Gorga wildfire (2011, Alicante, Spain).
Purpose
The main purpose of the project is to evaluate the temporal variation of soil water repellency depending on a number of factors (severity of fire upon vegetation type, soil type and weather conditions) in Mediterranean soils affected by forest fires. The implications of water availability on the soil and the vegetation response, and the hydrologic behavior of these soils and the role of ashes are other key objectives of this project. It is also intended to extend the knowledge of changes in water repellency of fire-affected soils, trying to answer questions that still have no clear answer.
The project tries to provide support for continuing and improving methodology recently developed by the GEA on the estimation of fire severity on soil (reached temperatures and heating times) using near-infrared spectroscopy (NIR). Fire severity is a factor closely related to the changes that occur in soil as a result of burning, and is strongly linked with the occurrence, changes or destruction of soil water repellency. Both research groups in this project have a background of years of experience in the study of soils affected by forest fires, and in recent years, the study of soil water repellency has been part of other research projects. The project intends to continue and develop this research line regarding specifically the study of soil water repellency and its implications for the recovery of the affected areas.
The project tries to provide support for continuing and improving methodology recently developed by the GEA on the estimation of fire severity on soil (reached temperatures and heating times) using near-infrared spectroscopy (NIR). Fire severity is a factor closely related to the changes that occur in soil as a result of burning, and is strongly linked with the occurrence, changes or destruction of soil water repellency. Both research groups in this project have a background of years of experience in the study of soils affected by forest fires, and in recent years, the study of soil water repellency has been part of other research projects. The project intends to continue and develop this research line regarding specifically the study of soil water repellency and its implications for the recovery of the affected areas.
Experimental area at Gorga (Alicante, Spain).
What is soil water repellency?
Under certain circumstances, some soils may develop water repellency (hydrophobicity) and are considered hydrophobic. Infiltration may be delayed or limited in water repellent soils (Brandt, 1969). Depending on the persistence of hydrophobicity water will penetrate the surface after a period of time (a few seconds to hours or even days if extreme repellency exists). Schreiner and Edmund (1910) were the first to observe and describe this property. According to Doerr et al. (2000), the grade of affinity/repellency between water and a solid surface is caused by a balance between the forces of adhesion between both phases and the cohesion between water molecules. Thus, when the adhesion forces outweigh cohesion the liquid readily penetrates the soil. When the adhesion forces are smaller than cohesion, the water stays on the soil surface, and it is said that such soils are water repellent, and an angle greater than or equal to 90 degrees is formed between the surface of the solid and the water . It is generally assumed that a soil is water repellent if a drop of water takes longer than 5 seconds to be fully absorbed by the soil (Dekker and Jungerius, 1990).
Effects of soil water repellency
Water repellency has great implications on soil behaviour. The delay of infiltration at the surface can improve runoff rates and accelerate water erosion processes (Imeson et al., 1992), increasing the transfer of pollutants and nutrients by the generation of preferential flow paths (Deban, 2000), decrease production of certain crops as well as affect microbial activity and seed germination (Doerr et al., 2002).
Hydrophobicity has been observed by different authors virtually all over the world (Doerr et al., 2000). The accumulation of certain organic compounds such as aliphatic hydrocarbons and amphiphilic compounds, are responsible for soil water repellency (Doerr et al., 2000). Several extraction methods have been used to determine the chemical composition of compounds that may be related to the presence of water repellency (Franco et al., 2000, Doerr et al., 2005, Morley et al., 2005), although their accurate composition is far from fully established. Horne and McIntosh (2000) used different extraction procedures in sandy soils of New Zealand, and identified neutral lipids, consisting primarily of alkanes and triglycerides, fatty lipids, long chained fatty acids and a soluble fraction with amphiphilic character and with an important role in the development of hydrophobicity. It has been found that the degree of hydrophobicity depends on the soil water content and wetting and drying cycles that determine the orientation of amphiphilic compounds on the surface of minerals (Ma'shum and Farmer, 1985, Horne and McIntosh, 2000). Mainwaring et al. (2004) concluded that the water repellent soil samples they studied contain hydrophobic polar compounds of high molecular mass, and that the presence of this compound, in sufficient quantities, were responsible for soil water repellency. Morley et al. (2005) also found that the presence of polar compounds of high molecular weight, fatty acids / amides (with 23-24 carbon atoms), differentiated between hydrophilic and hydrophobic samples. Franco et al. (2000) found that some hydrophobic compounds in soil samples taken had a chemical composition similar to that encountered for materials from Eucalyptus sp.
Hydrophobicity has been observed by different authors virtually all over the world (Doerr et al., 2000). The accumulation of certain organic compounds such as aliphatic hydrocarbons and amphiphilic compounds, are responsible for soil water repellency (Doerr et al., 2000). Several extraction methods have been used to determine the chemical composition of compounds that may be related to the presence of water repellency (Franco et al., 2000, Doerr et al., 2005, Morley et al., 2005), although their accurate composition is far from fully established. Horne and McIntosh (2000) used different extraction procedures in sandy soils of New Zealand, and identified neutral lipids, consisting primarily of alkanes and triglycerides, fatty lipids, long chained fatty acids and a soluble fraction with amphiphilic character and with an important role in the development of hydrophobicity. It has been found that the degree of hydrophobicity depends on the soil water content and wetting and drying cycles that determine the orientation of amphiphilic compounds on the surface of minerals (Ma'shum and Farmer, 1985, Horne and McIntosh, 2000). Mainwaring et al. (2004) concluded that the water repellent soil samples they studied contain hydrophobic polar compounds of high molecular mass, and that the presence of this compound, in sufficient quantities, were responsible for soil water repellency. Morley et al. (2005) also found that the presence of polar compounds of high molecular weight, fatty acids / amides (with 23-24 carbon atoms), differentiated between hydrophilic and hydrophobic samples. Franco et al. (2000) found that some hydrophobic compounds in soil samples taken had a chemical composition similar to that encountered for materials from Eucalyptus sp.
Soil water repellency is influenced by vegetation
Hydrophobicity has been studied with special emphasis on soils under coniferous and eucalyptus, particularly after a fire. However, the number of works on other types of Mediterranean scrub vegetation is still low (Martinez-Zavala and Jordan-López, 2009). In Mediterranean climate areas, under humid microclimate conditions or sub-humid areas, the abundant production of biomass and soil acidity are factors that trigger the hydrophobicity. In turn, how t hese factors influence soil hydrophobicity is determined by vegetation. The hydrophobicity is often associated with particular plant species, fungi and other soil microorganisms, although this does not mean that these species always act with the same intensity or in the same direction (Doerr et al., 2000). The way in which various plant species favour or not hydrophobicity has to do with the amount and type of organic residues accumulated in the soil, as root exudates (Dekker and Ritsema, 1996), washed compounds from plant leaves (Doerr et al., 2000; Debano, 2000) or products of the decomposition of organic matter (McGhie and Posner , 1981). Different shrubs, trees, and grasses have been related to the presence of hydrophobicity in soils (Doerr et al., 2000).
In unburned forest soils in the province of Alicante it has been observed that about 20-30% of the samples studied show water repellency. It was been found that this property was closely connected with the plant species present at each sampling point (Mataix-Solera et al., 2007), being higher in samples under pine (Pinus halepensis), compared with other species like the Kermes oak (Quercus coccifera), juniper (Juniperus oxycedrus) or rosemary (Rosmarinus officinalis). This heterogeneous pattern dependent on plant species present was also observed by Jordan et al. (2008), in Mediterranean soils of western Andalusia. In heathlands from Sierra del Algibe (Cádiz and Málaga), dominated by Erica australisand Calluna vulgaris, a strong to severe soil water repellency was reported, whereas the degree of repellency was lower under oaks (Quercus suber and Q. canariensis) or acebuchal (Olea europea). The presence of strongly hydrophobic organic compounds in tissues of Ericaor Calluna or soils associated with them (Carballeira, 1980) and the low rate of mineralization of its organic wastes, can explain these results.
Thus, the presence of a mosaic of vegetation types in the same hillslope allows different zones to serve as areas of runoff generation or infiltration in function of the degree of repellency at their surface. In the case of these Mediterranean heaths, the existence of hydrophobic compounds associated with humic substances may partly explain the high degree of hydrophobicity in soil (Martinez-Zavala and Jordan-López, 2009). The great power of acidification of plant residues from Ericaceae (Halal & Read, 1987, Nielsen et al., 1987; Mallik, 1995) and its slow decomposition contribute to an increase in the organic matter content of the soil and stimulates the development of hydrophobicity. With regard to microorganisms, Schantz and Piemeisel (1917) found relationship between hydrophobicity and the presence of a dense accumulation of fungal hyphae. They also suggested that different proteins produced by fungi, such as glomalin, can induce hydrophobicity in soils (Rillig, 2005). With regard to fire affected soils, water repellency may be altered depending on the temperature reached during burning (Deban et al., 1976). This effect has been observed by several authors since the late '60s (Deban et al., 1970). During a fire, in the first centimeters of soil, distillation of certain organic compounds is produced, and some of these gases can move into the soil and condense around the aggregates and mineral particles where temperatures are lower (Deban et al., 1970). After a fire, the study of water repellency is of great interest because it is responsible, along with other factors, for increased surface runoff and soil erosion (Martin and Moody, 2001).
The presence of water repellency in semiarid soils, where the amount of water is usually low, may be a factor in the soil water balance. The general pattern found in Mediterranean calcareous soils shows that fire increases soil water repellency (Mataix-Solera and Doerr, 2004; Arcenegui et al., 2008b). We have found that the presence of hydrophobicity is around 20-30% of the samples from burned soils in areas from southeastern Spanish, and these values increase to 70-80% immediately after fire, (Mataix-Solera et al ., 2007; Arcenegui et al., 2008b). The values found, still very variable, generally are lower with respect to the persistence of water repellency found in acid pH soils (Doerr et al., 1996, Varela et al., 2005). It is common to observe samples with low degree of water repellency (10-30 seconds using the test of time from penetration of water droplets, WDPT), and others with extreme values (> 3600 seconds) . The high variability found is not surprising given that fire can induce, enhance or destroy the hydrophobicity depending on several factors including the severity of fire, soil texture and mineralogy and the type and amount of vegetation (Doerr et al. 2000; Arcenegui et al., 2007).
In unburned forest soils in the province of Alicante it has been observed that about 20-30% of the samples studied show water repellency. It was been found that this property was closely connected with the plant species present at each sampling point (Mataix-Solera et al., 2007), being higher in samples under pine (Pinus halepensis), compared with other species like the Kermes oak (Quercus coccifera), juniper (Juniperus oxycedrus) or rosemary (Rosmarinus officinalis). This heterogeneous pattern dependent on plant species present was also observed by Jordan et al. (2008), in Mediterranean soils of western Andalusia. In heathlands from Sierra del Algibe (Cádiz and Málaga), dominated by Erica australisand Calluna vulgaris, a strong to severe soil water repellency was reported, whereas the degree of repellency was lower under oaks (Quercus suber and Q. canariensis) or acebuchal (Olea europea). The presence of strongly hydrophobic organic compounds in tissues of Ericaor Calluna or soils associated with them (Carballeira, 1980) and the low rate of mineralization of its organic wastes, can explain these results.
Thus, the presence of a mosaic of vegetation types in the same hillslope allows different zones to serve as areas of runoff generation or infiltration in function of the degree of repellency at their surface. In the case of these Mediterranean heaths, the existence of hydrophobic compounds associated with humic substances may partly explain the high degree of hydrophobicity in soil (Martinez-Zavala and Jordan-López, 2009). The great power of acidification of plant residues from Ericaceae (Halal & Read, 1987, Nielsen et al., 1987; Mallik, 1995) and its slow decomposition contribute to an increase in the organic matter content of the soil and stimulates the development of hydrophobicity. With regard to microorganisms, Schantz and Piemeisel (1917) found relationship between hydrophobicity and the presence of a dense accumulation of fungal hyphae. They also suggested that different proteins produced by fungi, such as glomalin, can induce hydrophobicity in soils (Rillig, 2005). With regard to fire affected soils, water repellency may be altered depending on the temperature reached during burning (Deban et al., 1976). This effect has been observed by several authors since the late '60s (Deban et al., 1970). During a fire, in the first centimeters of soil, distillation of certain organic compounds is produced, and some of these gases can move into the soil and condense around the aggregates and mineral particles where temperatures are lower (Deban et al., 1970). After a fire, the study of water repellency is of great interest because it is responsible, along with other factors, for increased surface runoff and soil erosion (Martin and Moody, 2001).
The presence of water repellency in semiarid soils, where the amount of water is usually low, may be a factor in the soil water balance. The general pattern found in Mediterranean calcareous soils shows that fire increases soil water repellency (Mataix-Solera and Doerr, 2004; Arcenegui et al., 2008b). We have found that the presence of hydrophobicity is around 20-30% of the samples from burned soils in areas from southeastern Spanish, and these values increase to 70-80% immediately after fire, (Mataix-Solera et al ., 2007; Arcenegui et al., 2008b). The values found, still very variable, generally are lower with respect to the persistence of water repellency found in acid pH soils (Doerr et al., 1996, Varela et al., 2005). It is common to observe samples with low degree of water repellency (10-30 seconds using the test of time from penetration of water droplets, WDPT), and others with extreme values (> 3600 seconds) . The high variability found is not surprising given that fire can induce, enhance or destroy the hydrophobicity depending on several factors including the severity of fire, soil texture and mineralogy and the type and amount of vegetation (Doerr et al. 2000; Arcenegui et al., 2007).
Effects of fire temperature
It is well known that fire severity is one of the most important factors controlling the degree of water repellency in burned soils. This property can be induced or increased if the temperature reached is between 200 and 250 ºC (Osborn et al., 1964) or destroyed if the temperature in the soil during burning is between 270 and 300 ºC (Debano et al., 1976). These temperature ranges may vary in other soils, and also depend on the time of residence of these temperatures. We have seen also that the type of vegetation and the amount of burned fuel affect repellency (Arcenegui et al., 2007), and in the same way noted the importance of soil properties as factors controlling the degree of hydrophobicity that may develop as a result of combustion. The content of soil organic matter, the presence of clay and its mineralogy are key factors that seem to prevent the occurrence of water repellency in certain soils (Arcenegui et al., 2007,Mataix-Solera et al., 2008).
Effects of soil mineralogy
One of the most recent work from our group on this topic focused on checking whether the low susceptibility for developing hydrophobicity by heat observed in some cases by soil type was a general pattern due to soil properties. The soil type in question is commonly known as "terra rossa", mainly classified as Rhodoxeralfs in Soil Taxonomy (Soil Survey Staff, 2006) or Chromic Luvisols in the WRB (FAO, 2006). Samples were taken from this type of soil from 14 different zones and after controlled laboratory experiments heating the soil at different temperatures with and without addition of litter, it was concluded that it was indeed a common pattern, and this soil type is much less susceptible to develop water repellency. Moreover mineralogical and physico-chemical analysis led us to conclude that the reasons for this behavior were: i) a lower content of organic matter than other types of forest soils in the region, ii) an increased presence of clay in the soil, and iii) a high presence of kaolinite in the clay fraction (Mataix-Solera et al., 2008).
Effects of soil texture and structure
Regarding the distribution of water repellency among different size fractions of aggregates, it has been found that the finest fraction show higher values of repellency (Mataix-Solera et al. 2002; Mataix-Solera and Doerr, 2004; Arcenegui et al., 2007). This is consistent with previous studies by other authors (Bisdom et al., 1993, De Jonge et al., 1999). Crockford (1991) conversely found that the coarse fraction were the most hydrophobic. It remains unclear why this distribution occurs in both unburned and burned soils. Hydrophobic substances can be found coating particles or soil aggregates (Doerr et al., 2000) or freely between mineral particles and soil aggregates (Franco et al., 1995). As noted above, the source of these compounds is mainly vegetation. Mataix-Solera and Doerr (2004) suggested that the sieving of samples can concentrate hydrophobic individual particles in the finest fractions. Another possible explanation is that distillated gases can condense on particles and soil aggregates making them hydrophobic. Therefore, the finest fractions having higher specific surface will be more hydrophobic. It is suspected that this has implications for soil fertility as an impediment to contact with the water of this fine fraction exists in soils when thay are not saturated with water. In addition, these processes can be leading to an inadequate cation exchange necessary for plant nutrition.
Resprouting after a wildfire (Santiago de Compostela, 2010).
Effect of ash
During the combustion process taking place in a fire, a great part of the litter is consumed, while some of the released energy is transferred to the soil surface. In such cases, non consumed litter and large amounts of ash are deposited on the soil surface (Neary et al., 2005). The effect of litter debris and ash on the soil mineral surface persist until external agents (heavy rain, wind, etc.) spread or redistribute them. Cerdà and Doerr (2008) analyzed and quantified the effect of the ashes and the litter layer in the generation of runoff and soil erosion risk by simulation of rain in the immediate period (10 days) after a wildfire in Sierra Calderona (Valencia). Subsequently, Zavala et al. (2009a) analyzed the hydrological response of soil and the hydrophobicity of the exposed surface of the soil during the first 7 days after an experimental fire in the Sierra de Algeciras (Cádiz) and compared these results with those observed a year later, after recovery of vegetation. Although fires are frequently associated with increased risk of erosion, after this work it was suggested that small-scale erosion risk is limited to a certain degree while the layers of ash and leaves remain on the soil surface. When a thick layer of ash is formed after a fire, it may delay runoff generation and reduce the erosion risk.
Recently, however the wettability of the ash generated by combustion is being investigated, and we have proved that under controlled laboratory conditions and in the field, ashes can also be hydrophobic under certain circumstances. Ash is considered the material resulting from biomass combustion, in which organic residues are found (charcoal, soot, partially charred material) and mineral matter. It has been shown that both plant species and the degree of combustion are factors that influence the wettability/water repellency of ashes. The ashes resulting from a weak combustion of species such as pine (Pinus halepensis) and Kermes oak (Quercus coccifera), have shown properties that affect soil water repellency (Bodí et al., 2009). There is currently a research line underway as the role of ash in these soils is one of those least studied by scientists. In the short term the ashes play a role in the hydrology of the affected area and this can influence the good regeneration of post-fire vegetation cover.
Recently, however the wettability of the ash generated by combustion is being investigated, and we have proved that under controlled laboratory conditions and in the field, ashes can also be hydrophobic under certain circumstances. Ash is considered the material resulting from biomass combustion, in which organic residues are found (charcoal, soot, partially charred material) and mineral matter. It has been shown that both plant species and the degree of combustion are factors that influence the wettability/water repellency of ashes. The ashes resulting from a weak combustion of species such as pine (Pinus halepensis) and Kermes oak (Quercus coccifera), have shown properties that affect soil water repellency (Bodí et al., 2009). There is currently a research line underway as the role of ash in these soils is one of those least studied by scientists. In the short term the ashes play a role in the hydrology of the affected area and this can influence the good regeneration of post-fire vegetation cover.
New approaches: using near-infrared spectroscopy (NIR) to estimate fire severity on soil
Fire intensity can be defined as the rate of energy release in a forest fire, and this may vary spatially and temporally depending on such factors as the quantity and type of fuel, moisture, atmospheric conditions, etc. The impact of a fire in the soil may be more or less severe depending on the intensity of the fire, but also depending on soil behavior, environmental conditions and its conditions during burning (eg, characteristics of soil). The severity of fire is one of the most important aspects in the study of areas affected by wildfires. The development of methodologies for quantifying, estimating temperatures reached, residence times of certain temperatures or severity indices that combine the two parameters is of greatest interest (Lewis et al., 2006). Most degradation processes are associated with the severity of the fire (Neary et al., 1999, Lewis et al., 2006). The changes occurring in soil properties are dependent inter alia on the severity of the fire. Recently, members from the GEA group (UMH) have succeeded in developing a methodology for estimating the maximum temperatures reached (MTR) in burned soils (Guerrero et al., 2007b; Arcenegui, 2008a), and are now attempting to optimize this technique to improve the reliability when applied to field samples (Arcenegui et al., 2008a, 2008b).
A spectrum of soil in the near infrared region is dominated by weak overtones and combinations of vibrational bands mainly caused by O-H, C-H, N-H, S-H and C=O type links, so the NIR provides information on the relative proportions of these links, which are the main constituents of organic molecules (Cozzolino and Moron, 2004). Many soil properties are modified by the effect of fire, and are generally dependent on the temperature reached (Raison, 1979; Pietikäinen et al., 2000; Certini, 2005, Guerrero et al., 2005; Arcenegui et al. 2008a). This made possible the development of models to estimate the maximum temperature reached on burned soils (Guerrero et al., 2007b). Models were constructed with soils from different areas of the province of Alicante burned at different temperatures and during different periods in the laboratory. The MTR was recorded by monitoring it in each of the burned samples by using thermocouples. Subsequently the spectra were obtained in the NIR using a Fourier-Transform near infrared spectrophotometer (FT-NIR) equipped with a prism and PbS detector (MPA, Bruker Optik GmbH, Germany). Once the NIR spectra from these samples were obtained, spectral information was related to the peak temperature by chemometric analysis of. As a result, several models were built and various types of validations were carried out: the first one by cross-validation and the second one by external validation. For example, models were bult using all samples from all zones, models with samples from four areas (validated with a control area), or "local models" using only samples from one area were built. In all cases excellent models were obtained to estimate the maximum temperature reached (MTR) (Guerrero et al. 2007b).
These models were developed with burned soil samples in the laboratory, and we then wondered what would happen in a field sample after a wildfire, where ashes are present? Would ashes affect the estimation of the MTR on the soil? With this objective we designed experiments in which ashes from different species, at different doses and with varying degrees of combustion were mixed with burned soil samples with known MTR.Arcenegui et al. (2008a) showed that the presence of ash interfered with the proper assessment previously obtained from the MTR in soils. Despite this, a very good classification of samples burnt at different temperatures was performed using a discriminant analysis, showing that the presence of ashes is a potential problem that can be easily solved.
Recently, we have investigated how soil moisture content would affect the estimation of MTR. Depending on the treatment prior to processing a sample, it has a certain moisture content. For example, the soil moisture content is different depending on whether the sample is air-dried or oven-dried. In addition, the modifications produced on the soil during burning, affect the moisture content, and we know that these changes are different depending on the severity of the fire. It is known that NIR spectra are sensitive to moisture content of samples. In order to understand how moisture affects the estimation and what is the most appropriate processing of the burned samples to estimate more precisely the MTR in burnt soils experiments in which models were built and validated using samples with different moisture content were designed (Arcenegui et al., 2008a). We used air-dried and oven-dried samples, taking into account that the fire may affect some properties and condition re-wetting, re-wetted and both air-dried and oven-dried samples were also used, taking into account the possible scenario that a storm rewets soil between fire and sampling. The results showed that depending on whether we know or not if the soil has been wetted after the fire, and selecting the most suitable pretreatment of the samples, models to accurately estimate the MTR can be built.
Therefore we are at the stage of testing this methodology with real burned samples in the field to validate, improve and perfect it, and to make progress in the development of models to estimate the length of time a soil has reached a certain temperature and try to combine these two parameters (temperatures reached and time) to propose indexes of fire severity. The project that we propose, working on real fire-affected areas and studying a property so linked to the severity of the fire as water repellency, represents a great opportunity to also continue with this line of maximum interest in the study of fire-affected areas.
A spectrum of soil in the near infrared region is dominated by weak overtones and combinations of vibrational bands mainly caused by O-H, C-H, N-H, S-H and C=O type links, so the NIR provides information on the relative proportions of these links, which are the main constituents of organic molecules (Cozzolino and Moron, 2004). Many soil properties are modified by the effect of fire, and are generally dependent on the temperature reached (Raison, 1979; Pietikäinen et al., 2000; Certini, 2005, Guerrero et al., 2005; Arcenegui et al. 2008a). This made possible the development of models to estimate the maximum temperature reached on burned soils (Guerrero et al., 2007b). Models were constructed with soils from different areas of the province of Alicante burned at different temperatures and during different periods in the laboratory. The MTR was recorded by monitoring it in each of the burned samples by using thermocouples. Subsequently the spectra were obtained in the NIR using a Fourier-Transform near infrared spectrophotometer (FT-NIR) equipped with a prism and PbS detector (MPA, Bruker Optik GmbH, Germany). Once the NIR spectra from these samples were obtained, spectral information was related to the peak temperature by chemometric analysis of. As a result, several models were built and various types of validations were carried out: the first one by cross-validation and the second one by external validation. For example, models were bult using all samples from all zones, models with samples from four areas (validated with a control area), or "local models" using only samples from one area were built. In all cases excellent models were obtained to estimate the maximum temperature reached (MTR) (Guerrero et al. 2007b).
These models were developed with burned soil samples in the laboratory, and we then wondered what would happen in a field sample after a wildfire, where ashes are present? Would ashes affect the estimation of the MTR on the soil? With this objective we designed experiments in which ashes from different species, at different doses and with varying degrees of combustion were mixed with burned soil samples with known MTR.Arcenegui et al. (2008a) showed that the presence of ash interfered with the proper assessment previously obtained from the MTR in soils. Despite this, a very good classification of samples burnt at different temperatures was performed using a discriminant analysis, showing that the presence of ashes is a potential problem that can be easily solved.
Recently, we have investigated how soil moisture content would affect the estimation of MTR. Depending on the treatment prior to processing a sample, it has a certain moisture content. For example, the soil moisture content is different depending on whether the sample is air-dried or oven-dried. In addition, the modifications produced on the soil during burning, affect the moisture content, and we know that these changes are different depending on the severity of the fire. It is known that NIR spectra are sensitive to moisture content of samples. In order to understand how moisture affects the estimation and what is the most appropriate processing of the burned samples to estimate more precisely the MTR in burnt soils experiments in which models were built and validated using samples with different moisture content were designed (Arcenegui et al., 2008a). We used air-dried and oven-dried samples, taking into account that the fire may affect some properties and condition re-wetting, re-wetted and both air-dried and oven-dried samples were also used, taking into account the possible scenario that a storm rewets soil between fire and sampling. The results showed that depending on whether we know or not if the soil has been wetted after the fire, and selecting the most suitable pretreatment of the samples, models to accurately estimate the MTR can be built.
Therefore we are at the stage of testing this methodology with real burned samples in the field to validate, improve and perfect it, and to make progress in the development of models to estimate the length of time a soil has reached a certain temperature and try to combine these two parameters (temperatures reached and time) to propose indexes of fire severity. The project that we propose, working on real fire-affected areas and studying a property so linked to the severity of the fire as water repellency, represents a great opportunity to also continue with this line of maximum interest in the study of fire-affected areas.