When air temperatures fall below 0 °C, sensitive crops can be injured, with significant effects on production. For example, in the USA, there are more economic losses to frost damage than to any other weather-related phenomenon (White and Haas, 1975). Therefore, impacts on affected farmers and the local economy are often devastating. Although it is clearly important, information on how to protect crops from freezing is relatively limited. Consequently, there is a need for a widely available, simplified source of information to help farmers address this serious problem. In this book, the distribution, economics, history, physical and biological aspects of frost damage are presented and discussed, together with methods of protection.
This book contains a broad range of information from basic to complex; however, it was mainly written to help growers to better understand freeze protection and to develop strategies to combat crop losses due to freezing. References are provided for those who want to further investigate the science of frost protection. However, the objective is to provide a guidebook for practitioners, rather than a literature review. Because some aspects of frost protection are complex, user-friendly computer programs for some applications are included with the book. In addition, useful information on simple, inexpensive measurements and applications using charts and tables are provided, along with the algorithms used to make them.
For those readers who are mainly interested in management rather than science, read Chapter 2 on Recommended Methods of Frost Protection, which provides relatively non-technical information on all aspects of freeze protection. For those readers who want more detailed explanations, Chapters 3 to 8 thoroughly discuss most aspects of frost protection, including the scientific basis. Volume II of this book covers the probability, risk and economics of frost protection. While there is useful information for meteorologists, the book covers neither mesoscale or synoptic scale forecasting nor frost risk modelling. These are reviewed in other, more technical, publications (e.g. Kalma et al., 1992). However, for the local grower and farm advisor, this book should provide most of the information needed to make wise decisions about frost protection, thus helping growers and local communities to minimize the devastating effects of frost damage.
Technically, the word "frost" refers to the formation of ice crystals on surfaces, either by freezing of dew or a phase change from vapour to ice (Blanc et al., 1963; Bettencourt, 1980; Mota, 1981; Cunha, 1982); however, the word is widely used by the public to describe a meteorological event when crops and other plants experience freezing injury. Growers often use the terms "frost" and "freeze" interchangeably, with the vague definition being "an air temperature less than or equal to 0 °C". Examples of frost definitions in the literature include:
the occurrence of a temperature less than or equal to 0 °C measured in a "Stevenson-screen" shelter at a height between 1.25 and 2.0 m (Hogg, 1950, 1971; Lawrence, 1952);
the occurrence of an air temperature less than 0 °C, without defining the shelter type and height (Raposo, 1967; Hewett, 1971);
when the surface temperature drops below 0 °C (Cunha, 1952); and the existence of a low air temperature that causes damage or death to the plants, without reference to ice formation (Ventskevich, 1958; Vitkevich, 1960).
Snyder, Paw U and Thompson (1987) and Kalma et al. (1992) have defined frost as falling into two categories: "advective" and "radiative". Advective frosts are associated with large-scale incursions of cold air with a well-mixed, windy atmosphere and a temperature that is often subzero, even during the daytime (Table 1.1). Radiative frosts are associated with cooling due to energy loss through radiant exchange during clear, calm nights, and with temperature inversions (i.e. temperature increases with height). In some cases, a combination of both advective and radiative conditions will occur. For example, it is not uncommon to have advective conditions bring a cold air mass into a region, resulting in an advection frost. This may be followed by several days of clear, calm conditions that are conducive to radiation frosts. In addition, the authors have observed conditions that are considered as "micro-scale-advection frosts". These occur when the region is exposed to radiation-type frost conditions, but local cold air drainage leads to rapid drops in temperature on a small scale within the radiation frost area.
TABLE 1.1
Frost event terminology and typical
characteristics
FROST TYPE |
CHARACTERISTICS |
Radiation |
Clear; calm; inversion; temperature greater than 0 °C during day |
Advection |
Windy; no inversion; temperature can be less than 0 °C during day |
Freeze and frost definitions in dictionaries and in the literature are variable and confusing; however, on a worldwide basis, the term frost protection is more commonly used than freeze protection. Based on the literature, it was decided that the following definitions are appropriate and will be used in this book.
A "frost" is the occurrence of an air temperature of 0 °C or lower, measured at a height of between 1.25 and 2.0 m above soil level, inside an appropriate weather shelter. Water within plants may or may not freeze during a frost event, depending on several avoidance factors (e.g. supercooling and concentration of ice nucleating bacteria). A "freeze" occurs when extracellular water within the plant freezes (i.e. changes from liquid to ice). This may or may not lead to damage of the plant tissue, depending on tolerance factors (e.g. solute content of the cells). A frost event becomes a freeze event when extracellular ice forms inside of the plants. Freeze injury occurs when the plant tissue temperature falls below a critical value where there is an irreversible physiological condition that is conducive to death or malfunction of the plant cells. This damaging plant tissue temperature is correlated with air temperatures called "critical temperatures" measured in standard instrument shelters. Subzero air temperatures are caused by reductions in sensible heat content of the air near the surface, mainly resulting from (1) a net energy loss through radiation from the surface to the sky (i.e. radiation frost); (2) wind blowing in subzero air to replace warmer air (i.e. advection frost); or (3) some combination of the two processes.
Radiation frosts are common occurrences. They are characterized by a clear sky, calm or very little wind, temperature inversion, low dew-point temperatures and air temperatures that typically fall below 0 °C during the night but are above 0 °C during the day. The dew-point temperature is the temperature reached when the air is cooled until it reaches 100 percent relative humidity, and it is a direct measure of the water vapour content of the air. To illustrate the difference between advection and radiation frost, data from the two worst frost events in the twentieth century in the main California citrus growing region are shown in Figures 1.1 and 1.2. Notice that the daytime maximum temperatures dropped considerably as cold air moved into the region. Based on wind speed, it would not be considered an advection frost event, because there was little or no wind during the night, when temperatures were subzero. However, because it was cloudy during the first few days of the events, the subzero temperatures are attributed to advection of cold air into the area rather than resulting from a net radiation loss. Similar events to the two frosts had occurred previously in 1913 and 1937, so they are relatively rare occurrences. However, this may not be the case in more continental climate areas where subzero temperatures are more common.
Under clear night-time skies, more heat is radiated away from the surface than is received, so the temperature drops. The temperature falls faster near the radiating surface causing a temperature inversion to form (i.e. temperature increases with height above the ground). The process is shown in Figure 1.3. As there is a net loss of energy through radiation from the surface, the sensible heat content of the soil surface and air near the surface decreases. There is a flux of sensible heat downward from the air and upward from within the soil to the surface to replace the lost sensible heat. This causes the temperature to decrease aloft as well, but not as rapidly as at the surface. The depth to the top of the temperature inversion is variable depending on local topography and weather conditions, but generally ranges from 9 to 60 m (Perry, 1994).
FIGURE 1.1
Mean air and dew-point temperatures at 1.5 m
height and mean wind speed at 2.0 m height recorded during the December 1990
event at Lindcove, California, USA
FIGURE 1.2
Mean air and dew-point temperatures at 1.5 m
height and mean wind speed at 2.0 m height during the December 1998 event at
Lindcove, California, USA
FIGURE 1.3
Development of an inversion over an apple
orchard in northern Portugal
If air temperature is measured at a sufficient height above the soil surface, it will reach the point where it begins to decrease with height (a lapse condition). The level where the temperature profile changes from an inversion to a lapse condition is called the ceiling. A weak inversion (high ceiling) occurs when the temperatures aloft are only slightly higher than near the surface and a strong inversion (low ceiling) has rapidly increasing temperature with height. Energy-intensive protection methods are most effective during the low ceiling, strong inversion conditions that are typical of radiation frosts.
There are two subcategories of radiation frosts. A "hoar frost" occurs when water vapour deposits onto the surface and forms a white coating of ice that is commonly called "frost". A "black" frost occurs when temperature falls below 0 °C and no ice forms on the surface. If the humidity is sufficiently low, then the surface temperature might not reach the ice point temperature and no frost will form. When the humidity is high, ice is more likely to deposit and a "hoar frost" can occur. Because heat is released during the ice deposition process, hoar frosts usually cause less damage than black frosts.
Note that the plots of daily air temperature for the December 1990 and 1998 frosts in California (Figures 1.1 and 1.2) had similar shapes in both years; however, the dew-point temperature trends were different in the two years. Because the air temperature plots have a similar shape during most radiation frost nights, a good approximation for changes in night-time air temperature can be made with an empirical model. However, because of variability, it nearly impossible to generalize about dew-point temperature changes during the night.
One clear characteristic of air temperature on radiation frost nights is that most of the temperature drop occurs in a few hours around sunset, when the net radiation on the surface rapidly changes from positive to negative. This rapid change in net radiation occurs because solar radiation decreases from its highest value at midday to zero at sunset, and the net long-wave radiation is always negative. This is explained in more detail in Chapter 3. Figure 1.4 shows typical temperature, radiation and soil heat flux density trends during a radiation frost night. In this example, the temperature fell about 10 °C during the first hour after the net radiation became negative. After the net radiation reached its most negative value, the temperature only fell 10 °C more during the remainder of the night. The rate of temperature change was small (e.g. less than 1.0 °C h-1) from two hours after sunset until sunrise.
FIGURE 1.4
Air (Ta) and dew-point
(Td) temperatures at 1.5 m height, net radiation
(Rn) and soil heat flux density (G) measured in a
walnut orchard in Indian Valley in Northern California, USA
Advection frosts occur when cold air blows into an area to replace warmer air that was present before the weather change. It is associated with cloudy conditions, moderate to strong winds, no temperature inversion and low humidity. Often temperatures will drop below the melting point (0 °C) and will stay there all day. Because many of the active protection methods work better in the presence of an inversion, advection frosts are difficult to combat. In many cases, a series of subzero nights will start as an advection frost and will later change to radiation frost nights. For example, the major California frosts of 1990 and 1998 shown in Figures 1.1 and 1.2 both started as advection frost events. Although the wind speeds were low, there were cloudy conditions from 18 to 20 December 1990 and from 18 to 22 December 1998. However, the temperature still fell to minimums well below 0 °C during these periods. After the skies cleared (i.e. 21-25 December 1990 and 23-25 December 1998), the subzero temperature resulted from radiation losses rather than advection of cold air.
Major frosts occur in Mediterranean climates, but they tend to be more common in the eastern part of continents where cold continental air masses occasionally advect from arctic regions into subtropical areas. Some of the best examples are in the Florida, USA, citrus growing region. Attaway (1997) describes the first "major impact" frost, which occurred in 1835, by citing John Lee Williams' account of the frost, which stated that "the northwest wind blew for 10 days and the temperature fell as low as -13.9 °C. Even the local river froze and all kinds of fruit trees were killed to the ground as far south as 28 °N latitude." Clearly, there is a big difference when trying to protect against subzero temperatures in windy conditions without an inversion than to protect against a relatively mild radiation frost. The saving grace is that major frost events tend to be sporadic, whereas radiation frost events occur often.
Frost protection techniques are often separated into indirect and direct methods (Bagdonas, Georg and Gerber, 1978), or passive and active methods (Kalma et al., 1992). Passive methods are those that act in preventive terms, normally for a long period of time and whose action becomes particularly beneficial when freezing conditions occur. Active methods are temporary and they are energy or labour intensive, or both. Passive methods relate to biological and ecological techniques, including practices carried out before a frost night to reduce the potential for damage. Active methods are physically based and energy intensive. They require effort on the day preceding or during the night of the frost event. Active protection includes heaters, sprinklers and wind machines, which are used during the frost night to replace natural energy losses. A classification of methods is presented in Table 1.2.
Frost damage can occur in almost any location, outside of tropical zones, where the temperature dips below the melting point of water (0 °C). The amount of injury depends on the crop's sensitivity to freezing at the time of the event and the length of time the temperature is below the "critical damage" temperature (Tc). For example, Argentina, Australia, Canada, Finland, France, Greece, Israel, Japan, Jordan, New Zealand, Portugal, Switzerland, United States of America and Zambia have developed minimum temperature forecasting techniques (Bagdonas, Georg and Gerber, 1978) to aid frost protection. Of course, many other countries in temperate and arid climates and at high elevations also have problems with frost damage.
TABLE 1.2
Categories and sub-categories for methods of
frost protection
CATEGORY |
SUB-CATEGORY |
PROTECTION METHOD |
Passive |
Biological (avoidance or resistance) |
Induction of resistance to freezing without modifying plant genetics |
Treatment of the seeds with chemicals |
||
Plant selection and genetic improvement |
||
Selecting species for timing of phenological development |
||
Selecting planting dates for annual crops after the probability of freezing lessens in the spring |
||
Growth regulators and other chemical substances |
||
Ecological |
Site selection for cropping |
|
Modification of the landscape and microclimate |
||
Controlling nutritional status |
||
Soil management |
||
Cover crop (weed) control and mulches |
||
Active |
Covers and Radiation |
Organic materials Covers without supports |
Covers with supports |
||
Water |
Over-plant sprinklers |
|
Under-plant sprinklers |
||
Microsprinklers |
||
Surface irrigation |
||
Artificial fog |
||
Heaters |
Solid fuel |
|
Liquid fuel |
||
Propane |
||
Wind machines |
Horizontal |
|
Vertical |
||
Helicopters |
||
Combinations |
Fans and heaters |
|
Fans and water |
To a large extent, the potential for frost damage depends on local conditions. Therefore, it is difficult to present a geographical assessment of potential damage. The average length of the frost-free period, which lasts from the occurrence of the last subzero temperature in the spring to the first in the autumn, is sometimes used to geographically characterize the potential for damage.
A world map of average length of frost-free period (Figure 1.5) clearly shows that the greatest potential for frost damage increases as one moves poleward. Only at latitudes between the tropics of Cancer and Capricorn are there relative large areas with little or no subzero temperatures. Even in these tropical areas, frost damage sometimes occurs at high elevations. Damage is somewhat less likely when the land mass is downwind or surrounded by large bodies of water, because of the moderating effect of the maritime environment on humidity and temperature, and hence temperature fluctuations and dew or frost formation.
Although the map of the average length of frost-free period provides a useful general guide as to where the potential for frost damage is greater, it is not a detailed map. Again, the probability of freezing temperatures is affected by local conditions that cannot be properly shown on a global map. In fact, farmers can experience some economic losses from frost damage even if it occurs infrequently.
Although outside of the scope of this book, considerable effort has recently been expended in improving the geographical characterization of regional-scale frost damage risk. Kalma et al. (1992) published an extensive review on the geographical characterization of frost risk. For example, Lomas et al. (1989) prepared an atlas of frost-risk maps for Israel. They used more than 25 years of temperature data and topographical information to develop the maps, which clearly show a close relationship between elevation and risk of subzero temperature. Others have used mobile temperature surveys or topographical and soil information, without temperature data, to derive risk maps. Case studies on developing a frost-risk map using an elevation model were presented by Kalma et al. (1992) based on Laughlin and Kalma (1987, 1990), and by Zinoni et al. (2002b).
While more and better spatial information on risk of frost damage is needed, there is no substitute for good local information and monitoring. Most farmers have a good idea about the location of cold spots in their locality. It is definitely worthwhile to consult neighbours before planting sensitive crops at a specific site. Generally, low spots, where cold air ponds, should be avoided. Also, avoid areas where the natural or modified topography dams cold air drainage from the site. Because ground fog forms in low spots first, a good rule of thumb is to avoid places where ground fog forms early. Definitely, one should review local topographical maps before planting frost-sensitive crops on high-risk sites. For example, because bloom is late, there is rarely a need for frost protection of walnut orchards in California, but the authors have noted that a few orchards that are planted in cold spots commonly experience damage. This could easily have been avoided by checking local weather records and topographical maps. Site selection is discussed in more detail later, in the section on passive protection.
FIGURE 1.5
Geographical distribution of the average
length of frost free period. See the file: "Frost free map.jpeg" on the programs
CD to view the distribution in colour
More economic losses are caused by freezing of crops in the USA than by any other weather hazard. In the State of Florida, the citrus industry has been devastated by frost damage on several occasions, resulting in fruit and tree costing billions of dollars (Cooper, Young and Turrell, 1964; Martsolf et al., 1984; Attaway, 1997). In California, the December 1990 frost caused about $ 500 million in fruit losses and damage to about 450 000 ha of trees (Attaway, 1997). There was about $ 700 million in damage during the December 1998 frost (Tiefenbacher, Hagelman and Secora, 2000). Similarly, huge economic losses to other sensitive horticultural crops are frequently observed throughout the world.
For example, Hewitt (1983) described the effects of freezing on coffee production in Brazil during the 1960s and 1970s. Winterkill of cereals is also a major problem (Stebelsky, 1983; Caprio and Snyder, 1984a, 1984b; Cox, Larsen and Brun, 1986).
Although the losses to farmers can be huge, there are also many secondary effects on local and regional communities. For example, if there is no fruit to pick, the pickers are unemployed, the processors have little or no fruit, so their employees are unemployed, and, because of unemployment, there is less money in circulation and the local economy suffers. Consequently, considerable effort is expended to reduce damage.
The cost-effectiveness of frost protection depends on the frequency of occurrence, cost of the protection method and the value of the crop. Generally, passive frost protection is easily justified. The cost-effectiveness of active protection depends on the value of the crop and cost of the method. In this book, both passive and active methods are discussed, as well as the economics of protection.
Frost damage to crops has been a problem for humans since the first crops were cultivated. Even if all aspects of crop production are well managed, one night of freezing temperatures can lead to complete crop loss. Except for tropical latitudes, where temperatures seldom fall below the melting point, damage due to freezing temperatures is a worldwide problem. Usually, frost damage in subtropical climates is associated with slow moving cold air masses that may bring 2-4 nights of 8-10 hours of subzero temperature (Bagdonas, Georg and Gerber, 1978). In eastern continental locations, damaging events are typically advective, with weak inversions. In western continental and marine climates, frost events with calm conditions and stronger inversions are more typical. The damaging events typically start with advection of cold air followed by a few nights of radiation frost. In temperate climates, frost periods are shorter in duration and occur more frequently than in other climates (Bagdonas, Georg and Gerber, 1978).
For deciduous fruit and nut trees, damaging frost events occur mainly in the spring, but sometimes in the autumn as well. For subtropical fruits, damage to the crops typically occurs during the winter. In tropical climates, there is normally no freezing except at higher elevations. Therefore, when tropical crops are damaged by cold, the temperature is usually above zero. When the damage occurs at temperatures above 0 °C, it is called "chilling" rather than "freeze" injury. In temperate climates, damage to grain crops can also occur before booting, under severe conditions, or to flowers even in mild frosts.
For grain farmers, the main response is to plant crops or varieties that are less susceptible to damage (e.g. planting spring wheat rather than winter wheat), or to not plant sensitive crops in the area if damage occurs too frequently. In any case, the date of planting should be adjusted to the crop, variety and microclimate. Similarly, if subzero temperatures occur too frequently, subtropical crops are preferentially grown in regions with less occurrence of damage. A good example of this is the movement of the citrus industry further south in Florida in response to several severe frosts during the 1980s and 1990s (Attaway, 1997). At the same time, due to more favourable temperatures, the olive industry is moving northward in Italy where soil and climate factors allow for production of high quality olive oil. However, this has led to an increase in frost damage to olives during severe winters in 1985, 1991 and 1996 (Rotondi and Magli, 1998). Generally speaking, the dates of the last frost occurrence in the spring and the first occurrence in the autumn will determine where particular crops are grown. For example, many of the deciduous fruit and nut crops tend to be grown in Mediterranean climates because the probability of losing a crop to frost damage is less than in more continental climates. The science of frost protection has mainly developed in response to the occurrence of intermittent damage in relatively favourable climates. If the damage occurs regularly, the best strategy is to grow the crop elsewhere, in a more favourable location.
In some cases, cropping locations change in response to climate change. For example, Attaway (1997) noted that prior to 1835 orange trees were commonly grown in South Carolina, Georgia and northern Florida, where, because of potential losses to frost damage, people today would not consider commercial production of oranges. He cited several examples of subtropical orchards that had survived up until about 1835, when a severe frost occurred. In fact, there were citations of documents recommending that subtropical fruits be grown in the American southeast to help compete with fruit produced in Mediterranean countries of Europe. With today's climate, subtropical fruit production would not be considered in these areas. Attaway (1997) makes the point that his observations are based on grower experience rather than climatology, but fewer damaging frost events must have occurred during the 1700-1800s for farmers to be producing subtropical fruits where none can be economically produced today.
The history of frost damage is more sporadic in the Mediterranean climate of California. There have been some major losses from time to time, but the diversity of crops and timing of the frosts leads to less extensive impacts in California. Recently, California suffered two major damaging events in the citrus industry. One occurred in December 1990 and the other in December 1998. The 1990 frost caused the most damage to citrus production since the 1913 and 1937 frosts (Attaway, 1997). Interestingly, some regions had little damage, while others were devastated. Attaway (1997) noted that, although the damage to fruit was immense,
"most trees were in relatively good condition although they had endured temperatures which would have killed trees in Florida. We attribute this to the fact that morning lows in the upper 20s and low 30s [i.e. between about -4 °C and +2 °C] had occurred for the two weeks prior to the frost, putting the trees in an almost completely dormant state."
The December 2000 frost was a good example of how hardening can provide protection against frost damage. In Florida, before a cold front passes and drops the air to subzero temperatures, relatively warm temperature often precedes a severe frost. Consequently, the trees are less hardened against frost damage than those exposed to the two California frosts. Interestingly, Attaway (1997) emphasized the inconsistent nature of frost damage that was observed following the frost. For example, within a relatively small region, he noted losses of 70 to 80 percent of the oranges in Ojai Valley, 60 percent to 70 percent losses in Santa Paula Canyon, but only 20 percent losses in the Santa Clara Valley, which is relatively close. This illustrates the site-specific nature of frost damage to crops, especially in hilly and mountainous regions like Ventura County in California.
The December 1998 frost was not as bad for California citrus growers as that of 1990; however, it still is considered one of the major frosts of the twentieth century. The economic losses were high; however, unlike the 1990 frost, most growers were able to survive (Tiefenbacher, Hagelman and Secora, 2000). In their review of the December 1998 frost in California's San Joaquin Valley, Tiefenbacher, Hagelman and Secora (2000) noted that there was a clear relationship between latitude and damage and latitude and harvesting in anticipation of a frost. They noted that more northerly orchards suffered more frost damage, but they also harvested considerably earlier than the first frost, which allowed them to survive with less economic loss. They also noted a relationship between longitude and the age and size of orchards, which is also related to elevation. In the San Joaquin Valley, older orchards are located on the east side at higher elevations, with younger orchards to the west at lower elevation in the Valley. The reviewers recommended that micrometeorological models, combined with digital elevation data and detailed damage information, could help to understand spatial patterns of damage risk.
Tiefenbacher, Hagelman and Secora (2000) observed that larger operations proportionally lost more crop production, whereas smaller growers and cooperative members lost less. This was partially attributed to communication between cooperative organizations and the fact that many small growers harvested before the frost. After the 1990 frost, many farmers began to purchase catastrophic crop insurance and growers with insurance experienced more damage in 1998. This might have occurred because their orchards are more prone to damage or it might be that there was less effort to use protection methods because they had insurance. The answer is unknown. In addition, Tiefenbacher, Hagelman and Secora (2000) noted that government disaster assistance might be influencing frost protection activities by growers. In both 1990 and 1998, the government provided disaster funding to help growers recoup their losses. While this disaster relief is helpful to the farmers, it might discourage the use of active protection methods and it might encourage expansion of the industry into areas where the risk of frost damage is higher (Tiefenbacher, Hagelman and Secora, 2000).
Historically, heaters have been used to protect plants from freezing for more than 2000 years (Powell and Himelrick, 2000). Originally, the heaters were mostly open fires; however, in recent history, metal containers for the fire were used to better retain the heat for radiation and convection to the crop. Powell and Himelrick (2000) wrote that about 75 percent of the energy from stack heaters is used to directly heat the air, which then is convected to the crop directly or indirectly by mixing with air within the inversion layer. They attributed the additional 25 percent of energy as transferring from the heater stacks to the plants as direct radiation, which is effective even during advection frost events.
The earliest known metal-container heaters (i.e. stack heaters or smudge pots) for frost protection were introduced by W.C. Scheu in 1907 in Grand Junction, Colorado, USA. He found an oil-burning device for heating that was more efficient than open fires. It later became known as the HY-LO orchard heater, which was produced by the Scheu Manufacturing Company, which today produces portable space heaters. Even before the HY-LO orchard heater, growers used simple metal containers that burned heavy oils or old rubber tyres containing sawdust. These fires produced considerable oily smoke that for a long time was believed to provide protection against freezing by blocking net radiation losses from the surface. In fact, it is now known that little or no protection is afforded by adding smoke particles to the air with orchard heaters (Mee and Bartholic, 1979). The use of orchard heaters was standard practice worldwide for some time, but the smoke was terribly polluting and the use of smoke-producing orchard heaters was later banned in the USA for health and environmental reasons. It took a strong public outcry to eventually eliminate the use of smoke-producing heaters. For example, the Pasadena Star-News, 20 October 1947, published a request from Louis C. McCabe, director of the newly formed Los Angeles Air Pollution Control District, to eliminate smoke from more than 4 million orchard heaters. The Orange County Air Pollution Control District and seven other Districts in California adopted regulations banning the use of dirty fuels and smoke-producing smudge pots (SCAQMD, 2002).
In the USA, growers were given a few years to find a less polluting method of frost protection. Eventually, the "return stack" heater, which recirculates smoke and vapour, was developed and used for some time (Leonard, 1951). Today, return stack heaters and clean-burning propane-fuel heaters are legal in many locations; however, before using any type of heater, local regulations should be checked. However, the perception of increased fuel costs and pollution issues during the mid-1900s has led to the demise of most heaters for frost protection. During the 1950s, wind machines began to replace heaters as the preferred method of frost protection. They were more expensive to purchase, but the labour and operational costs were lower. By the 1970s, the use of heaters for frost protection was almost non-existent in California. Small fires and solid-fuel heaters are still used in some parts of the world. However, it is likely that the use of all but clean burning heaters will stop eventually.