudeau et al., 2001). Additional data are needed in this area because a negative impact on the weight of piglets would increase the estimated losses to heat stress in sows considerably.
The economic losses in growing-finishing pigs are noticeably more than in sows (Table 14). Heat abatement would optimally be required in North Carolina but not in Iowa. The economic effectiveness of heat abatement is very small in grow-finish hogs. Essentially, the gains in productivity are nearly all negated by the additional capital and operating costs. Nationally, total economic losses in grow-finish pigs are estimated at $202 million per year. Combined with sow production, annual losses to the swine industry are estimated at $299 to $316 million, depending on the proportion of the production achieved under optimal heat abatement intensity.
In poultry, economic losses in broiler production never justify the additional cost of heat abatement (Table 15). Nationally, the annual total economic losses are estimated at $ 51.8 million, a very small amount in an industry that generates an estimated $20 to $25 billion of gross revenue per year.
|
). High heat abatement intensity is economically optimal in all states. Optimum heat abatement reduces annual total economic losses from $98.1 to $61.4 million. In turkey production, total annual losses are estimated at $14.4 million nationally, with little effect of heat abatement intensity. This loss seems insignificant in an industry that generates approximately $4 billion in gross returns per year.
Across all animal classes, the estimated national annual losses to heat stress are estimated at $2.4 billion in the absence of heat abatement and $1.7 billion under optimum heat abatement intensity. The actual number would be bounded by these two values and would be dependent on the proportion of all livestock raised under optimal heat abatement intensity. Considering the magnitude of the errors in estimating the effects of heat stress on animal performance, the national estimate of losses should be rounded to $2 billion per year.
Nationally, losses under optimum heat abatement intensity average 71.9% of estimated losses without heat abatement (Figure 3). This proportion varies considerably across the nation depending on the nature of the production, the severity of heat stress, and the efficiency of the optimal system (Figure 3).
|
Limitations
Some of the limitations to our knowledge on the effects of heat stress on animal productivity have been previously identified. There are many areas in which the mechanisms of heat stress are relatively well understood but for which the quantification of the response is poor (e.g., animal mortality). The paucity of information regarding the probability of mortality across major farm species given specific environmental conditions makes the quantification of this loss difficult. The integration of all major factors involved in creating heat stress is still very much incomplete. The THI scale is a weighted average of dry-bulb temperature (65%) and wet-bulb temperature (35%). Possibly, the weights assigned to each component should vary among species (Ravagnolo and Mistal, 2000) and may include nonlinear terms. The carryover effects of heat stress and the acclimation of animals seem important, yet the quantification of these two processes is difficult and generally lacking.
The model that we developed had as a primary objective the quantification of the total economic losses to heat stress across all major food-producing animals in the United States. Aggregating weather data to the state level induced some errors that were negligible in this context. There is a need, however, to design models for decision support at the farm level. These models will require much less aggregated weather data because enough climatic variation exists within many states to induce variation in the optimal cooling system within states and species.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
|
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Received for publication August 8, 2002. Accepted for publication December 2, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESEARCH AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
