Assessment of soil respiration patterns in an irrigated corn field based on spectral information acquired by field spectroscopy

Assessment of soil respiration patterns in an irrigated corn field based on spectral information acquired by field spectroscopy

Agriculture, Ecosystems and Environment 212 (2015) 158–167 Contents lists available at ScienceDirect Agriculture, Ecosystems and Environment journal...

2MB Sizes 0 Downloads 50 Views

Agriculture, Ecosystems and Environment 212 (2015) 158–167

Contents lists available at ScienceDirect

Agriculture, Ecosystems and Environment journal homepage: www.elsevier.com/locate/agee

Assessment of soil respiration patterns in an irrigated corn field based on spectral information acquired by field spectroscopy Víctor Cicuéndeza , Manuel Rodríguez-Rastrerob , Margarita Huescaa , Carla Uribeb , Thomas Schmidb , Rosa Inclánb , Javier Litagoc , Víctor Sánchez-Girónd , Silvia Merino-de-Miguele, Alicia Palacios-Oruetaa,* a

Departamento de Sistemas y Recursos Naturales, E.T.S.I.M., Universidad Politécnica de Madrid, Spain Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas, Madrid, Spain c Departamento de Economía Agraria, Estadística y Gestión de empresas, E.T.S.I.A., Universidad Politécnica de Madrid, Spain d Departamento de Ingeniería Agroforestal, E.T.S.I.A., Universidad Politécnica de Madrid, Spain e Departamento de Ingeniería y Gestión Forestal Ambiental, E.T.S.I.A. E.T.S.I.M., Universidad Politécnica de Madrid, Spain b

A R T I C L E I N F O

A B S T R A C T

Article history: Received 31 October 2014 Received in revised form 21 April 2015 Accepted 25 June 2015 Available online xxx

The assessment of soil respiration processes in agroecosystems is essential to understand the C balance and to study the effects of soil respiration on climate change. The use of spectral data through remote sensing techniques constitutes a valuable tool to study ecological processes such as the C cycle dynamics. The objective of this work was to evaluate the potential to assess total (Rs) and autotrophic (Ra) soil respiration through spectral information acquired by field spectroscopy in a row irrigated corn crop (Zea mays L.) throughout the growing period. The relationships between Rs and Ra with leaf area index (LAI), spectral indexes and abiotic factors (soil moisture and soil temperature) were assessed by linear regression models using the adjusted coefficient of determination (Radj2) to measure and compare the proportion of variance explained by the models. Results showed significant differences and a high variability in the relationships between Rs and Ra with spectral indexes within the corn field during the phenological stages and in measurements under the plants and between the rows. Best results were obtained when assessing Ra during vegetative stages. However, during the reproductive stages, spectral indexes were better related to Rs which could be related to the presence of rhizomicrobial respiration linked to vegetation activity. Spectral indexes contain significant functional information, beyond mere structural changes, that could be related to carbon fluxes. However, specific models should be applied for the different phenological stages and there is a need to be cautious when upscaling remote sensing models. The results obtained confirm that in irrigated crop systems remote sensing data can produce relevant information to assess both Rs and Ra through spectral indexes. ã 2015 Elsevier B.V. All rights reserved.

Keywords: Soil respiration Autotrophic soil respiration Spectral indexes Remote sensing Field spectroscopy Phenological stages Corn

1. Introduction The assessment of the CO2 balance and the production dynamics through photosynthesis and respiration processes in ecosystems has become a fundamental research issue due to the potential impact that this may have on the climate change (Cramer et al., 2001; Dixon et al., 1994; Knapp and Smith, 2001). Agricultural activities are responsible for approximately 13% of the anthropogenic emissions of greenhouse gases (Pachauri and

* Corresponding author at: Alicia Palacios-Orueta Departamento de Sistemas y Recursos Naturales, ETSIM, Universidad Politécnica de Madrid, Ciudad Universitaria, 28040 Madrid, Spain. E-mail address: [email protected] (A. Palacios-Orueta). http://dx.doi.org/10.1016/j.agee.2015.06.020 0167-8809/ã 2015 Elsevier B.V. All rights reserved.

Reisinger, 2007). Intensive agricultural practices have significantly increased in the last century due to human overpopulation, to satisfy market requirements and to ensure farmers benefits causing serious social and environmental impacts (FAO, 2002; Matson et al., 1997; Tilman et al., 2002). Therefore, it is essential to look for improved agricultural practices which reduce the impacts of agroecosystems (Sanchez-Martín et al., 2010; Snyder et al., 2009). Soil respiration (Rs) is responsible for 60–90% of the total ecosystem respiration (Goulden et al., 1998, 1996). The production of CO2 by soil respiration is the second most important flux of C in the majority of ecosystems after the photosynthesis process (Raich and Schlesinger, 1992; Schlesinger and Andrews, 2000). Mainly, Rs is composed of autotrophic respiration (Ra) from the metabolism of roots and of heterotrophic respiration (Rh) from microorganisms

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

that decompose soil organic matter (Kuzyakov, 2006; Prolingheuer et al., 2010). In agricultural fields spatial distribution of Rs components can be highly variable depending on the type of crop spatial distribution. Rs is higher close to the roots in row crops due to a higher proportion of autotrophic respiration (Amos et al., 2005; Rochette et al., 1991). Prolingheuer et al. (2010) found highly different spatial structure in autotrophic and in heterotrophic components in a winter wheat field, and suggested that an adequate spatial sampling design would be needed in order to assess accurately soil respiration patterns. Thus, the assessment of these soil respiration processes is essential to understand the C balance and to study the effects of soil respiration on climate change (Buchmann, 2000; Han et al., 2007). It is important to know whether agroecosystems act as sinks or sources of CO2 over time and space depending on the various factors that can interact on them such as management practices (Lal, 2011; Mosier et al., 2005; Paustian et al., 1997; Sauerbeck, 2001; Uribe et al., 2013). The improvement of technology has allowed quantifying more accurately the flux of CO2 between the atmosphere and ecosystems. In this sense, the use of spectral data through remote sensing techniques constitutes a valuable tool to study ecosystem’s dynamics and ecological processes for improving land use management and combine economic and environmental revenues. The availability of high frequency remote sensing time series providing continuous land surface observations allows characterizing and monitoring ecosystem’s phenology at different temporal and spatial scales (Palacios-Orueta et al., 2012; Tucker et al., 2001; Zhang et al., 2003). Besides other applications, these capabilities are extremely useful to assess the C cycle dynamics (Dong et al., 2003; Heinsch et al., 2006; Turner et al., 2004). However, there are few studies that are focused on estimating Rs from remote sensing products and this is mostly carried out indirectly through spectral indexes related to ecosystem activity (Huang and Niu, 2013; Huang et al., 2012). Spectral data is usually summarized as spectral indexes which are easy to calculate and could be used as input variables in biophysical models to calculate properties such as moisture content or leaf area index (LAI), and to relate them to physiological processes such as photosynthesis, respiration and gross and net primary production of ecosystems (Baret and Guyot, 1991; Viña et al., 2011; Wang et al., 2005). The most common indexes used are ratio indexes based on measurements obtained within the visible and near-infrared (VIS/NIR) region of the electromagnetic spectrum such as the normalized difference vegetation index (NDVI) (Tucker, 1979) and the enhanced vegetation index (EVI) (Huete et al., 2002). Both indexes are strongly related to photosynthetic activity and therefore offer an important and convenient measure for canopy photosynthesis (Gitelson et al., 2006; Huang and Niu, 2013; Sims et al., 2006). Since it has been shown that NDVI saturates at a high chlorophyll content, i.e., at high LAI values (Buschmann and Nagel, 1993; Gitelson et al., 2003b, 2002; Viña and Gitelson, 2005), EVI was developed to avoid this problem (Huete et al., 2002). Therefore, they seem to be an appropriate basis for assessing ecosystem functioning when vegetation is active during the growing period but probably not at other phenological stages (Palacios-Orueta et al., 2012). In terms of Rs, these indexes could be related to the autotrophic respiration (Ra) of roots due to the relationship between plant photosynthesis and root respiration during the growing period (Kuzyakov and Cheng, 2004, 2001; Xu et al., 2008); however this relationship may not hold in the same manner at different phenological stages (Huang et al., 2012) in which Rh is more significant (Prolingheuer et al., 2010). Indexes based on the shortwave infrared (SWIR) spectral region such as the normalized difference water index (NDWI) (Gao, 1996) are more directly related to soil and vegetation moisture capturing probably

159

different features in the phenological dynamics of the agroecosystems. In the recently proposed spectral shape indices (SSI) (Palacios-Orueta et al., 2006) the shape of the spectral signature in a specific wavelength range is represented using information from three bands instead of two. Depending on the spectral range, different applications of these indexes have been demonstrated (Khanna et al., 2013; Palacios-Orueta et al., 2012; Zhang et al., 2014). Among the SSI indexes, ANIR (the angle at the near-infrared band) is a combination of reflectance values in the Red, NIR and SWIR1 bands and it has been used in several applications (Khanna et al., 2007). The objective of this work was to evaluate the potential to assess total and autotrophic soil respiration through spectral information acquired by field spectroscopy in a row irrigated corn crop (Zea mays L.) throughout the growing period. In order to accomplish this objective, the following issues are taken into consideration:  The spatial variability of soil respiration within the corn field, specifically underneath the corn plants and between crop rows.  Differences in dynamics between vegetative and reproductive stages.  The influence of abiotic factors.

2. Material and methods 2.1. Study area The study has been carried out during two crop seasons (2011– 2012) in an irrigated experimental corn field of the College of Agricultural Engineering of the Technical University of Madrid (Central Spain). The study area is located at UTM X: 437410; UTM Y: 4476875 (Zone 30N; Datum ETRS89) at an altitude of 595 m a.s.l. and is characterized by a Mediterranean climate defined as Csa (temperate with dry or hot summer) according to the Köppen– Geiger classification (AEMET/IMP, 2011). The average annual temperature is 13.1  C (4.3  C in January and 24.1  C in July), and the mean annual precipitation is 455 mm, with autumn–spring maxima and a marked minimum in summer (12 mm in August). The experimental zone was developed in an area that was leveled approximately 30 years ago by adding earthy and rubble materials over a gentle slope. Therefore, the study area is characterized by an anthropogenic soil with higher pH and higher electrical conductivity than a natural soil which shows an A/C morphology. The A horizon is formed of a 30–40 cm layer, fine textured, with a neutral pH and moderate organic matter content. The C horizon consists of an anthropogenic deposit with a coarse texture, and rich in rubble materials which are constituted mainly by remnants of bricks, ceramic fragments and other construction materials (5% of the soil volume). Main soil parameters are obtained from 3 soil profiles made in the study site (A horizons: pH: 7.6; ECse: 2.8 dSm 1; CaCO3: 2.3%. C horizons: pH: 8.1; ECse: 1.6 dSm 1; CaCO3: 1.7%; and an average “sandy clay loam” sclass). The study area has been under cultivation approximately for 30 years. Over the last decade, the soil has been subjected to various rainfed and irrigated crops: barley, alfalfa, ryegrass or vetch and other crops such as corn, sunflower, cotton, or beet. Cultivation conditions include manuring during plough practices. Available data indicate an organic matter content of 1.8% in the surface horizon (1.0% of organic C). 2.2. Crop characteristics and phenological stages Maize (Zea mays L.) was sowed in mid-May in 2011 and lateApril in 2012. In both cases the plants were irrigated 3 h per week.

160

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

In all the study area including the bare soil plots, primary tillage consisted of moldboard plowing to an average depth of 35 cm followed by one disk harrowing to a depth of 10 cm for the seedbed preparation. While performing the latter operation, rabbit compost at a rate of 50–60 Mg ha 1 was incorporated to the soil as manure amendment. Prior to sowing, maize seeds (variety NK factor, Italy, purity 98%) were treated with fludioxonil (2.4%) and metalaxil-m (1%). In the maize plots, seeds were placed in rows spaced 70 cm apart and having in-row distance of 20 cm. The drill planter was equipped with fluted coulters and double disk furrow openers. Herbicide was banded behind the planter and it was also applied to the bare soil area. The main phenological stages according to Iowa State University (Hanway, 1971) were identified by visual inspection and recorded during the sampling campaigns. The Table 1 shows the Day of the Year (DOY) corresponding to each phenological stage for both years. Maize phenology is generally divided into the vegetative and the reproductive stages (Hanway, 1971; Viña et al., 2004). Thus, in this research the growing period has been divided into the corresponding stages. The vegetative stages are characterized by structural changes and the reproductive stages are characterized by biochemical changes. 2.3. Experimental design The phenological evolution of the corn crop has been assessed from April to October in both years in two different fields (Fig. 1). Each year, the study field (29  9 m2) was divided into three plots: two cultivated plots that are homogeneous in terms of soil and management (2011: MZ211 and MZ311; 2012: MZ212 and MZ312) and one bare soil used as a control plot (MZ111 and MZ112). Measurements were taken in two types of locations within each cultivated plot: at the base of the plants with-in rows (P) and between rows of plants (I). Additionally, measurements were taken on bare soil (S) in the control plot which was cleaned of weeds during the measurement campaign. 2.4. Soil respiration measurements Instantaneous soil CO2 efflux, i.e., Rs, was measured (between 9 a.m. and 12 p.m.) using a portable automated soil CO2 infrared gas analyzer (Li-8100, LI-COR Inc., Lincoln, NE, USA) equipped with a 10 cm survey chamber (Model 8100-102). Measurements were taken over PVC collars (105 mm in diameter and 90 mm high) which were inserted at 5–6 cm depth into the soil each year in the three plots. The collar area was cleaned of vegetation during the measurement campaign. In each plot of MZ2 and MZ3, four collars were located between plants in the same row (P) and four between plants from adjacent

Fig. 1. Location of the study plots for both years. Each year, the study field was divided into three plots: two cultivated plots (2011: MZ211 and MZ311; 2012: MZ212 and MZ312) and one bare soil used as a control plot (MZ111 and MZ112).

rows (I). Furthermore, five cylinders were distributed randomly in the bare soil plot MZ1 (S). The collars were located in different places each year. The autotrophic respiration (Ra) was estimated by a simple partition method (Kuzyakov, 2006) as the difference between the respiration measured at each site (Rs) and the respiration measured in the bare soil plot without plants (under the assumption that this one represented the heterotrophic respiration, Rh). 2.5. Spectroradiometric measurements Hyperspectral ground measurements between 350 and 2500 nm were taken weekly with an ASD FieldSpec3 Spectroradiometer (Analytical Spectral Devices Inc., Boulder, CO, USA), approximately between 12 p.m. and 4 p.m. and selecting days with clear sky conditions. Furthermore, the field was cleaned of weeds during the measurement campaign. In each plot of MZ2 and MZ3, measurements were acquired over five plants from the same row representing P sites and five sites between rows representing I sites. In addition, six points were selected in the bare soil plot MZ1 (S). Three measurements were taken with the spectroradiometer using the bare optical fiber with an instantaneous field of view of 44.3 cm over each maize plant canopy or bare soil surface. Measurements were made at a height of 70 cm over the plants and at a height of 1 m over the soil. Calibration of the spectroradiometer was carried out between measurements using a calibrated diffuse white reference panel. The measurements were corrected for sensor discrepancy using a splice correction (ViewSpec Pro 5.6). Thereafter, an average of each of the three measurements was carried out. Thereafter, a

Table 1 DOY of each phenological stage according to Iowa State University in the corn plots for the two years of the study.

Vegetative stages

Flowering and reproductive stages

Growth stage

Julian date 2011

S VE V6 V8 V11 V13

136 145 160

VT R1 R2 R3 R6

208 211 224 236 266

Julian date 2012

166 174 181 198 202 214 226 247

Note: S: Sowing; VE: emergence; Vx: vegetative growth where x indicates the number of leaves; VT: flowering stage; Rx: are the reproductive stages: R1: silking stage; R2: blister stage; R3: milk stage; and R6: kernels have reached their maximum dry matter accumulation.

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

Savitzky-Golay filter was applied using a second order polynomial with a window size of five spectral bands where the measurements between 1350 and 1460 nm and between 1790 and 1960 nm were eliminated to exclude atmospheric effects within the visible and near infrared spectra. Finally, the spectral indexes NDVI, EVI, NDWI, and ANIR were calculated (Table 2) after a resampling method to MODIS bands applying the software ENVI 4.2 (Research Systems Inc., Boulder, CO, USA).

161

Table 3 Names used to represent each of the variables.

Soil respiration (Rs) Leaf area index Spectral indexes

Plant (P)

Inter-row (I)

Bare soil (S)

P_Rs_MZyear P_LAI_MZyear P_Index_MZyear

I_Rs_MZyear I_LAI_MZyear I_Index_MZyear

S_Rs_MZyear S_Index_MZyear

3.1. LAI and Rs 2.6. Leaf area index Leaf area index measurements were used to assess structural changes in the crop throughout the growing period. The LAI data were measured over the vegetation period with a LAI-2200 Plant Canopy Analyzer equipment (LI-COR Inc., Lincoln, NE, USA) and data were processed using the FV2200 software (1.2.1. version, LI-COR Biosciences). In each plot of MZ2 and MZ3, eight measurements were collected at the base of the stem for each of the same five plants that were previously selected for obtaining the spectroradiometric data and eight measurements were taken over 5–10 cm of the ground in the corresponding Inter-row sites. In 2011, LAI was measured only until July (DOY 202) due to technical problems. 2.7. Soil temperature and moisture data Soil temperature (T10) and soil moisture (H10) data were obtained from RT-1 and 10HS sensors (Decagon Services Inc., WA, USA) and recorded in their corresponding Em5b Analog Data Loggers (Decagon Services Inc., WA, USA). The sensors were located at 10 cm depth in the in P and I sites in the MZ2 and MZ3 cultivated plots and in the bare soil plot MZ1. 2.8. Statistical analysis

Fig. 2a and b shows LAI and Rs evolution for the years 2011 and 2012, respectively. LAI values (m2 m 2) were higher in 2011 than in 2012 reaching maximum values of 3.5 m2 m 2 and 2.8 m2 m 2, respectively. In addition in 2011, maximum LAI showed similar values in P and I sites while in 2012 LAI values in P sites were always higher than those in the I sites. The standard deviation of the measurements showed the maximum values during the last vegetative stages being higher in the P than in I sites (2011: 0.43 m2 m 2 in P sites and 0.24 m2 m 2 in I sites; 2012: 0.43 in P sites and 0.33 m2 m 2 in I sites). LAI increased sharply from VE to V18 in 2011 and to V11 in 2012. In I sites, LAI started increasing around a month later than in the P sites, that is, when the green canopy was high enough to invade the space between the rows. In 2012, LAI remained constant during the first reproductive stages decreasing slightly during the last reproductive stages before harvest. The Rs (mmol m 2 s 1) dynamics was similar in the P and the I sites, increasing from the first stages of corn growth (VE–V6) to V13–V14 in 2011 and to V11 in 2012 and decreasing slowly through the reproductive stages (R1–R6). Values in I sites are slightly lower than those in P sites reaching similar values to bare soil after harvest. Rs in bare soil plots remained approximately constant during the growing period reaching values around

For all variables, average values from each location (P, I and S) were computed resulting in three time series per year. Table 3 shows the names used to represent each of the variables. The relationships between Rs, Ra, LAI, and spectral indexes were assessed by linear regression models using all the measurements from the data set, including both years. In addition, soil temperature and moisture were included in the models to evaluate their effect on Rs and Ra. The models were adjusted according to the least square method and an analysis of variance (ANOVA) was performed to evaluate the significance of the model by means of the F-statistic. Significance of model’s coefficients was evaluated by the T-statistic. The adjusted coefficient of determination (Radj2) was used to measure and compare the proportion of variance explained by the regression models. Statgraphics Centurion XVI (StatPoint Technologies Inc., VA, USA) and SAS software (SAS 9.4 Software, SAS Institute, NC, USA) were used for the statistical analyses. 3. Results Average values for all the variables measured for years 2011 and 2012 are shown in Figs. 2–4. The uncertainty in the measurements was assessed by the standard deviation; however, for clarity purposes was not included in the graphs. Table 2 The spectral indexes used based on MODIS wavelengths. Index

Value

Reference

NDVI EVI NDWI ANIR

(rnir rred)/(rnir + rred) (rnir rred)/(rnir + C1*rred C2*rblue + 1) (rnir rswir1)/(rnir + rswir1) Angle anir

Tucker (1979) Huete et al. (2002) Gao (1996) Khanna et al. (2007)

Fig. 2. Seasonal dynamics of leaf area index (LAI) and soil respiration (Rs) for the three types of locations: plant (P), inter-row (I) and bare soil (S) for both years 2011 (a) and 2012 (b) in an irrigated corn crop (Zea mays L.) in Madrid, Spain.

162

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

Fig. 3. Evolution of abiotic factors for both years for the three types of locations: plant (P), inter-row (I) and bare soil (S) in an irrigated corn crop (Zea mays L.) in Madrid, Spain: (a) Soil moisture determined at 10 cm depth and (b) soil temperature determined at 10 cm depth.

1.5 mmol m 2 s 1 in 2011 and around 3 mmol m 2 s 1 in 2012. The standard deviation showed maximum values during the last vegetative stages (2.51 mmol m 2 s 1 in P sites and 1.45 mmol m 2 s 1 in I sites) being higher in the P than in the I sites. In 2012, the standard deviation values were irregular during the growing period (values ranged between 0.39 and 1.81 mmol m 2 s 1 in P sites and 0.47 and 1.61 mmol m 2 s 1 in I sites). 3.2. Soil temperature and moisture data Fig. 3a and b shows the evolution of H10 and T10, respectively. Moisture values ranged between 0.20 and 0.40 m3 m 3 in both years. In 2011 maximum values occurred at the end of May and at the beginning of June, and showed a decrease since mid-August while in 2012 a clear trend could not be observed. Soil temperature showed similar dynamics in both years with an increase at the beginning of the crop and a decreasing trend since the V6–V8 stages. Values were slightly higher in 2011 than in 2012 and in the bare soil plot than in the cultivated plots. 3.3. Spectral indexes Fig. 4a–d shows the indexes evolution for the years 2011 and 2012 during the study period. In both sites NDVI, EVI and NDWI increased markedly from VE to V13 in 2011 and to V11 in 2012, maintaining close to constant values until R2 when they started decreasing slowly until harvest. The ANIR index showed the opposite behavior, decreasing during the vegetative stages and increasing during the reproductive stages.

Differences between the I and the P sites were significant at the beginning of the growing period for both years and similar to the LAI behavior. There is a clear difference in behavior between 2011 and 2012, while in 2011 indexes in I and P sites took on similar values after V13, in 2012 indexes at the I sites were consistently lower than those in the P sites. In all cases, the VE stage was not detected by indexes having similar values to those of bare soil. In the P sites, indexes began to discriminate green vegetation from bare soil when the maize plant had three leaves (V3), while in the I sites green vegetation was detected at the V6 stage when LAI of P sites was high enough to cover the space between rows. In general terms the measurements standard deviations ranged between 0.001 and 0.39 with maximum values in the crop during the vegetative periods and the beginning of senescence. Standard deviation in the bare soil plots showed slightly lower values than in vegetation. 3.4. Relationships between LAI, spectral indexes and Rs Table 4 shows the coefficients of determination of a linear regression model (LAI = a + b*index) between LAI and spectral indexes for the vegetative and reproductive stages separately. During the vegetative stages, all indexes present high coefficients of determination (Radj2) in both sites (P and I). During the reproductive stages, Radj2 are much lower in the P sites. However, in the I sites, Radj2 only decrease slightly. Table 5 shows the coefficients of determination of the simple linear regression models between the Rs, LAI and spectral indexes and of the multiple linear regression when H10 and T10 were also

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

163

Fig. 4. Spectral indexes values through the study period for both years for the three types of locations: plant (P), inter-row (I) and bare soil (S) in an irrigated corn crop (Zea mays L.) in Madrid, Spain. The spectral indexes are: (a) normalized difference vegetation index (NDVI), (b) enhanced vegetation index (EVI), (c) normalized difference water index (NDWI) and (d) the angle at the NIR band (ANIR).

164

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

Table 4 Coefficients of determination (Radj2) of a linear regression model between LAI and spectral indexes (LAI = a + b*Index) for both years during the vegetative and reproductive stages in an irrigated corn crop (Zea mays L.) in Madrid, Spain. Results are shown for both sites plant (P) and inter-row (I). Vegetative stages: structural changes Plant (P) NDVI EVI LAI 0.89 0.88 NDVI EVI Inter-row (I) LAI 0.77 0.76

NDWI 0.89 NDWI 0.82

ANIR 0.88 ANIR 0.77

Flowering and grain filling stages: biochemical changes Plant (P) NDVI EVI NDWI LAI 0.32 0.3 0.31 Inter-row (I) NDVI EVI NDWI 0.69 0.75 0.7 LAI

ANIR 0.18 ANIR 0.63

included in the models. In both cases, the growing period is divided into vegetative and reproductive stages. During the vegetative stages, the LAI explained more proportion of Rs variability than the spectral indexes in the P sites where the Radj2 increased slightly when the soil moisture was included in the model. On the contrary in the I sites, the LAI did not explain Rs at all, and the T10 together with spectral indexes could explain 25% of the Rs variability. During the reproductive stages up to 65% of the Rs variability was explained by indexes in both sites while the LAI did not explain any variability. By including the temperature, the models improved in the I sites whereas when H10 was included, the Radj2 increase only in the P sites. Table 6 shows the same type of results that Table 5 but with autotrophic respiration (Ra) instead of total soil respiration (Rs). During the vegetative stages in the P sites the LAI and the spectral indexes explained a high proportion of Ra variability. The inclusion of T10 and H10 in the models did not improve the Radj2. In the I sites, the coefficients of determination were lower; however, they increased considerably when T10 was included in the model. Table 5 Coefficients of determination (Radj2) of the simple and multiple linear regression models for both years during the vegetative and reproductive stages in an irrigated corn crop (Zea mays L.) in Madrid, Spain. Results are shown for both sites plant (P) and inter-row (I). The first column indicates that soil respiration (Rs) can be a function of only one of the indexes (LAI, NDVI, EVI, NDWI, or ANIR) (first row: Rs = a + b*Index), a function of one of the indexes with soil temperature determined at 10 cm depth (T10) (second row: Rs = a + b*Index + c*T10), with soil moisture determined at 10 cm depth (H10) (third row: Rs = a + b*Index + c*H10) or with T10 plus H10 (fourth row: Rs = a + b*Index + c*T10 + d*H10). Vegetative stages: structural changes Plant (P) LAI Rs = a + b*Index 0.52 Rs = a + b*Index + c*T10 0.51T Rs = a + b*Index + c*H10 0.58H Rs = a + b*Index + c*T10 + d*H10 0.57TH Inter-row (I) LAI Rs = a + b*Index 0.04* Rs = a + b*Index + c*T10 0.37 Rs = a + b*Index + c*H10 0.04*H Rs = a + b*Index + c*T10 + d*H10 0.37H

NDVI 0.4 0.36T 0.5 0.48T NDVI 0.09* 0.26 0.08*H 0.25H

EVI 0.39 0.36T 0.5 0.47T EVI 0.09* 0.26 0.08*H 0.25H

NDWI 0.41 0.38T 0.52 0.5T NDWI 0.07* 0.24 0.06*H 0 .23H

ANIR 0.44 0.41T 0.55 0.53T ANIR 0.06* 0.24 0.05*H 0.24H

Flowering and grain filling stages: biochemical changes Plant (P) LAI NDVI EVI Rs = a + b*Index 0.16* 0.62 0.61 Rs = a + b*Index + c*T10 0.11*T 0.61T 0.59T Rs = a + b*Index + c*H10 0.08*H 0.68 0.69 Rs = a + b*Index + c*T10 + d*H10 0.01*TH 0.67T 0.68T Inter-row (I) LAI NDVI EVI Rs = a + b*Index 0.05* 0.65 0.65 T Rs = a + b*Index + c*T10 0* 0.7 0.68T Rs = a + b*Index + c*H10 0.09*H 0.66H 0.65H Rs = a + b*Index + c*T10 + d*H10 0*TH 0.7TH 0.67TH

NDWI 0.56 0.54T 0.67 0.65T NDWI 0.56 0.68 0.55H 0.66H

ANIR 0.57 0.55T 0.63 0.62T ANIR 0.6 0.69 0.6H 0.68H

Significance of model’s coefficients was evaluated by the T-statistic (p < 0.05). *: index not significant, T: soil temperature not significant, and H: soil moisture not significant.

Table 6 Coefficients of determination (Radj2) of the simple and multiple linear regression models for both years during the vegetative and reproductive stages in an irrigated corn crop (Zea mays L.) in Madrid, Spain. Results are shown for both sites plant (P) and inter-row (I). The first column indicates that autotrophic respiration (Ra) can be a function of only one of the indexes (LAI, NDVI, EVI, NDWI, or ANIR) (first row: Ra = a + b*Index), a function of one of the indexes with soil temperature determined at 10 cm depth (T10) (second row: Ra = a + b*Index + c*T10), with soil moisture determined at 10 cm depth (H10) (third row: Ra = a + b*Index + c*H10) or with T10 plus H10 (fourth row: Ra = a + b*Index + c*T10 + d*H10). Vegetative stages: structural changes Plant (P) LAI Ra = a + b*Index 0.69 Ra = a + b*Index + c*T10 0.81 Ra = a + b*Index + c*H10 0.68H Ra = a + b*Index + c*T10 + d*H10 0.82H Inter-row (I) LAI Ra = a + b*Index 0.26 Ra = a + b*Index + c*T10 0.61 Ra = a + b*Index + c*H10 0.29H Ra = a + b*Index + c*T10 + d*H10 0.66H

NDVI 0.74 0.75T 0.73H 0.73TH NDVI 0.37 0.67 0.39H 0.7H

EVI 0.68 0.71T 0.67H 0.69TH EVI 0.37 0.68 0.39H 0.71H

NDWI 0.75 0.79T 0.75H 0.77TH NDWI 0.33 0.65 0.36H 0 .69H

ANIR 0.74 0.75T 0.73H 0.73TH ANIR 0.33 0.66 0.35H 0.7H

Flowering and grain filling stages: biochemical changes Plant (P) LAI NDVI EVI Ra = a + b*Index 0.34 0.46 0.44 Ra = a + b*Index + c*T10 0.28T 0.43T 0.42T H H Ra = a + b*Index + c*H10 0.42 0.45 0.42H Ra = a + b*Index + c*T10 + d*H10 0.36TH 0.43TH 0.4TH Inter-row (I) LAI NDVI EVI Ra = a + b*Index 0.59 0.44 0.4 Ra = a + b*Index + c*T10 0.59T 0.44T 0.4T Ra = a + b*Index + c*H10 0.55H 0.48H 0.41H Ra = a + b*Index + c*T10 + d*H10 0.58TH 0.49TH 0.43TH

NDWI 0.49 0.47T 0.48H 0.46TH NDWI 0.56 0.55T 0.59H 0.58TH

ANIR 0.36 0.34T 0.34H 0.31TH ANIR 0.51 0.5T 0.55H 0.55TH

Significance of model’s coefficients was evaluated by the T-statistic (p < 0.05). *: index not significant, T: soil temperature not significant and H: soil moisture not significant.

During the reproductive stages, in the P sites LAI and indexes explained smaller amount of Ra variability than in the vegetative stages while the opposite was observed in the I sites. The addition of T10 and H10 in the models did not improve the coefficients of determination. Table 7 shows the coefficients of determination of the linear regression models between both types of respiration and each abiotic factor (T10 and H10) by themselves and when they were both included in the model.

Table 7 Coefficients of determination (Radj2) of the simple linear regression model between total soil respiration (Rs) or autotrophic respiration (Ra) with each abiotic factor (T10 and H10) and the multiple linear regression model between total soil respiration (Rs) or autotrophic respiration (Ra) with both abiotic factors included in the model. Results are shown for three sites plant (P), inter-row (I) and bare soil (S). Vegetative stages: structural changes Plant (P) T10 Rs 0T Ra 0T Inter-row (I) T10 Rs 0.1T Ra 0T

H10 0H 0H H10 0H 0H

Flowering and grain filling stages: biochemical changes Plant (P) T10 H10 Rs 0.33 0.42 Ra 0.2 0.04H Inter-row (I) T10 H10 Rs 0.40 0H T Ra 0.03 0H Bare soil Bare soil (S) T10 H10 Rs 0.10T 0.77

T10+H10 0TH 0TH T10+H10 0.06TH 0TH

T10+H10 0.53 0.18H T10+H10 0.41H 0TH T10+H10 0.80T

Significance of model’s coefficients was evaluated by the T-statistic (p < 0.05). T: Soil temperature not significant and H: soil moisture not significant.

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

165

4. Discussion

4.2. Statistical relationship between LAI and spectral indexes

4.1. Phenological evolution of LAI, indexes and soil respiration

During the vegetative stages, statistical results (Table 4) showed that all indexes explained a high proportion of LAI’s variance assuming a linear regression model (between the LAI and the spectral indexes). The Radj2 were higher in the P sites than in the I sites probably due to the null relationship between the LAI and the spectral indexes during the first stages of corn growth when LAI was zero in the I sites. During the reproductive stages the indexes explained a high proportion of the LAI variability only in the I sites while moderate to low Radj2 values were found in the P sites. This corroborates the presence of functional variability in P sites captured by the spectral indexes but not yet by LAI which has just a response to structural changes, shown in the I sites.

The phenological evolution of the irrigated maize crop showed that the LAI and the spectral indexes increased simultaneously (ANIR showed an inverse pattern) during the vegetative period until the indexes saturate at high LAI values. In the P sites this increase is observed from VE, whereas in the I sites there is a delay and it is not observed until V6. During the first reproductive stages (Table 1: from flowering (VT) to blister stage (R2)), LAI and indexes remained constant. After R2 the indexes decreased in both sites while in LAI this decrease was only observed in the I sites. Similar dynamics in the indexes have been found by other researchers during the reproductive stages (Gitelson et al., 2003a,b; Viña et al., 2004). In fact, Huang et al. (2012) showed that NDVI, EVI and Clrededge decreased after VT during the reproductive stages until R5 (dent stage), although the decrease in NDVI was not as clear as in the other two indexes. The structural change observed (decrease of LAI) in the I sites was probably related with the loss of leaf turgidity which was more clear between rows because the leaf senescence process starts at the tip of the leaf and enlarges progressively toward the leaf base (Feller et al., 1977). Our results indicate that spectral indexes are able to capture functional information related to the senescence process in areas where no structural change is yet captured by LAI. Total soil respiration always showed higher values in the cultivated plots than in the bare soil plots in which Rs kept with low constant values throughout the study period. In addition, Rs is higher in the P than in the I sites. These facts show the significant contribution of autotrophic respiration under the plants with higher root density. The same effect has been shown for maize and other row crops in several studies (Amos et al., 2005; Han et al., 2007; Prolingheuer et al., 2010; Rochette et al., 1991). Rs increased during the first stages of corn growth as a response to the formation of the root system together with the effect of the rabbit manure and the C input being made available to the heterotrophs. As a matter of fact, in the 2012 experiment, Rs values were higher in the P sites than in the I and S sites even before the VE stage (i.e., before the coleoptile emerges from the soil surface) showing probably the effect of the seminal root system development. During this period in the I sites, where roots were not present yet, Rs values were similar to those of bare soil showing a delay with respect to the P sites. Amos et al. (2005) also found differences in Rs between P and I sites in the early maize phenological stages. After VE, in both the P and the I sites, Rs increased probably due to the development of the nodal root system linked to leaf growth and expansion. The Rs started decreasing steadily around three weeks before the maximum LAI, keeping this trend during the flowering and the reproductive stages. Huang et al. (2012), who found similar results, explained it as a result of the decrease in temperature and senescence of maize. Our results also showed that the Rs started decreasing around two weeks after the soil temperature reached its maximum and began to decrease. Probably, both the soil temperature and the Rs were responding to the shading effect of the crop when the LAI values were high enough. Leaf senescence has also an effect on the Rs due to the decrease of photosynthesis rate and chlorophyll content which is the main factor affecting the green vegetation fraction (Viña et al., 2004). In addition, during this period the assimilates go directly to the grain filling processes, thus decreasing the C translocation to the soil (Amos et al., 2005; Martens, 1990; Qian et al., 1997). At the end of the growing period, the Rs was equal to the one measured in the soil accounting just for the heterotrophic respiration (Prolingheuer et al., 2010).

4.3. Relationship between soil respiration and spectral indexes Due to the phenological dynamics of the crop it was decided to evaluate the capability of the spectral indexes to assess on one side total soil respiration and on the other autotrophic respiration, estimated as the difference between total soil respiration and the heterotrophic (basal) respiration which was assumed to be the respiration of the bare soil. Results showed clear different patterns between vegetative and reproductive stages in the two study cases. During the vegetative stages, in the P sites (Tables 5 and 6) spectral indexes explained a moderate proportion of Rs variance (40–50%) and a high proportion Ra variance (around 70%). This indicates that during the vegetative growth a significant part of Rs was autotrophic root respiration which is directly related to biomass accumulation and to crop structural changes (i.e., leaf and root development) (Amos et al., 2005; Han et al., 2007; Kuzyakov and Cheng, 2004). This is corroborated by the high relationship between indexes and LAI and between Rs and Ra with LAI. In the I sites, neither indexes nor LAI alone explained any Rs variance, while they explained around 30% of the Ra variance. It is expected that around the emergence Rs is primarily due to heterotrophic activity (probably more related to abiotic factors) and Ra contribute to variability mainly during the last vegetative stages as roots are invading inter-row spaces resulting in moderate Radj2 values. The fact that results are better for Ra than for Rs indicates that the capability of spectral information to capture vegetation dynamics provides useful tools to assess the evolution of autotrophic respiration during this period. The information provided by LAI measurements indicated the significant weight of structural changes in the evolution of Ra. In this line Prolingheuer et al. (2010) concluded that Rs seasonal variability in winter wheat basically originates from the variability of the autotrophic component. Although other studies have found differences between spectral indexes (Huang et al., 2012), in this case, a similar amount of variability was explained by the four indexes tested indicating that probably the variability was due to phenological evolution that can be captured by the general patterns of the reflectance spectra. During the reproductive stages, Radj2 values between the Rs and the LAI were irrelevant while when considering only Ra they increased. It seems that by removing the heterotrophic component it was possible to reveal the relationship between structural changes and Ra during the reproductive period, especially in the I sites as it has been already highlighted. In fact, the structural changes must be so distinct that Ra was better related to LAI than to vegetation indexes. Except in this case, a higher proportion of soil respiration (Ra and Rs) variability was better explained by spectral indexes than by the LAI, indicating that they contain significant

166

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167

information related to plant functioning, beyond mere structural changes. The Radj2 values showed that during the reproductive stages vegetation indexes were representing better total soil respiration than the autotrophic component. A hypothesis that could explain this effect is that the assumption that the heterotrophic respiration in the crop is equivalent to bare soil respiration is not always correct through the growing period. It would be expected that the proportion of heterotrophic respiration evolves during the growing period due for example to the presence of additional soil organic matter and root exudates in the rizhosphere which could result in a proportion of rhizomicrobial respiration (Kuzyakov, 2006), frequently considered as autotrophic respiration (Högberg and Högberg, 2002; Moyano et al., 2007), and linked to plant activity and hence vegetation indexes. 4.4. Significance of abiotic factors In the bare soil plot 77% of the Rs variability was explained by moisture as it was expected, the same effect has been reflected by authors (Davidson et al., 1998; Herbst et al., 2009; Prolingheuer et al., 2010). On the other hand no relationship with temperature was found as in other studies (Prolingheuer et al., 2010). In irrigated crops grown in Mediterranean conditions in summer temperature is not a limiting factor for the heterotrophic populations (Oyonarte et al., 2012), not having probably an effect on Rs variability. During the vegetative stages for Rs and the whole growing period for Ra neither moisture nor temperature explained any variability. This could indicate that in cases when Ra represents a high proportion of total soil respiration, the response to abiotic factors is masked by the variability due to plant activity during plant growth. This is in agreement with the fact that during these stages a high proportion of Ra variability was already accounted for by the indexes. During the reproductive stages, moisture explained 42% of the Rs variability in the P sites while in the I sites was irrelevant. This may be explained by an increase of the proportion of heterotrophic respiration associated to the rhizosphere, denser under the plant and which is also dependent on abiotic factors. This is in agreement with the fact that during these stages total respiration was better explained by the indexes than Ra as it was explained previously. This idea could be supported also by the fact that when including moisture into the indexes models the amount of variability explained increased slightly only for Rs in the P sites. While in the vegetative stages temperature by itself did not show any explanatory power, when included in the models it increase approximately 15% of the Rs and around 30% of the Ra variability. This indicates that the influence of temperature is maybe linked to vegetation activity, which is supported by the fact that the variable is irrelevant when explaining bare soil respiration. 5. Conclusion In this work the relationship between total soil respiration and autotrophic respiration with vegetation indexes and abiotic factors was assessed in a corn crop through the growing period. Leaf area index was used as an indicator of structural changes in the crop spatially and seasonally. Significant differences in the relationships between Rs and Ra with spectral indexes in the vegetative and reproductive stages and between measurements under the plants and between the rows were found. Best results were obtained when assessing autotrophic respiration during period with significant structural changes during vegetative stages. However, during the reproductive stages,

when the autotrophic component was not so significant, spectral indexes were better related to total soil respiration which could be related to the presence of rhizomicrobial respiration linked to vegetation activity. This showed that spectral indexes contain significant functional information, beyond mere structural changes, that could be related to carbon fluxes. No clear patterns were found among the different spectral indexes tested which probably indicate that the indexes were tracking phenological evolution that is captured by the general patterns of the reflectance spectra. The results obtained confirm that in irrigated crop systems remote sensing data can produce relevant information to assess the both autotrophic and the total soil respiration through vegetation indexes. In addition this work provides relevant information to the remote sensing scientific community for the development of methodologies to assess the evolution of carbon fluxes in agroecosystems. Specifically the following should be taken into account:  The high variability found in the relationship between soil respiration and spectral indexes within the corn field suggests that there is a need to be cautious when upscaling remote sensing models.  Furthermore, the variability observed in the results through the phenological cycle suggests that in order to assess the Rs and Ra from remote sensors, specific models should be applied for the different phenological stages.

Acknowledgments This research was conducted in the framework of the Spanish National Projects AGL-2010-17505 and the CGL2009-07031. The authors would like to express their gratitude to the College of Agricultural Engineering of the Technical University of Madrid for allowing them to carry out this research in their experimental fields. They would like to acknowledge all the staff of these fields, especially Roman Zurita. Also they would like to express their gratitude to David Manrique from CIEMAT for his collaboration in taking respiration measurements. References AEMET—Agencia Estatal de Meteorología/IMP—Instituto de Meteorología de Portugal, 2011. Atlas Climático Ibérico/Iberian Climate Atlas. Amos, B., Arkebauer, T.J., Doran, J.W., 2005. Soil surface fluxes of greenhouse gases in an irrigated maize-based agroecosystem. Soil Sci. Soc. Am. J. 69, 387–395. Baret, F., Guyot, G., 1991. Potentials and limits of vegetation indices for LAI and APAR assessment. Remote Sens. Environ. 35, 161–173. Buchmann, N., 2000. Biotic and abiotic factors controlling soil respiration rates in Picea abies stands. Soil Biol. Biochem. 32, 1625–1635. Buschmann, C., Nagel, E., 1993. In vivo spectroscopy and internal optics of leaves as basis for remote sensing of vegetation. J. Int. Remote Sens. 14, 711–722. Cramer, W., Bondeau, A., Woodward, F.I., Prentice, I.C., Betts, R.a., Brovkin, V., Cox, P. M., Fisher, V., Foley, J.a., Friend, A.D., Kucharik, C., Lomas, M.R., Ramankutty, N., Sitch, S., Smith, B., White, A., Young-Molling, C., 2001. Global response of terrestrial ecosystem structure and function to CO2 and climate change: results from six dynamic global vegetation models. Global Change Biol. 7, 357–373. Davidson, E., Belk, E., Boone, R.D., 1998. Soil water content and temperature as independent or confounded factors controlling soil respiration in a temperate mixed hardwood forest. Global Change Biol. 4 (2), 217–227. Dixon, R.K., Solomon, A.M., Brown, S., Houghton, R.A., Trexier, M.C., Wisniewski, J., 1994. Carbon pools and flux of global forest ecosystems. Science 263, 185–190. Dong, J., Kaufmann, R.K., Myneni, R.B., Tucker, C.J., Kauppi, P.E., Liski, J., Buermann, W., Alexeyev, V., Hughes, M.K., 2003. Remote sensing estimates of boreal and temperate forest woody biomass: carbon pools sources, and sinks. Remote Sens. Environ. 84, 393–410. FAO (Agriculture Organization of the United Nations), 2002. El Estado Mundial De La Agricultura y La Alimentación. FAO, Roma. Feller, U.K., Soong, T.S.T., Hageman, R.H., 1977. Leaf proteolytic activities and senescence during grain development of field-grown corn (Zea mays L.). Plant Physiol. 59 (2), 290–294.

V. Cicuéndez et al. / Agriculture, Ecosystems and Environment 212 (2015) 158–167 Gao, B., 1996. NDWI—a normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Environ. 58, 257–266. Gitelson, A.A., Kaufman, Y.J., Stark, R., Rundquist, D., 2002. Novel algorithms for remote estimation of vegetation fraction. Remote Sens. Environ. 80, 76–87. Gitelson, A.A., Verma, S.B., Viña, A., Rundquist, D.C., Keydan, G., Leavitt, B., Arkebauer, T.J., Burba, G.G., Suyker, A.E., 2003a. Novel technique for remote estimation of CO2 flux in maize. Geophys. Res. Lett. 30, 1486. Gitelson, A.A., Viña, A., Arkebauer, T.J., Rundquist, D.C., Keydan, G., Leavitt, B., 2003b. Remote estimation of leaf area index and green leaf biomass in maize canopies. Geophys. Res. Lett. 30, 1248. Gitelson, A.A., Viña, A., Verma, S.B., Rundquist, D.C., Arkebauer, T.J., Keydan, G., Leavitt, B., Ciganda, V., Burba, G.G., Suyker, A.E., 2006. Relationship between gross primary production and chlorophyll content in crops: Implications for the synoptic monitoring of vegetation productivity. J. Geophys. Res.: Atmos. 111, D08S11 (1984–2012). Goulden, M.L., Munger, J.W., Fan, S., Daube, B.C., Wofsy, S.C., 1996. Exchange of carbon dioxide by a deciduous forest: response to interannual climate variability. Science 271, 1576–1578. Goulden, M.L., Wofsy, S.C., Harden, J.W., Trumbore, S.E., Crill, P.M., Gower, S.T., Fries, T., Daube, B.C., Fan, S., Sutton, D.J., Bazzaz, A., Munger, J.W., 1998. Sensitivity of boreal forest carbon balance to soil thaw. Science 279, 214–217. Han, G., Zhou, G., Xu, Z., Yang, Y., Liu, J., Shi, K., 2007. Biotic and abiotic factors controlling the spatial and temporal variation of soil respiration in an agricultural ecosystem. Soil Biol. Biochem. 39, 418–425. Hanway, J.J., 1971. How a Corn Plant Develops, Special Report No. 48. Iowa State University Cooperative Extension Service, Ames, Iowa. Heinsch, F.A., Zhao, M., Running, S.W., Kimball, J.S., Nemani, R.R., Davis, K.J., Bolstad, P.V., Cook, B.D., Desai, A.R., Ricciuto, D.M., Law, B.E., Oechel, W.C., Kwon, H., Luo, H., Wofsy, S.C., Dunn, A.L., Munger, J.W., Baldocchi, D.D., Xu, L., Hollinger, D.Y., Richardson, A.D., Stoy, P.C., Siqueira, M.B.S., Monson, R.K., Burns, S.P., Flanagan, L. B., 2006. Evaluation of remote sensing based terrestrial productivity from MODIS using regional tower eddy flux network observations. IEEE Trans. Geosci. Remote Sens. 44, 1908–1925. Herbst, M., Prolingheuer, N., Graf, A., Huisman, J.A., Weihermüller, L., Vanderborght, J., 2009. Characterization and understanding of bare soil respiration spatial variability at plot scale. Vadose Zone J. 8 (3), 762–771. Högberg, M.N., Högberg, P., 2002. Extramatrical ectomycorrhizal mycelium contributes one-third of microbial biomass and produces, together with associated roots, half the dissolved organic carbon in a forest soil. New Phytol. 154, 791–795. Huang, N., Niu, Z., 2013. Estimating soil respiration using spectral vegetation indices and abiotic factors in irrigated and rainfed agroecosystems. Plant Soil 367, 535– 550. Huang, N., Niu, Z., Zhan, Y., Xu, S., Tappert, M.C., Wu, C., Huang, W., Gao, S., Hou, X., Cai, D., 2012. Relationships between soil respiration and photosynthesis-related spectral vegetation indices in two cropland ecosystems. Agric. For. Meteorol. 160, 80–89. Huete, A., Didan, K., Miura, T., Rodriguez, E.P., Gao, X., Ferreira, L.G., 2002. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sens. Environ. 83, 195–213. Khanna, S., Palacios-Orueta, A., Whiting, M.L., Ustin, S.L., Riaño, D., Litago, J., 2007. Development of angle indexes for soil moisture estimation, dry matter detection and land-cover discrimination. Remote Sens. Environ. 109, 154–165. Khanna, S., Santos, M.J., Ustin, S.L., Koltunov, A., Kokaly, R.F., Roberts, D.A., 2013. Detection of salt marsh vegetation stress and recovery after the deepwater horizon oil spill in Barataria Bay, Gulf of Mexico using AVIRIS data. PloS One 8, e78989. Knapp, A.K., Smith, M.D., 2001. Variation among biomes in temporal dynamics of aboveground primary production. Science 291, 481–484. Kuzyakov, Y., 2006. Sources of CO2 efflux from soil and review of partitioning methods. Soil Biol. Biochem. 38, 425–448. Kuzyakov, Y., Cheng, W., 2004. Photosynthesis controls of CO2 efflux from maize rhizosphere. Plant Soil 263, 85–99. Kuzyakov, Y., Cheng, W., 2001. Photosynthesis controls of rhizosphere respiration and organic matter decomposition. Soil Biol. Biochem. 33, 1915–1925. Lal, R., 2011. Sequestering carbon in soils of agro-ecosystems. Food Policy 36, S33– S39. Martens, R., 1990. Contribution of rhizodeposits to the maintenance and growth of soil microbial biomass. Soil Biol. Biochem. 22, 141–147. Matson, P.A., Parton, W.J., Power, A.G., Swift, M.J., 1997. Agricultural intensification and ecosystem properties. Science 277, 504–509. Mosier, A., Halvorson, A., Peterson, G., Robertson, G., Sherrod, L., 2005. Measurement of net global warming potential in three agroecosystems. Nutr. Cycling Agroecosyst. 72, 67–76. Moyano, F.E., Kutsch, W.L., Schulze, E.D., 2007. Response of mycorrhizal, rhizosphere and soil basal respiration to temperature and photosynthesis in a barley field. Soil Biol. Biochem. 39 (4), 843–853.

167

Oyonarte, C., Rey, A., Raimundo, J., Miralles, I., Escribano, P., 2012. The use of soil respiration as an ecological indicator in arid ecosystems of the SE of Spain: spatial variability and controlling factors. Ecol. Indic. 14 (1), 40–49. Pachauri, R.K., Reisinger, A., 2007. Cambio climático 2007: Informe de síntesis. Contribución de los Grupos de Trabajo I, II y III al Cuarto Informe de Evaluación del Grupo Intergubernamental de Expertos sobre el Cambio Climático. IPCC, Ginebra, Suiza 104 págs. Palacios-Orueta, A., Huesca, M., Whiting, M.L., Litago, J., Khanna, S., Garcia, M., Ustin, S.L., 2012. Derivation of phenological metrics by function fitting to time-series of Spectral Shape Indexes AS1 and AS2: mapping cotton phenological stages using MODIS time series. Remote Sens. Environ. 126, 148–159. Palacios-Orueta, A., Khanna, S., Litago, J., Whiting, M.L., Ustin, S.L., 2006. Assessment of NDVI and NDWI spectral indices using MODIS time series analysis and development of a new spectral index based on MODIS shortwave infrared bands. 1st International Conference on Remote Sensing and Geoinformation Processing in the Assessment and Monitoring of Land Degradation and Desertification, Trier, Germany, pp. 207–209. Paustian, K., Andrén, O., Janzen, H.H., Lal, R., Smith, P., Tian, G., Tiessen, H., Noordwijk, M.V., Woomer, P.L., 1997. Agricultural soils as a sink to mitigate CO2 emissions. Soil Use Manage. 13, 230–244. Prolingheuer, N., Scharnagl, B., Graf, A., Vereecken, H., Herbst, M., 2010. Spatial and seasonal variability of heterotrophic and autotrophic soil respiration in a winter wheat stand. Biogeosci. Discuss. 7, 9137–9173. Qian, J.H., Doran, J.W., Walters, D.T., 1997. Maize plant contributions to root zone available carbon and microbial transformations of nitrogen. Soil Biol. Biochem. 29, 1451–1462. Raich, J., Schlesinger, W.H., 1992. The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus B 44, 81–99. Rochette, P., Desjardins, R., Pattey, E., 1991. Spatial and temporal variability of soil respiration in agricultural fields. Can. J. Soil Sci. 71, 189–196. Sanchez-Martín, L., Meijide, A., Garcia-Torres, L., Vallejo, A., 2010. Combination of drip irrigation and organic fertilizer for mitigating emissions of nitrogen oxides in semiarid climate. Agric. Ecosyst. Environ. 137, 99–107. Sauerbeck, D., 2001. CO2 emissions and C sequestration by agriculture–perspectives and limitations. Nutr. Cycling Agroecosyst. 60, 253–266. Schlesinger, W.H., Andrews, J.A., 2000. Soil respiration and the global carbon cycle. Biogeochemistry 48, 7–20. Sims, D.A., Rahman, A.F., Cordova, V.D., El-Masri, B.Z., Baldocchi, D.D., Flanagan, L.B., Goldstein, A.H., Hollinger, D.Y., Misson, L., Monson, R.K., Oechel, W.C., Schmid, H. P., Wofsy, S.C., Xu, L., 2006. On the use of MODIS EVI to assess gross primary productivity of North American ecosystems. J. Geophys. Res. 111, G04015 (2005– 2012). Snyder, C., Bruulsema, T., Jensen, T., Fixen, P., 2009. Review of greenhouse gas emissions from crop production systems and fertilizer management effects. Agric. Ecosyst. Environ. 133, 247–266. Tilman, D., Cassman, K.G., Matson, P.A., Naylor, R., Polasky, S., 2002. Agricultural sustainability and intensive production practices. Nature 418, 671–677. Tucker, C.J., 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens. Environ. 8, 127–150. Tucker, C.J., Slayback, D.A., Pinzon, J.E., Los, S.O., Myneni, R.B., Taylor, M.G., 2001. Higher northern latitude normalized difference vegetation index and growing season trends from 1982 to 1999. Int. J. Biometeorol. 45, 184–190. Turner, D.P., Ollinger, S.V., Kimball, J.S., 2004. Integrating remote sensing and ecosystem process models for landscape-to regional-scale analysis of the carbon cycle. Bioscience 54, 573–584. Uribe, C., Inclán, R., Sánchez, D., Clavero, M., Fernández, A., Morante, R., Cardeña, A., Blanco, A., Van Miegroet, H., 2013. Effect of wildfires on soil respiration in three typical Mediterranean forest ecosystems in Madrid, Spain. Plant Soil 369, 403– 420. Viña, A., Gitelson, A.A., 2005. New developments in the remote estimation of the fraction of absorbed photosynthetically active radiation in crops. Geophys. Res. Lett. 32, L17403. Viña, A., Gitelson, A.A., Nguy-Robertson, A.L., Peng, Y., 2011. Comparison of different vegetation indices for the remote assessment of green leaf area index of crops. Remote Sens. Environ. 115, 3468–3478. Viña, A., Gitelson, A.A., Rundquist, D.C., Keydan, G., Leavitt, B., Schepers, J., 2004. Monitoring maize (L.) phenology with remote sensing. Agron. J. 96, 1139–1147. Wang, Q., Adiku, S., Tenhunen, J., Granier, A., 2005. On the relationship of NDVI with leaf area index in a deciduous forest site. Remote Sens. Environ. 94, 244–255. Xu, X., Kuzyakov, Y., Wanek, W., Richter, A., 2008. Root-derived respiration and nonstructural carbon of rice seedlings. Eur. J. Soil Biol. 44, 22–29. Zhang, M., Wu, B., Meng, J., 2014. Quantifying winter wheat residue biomass with a spectral angle index derived from China Environmental Satellite data. Int. J. Appl. Earth Obs. Geoinf. 32, 105–113. Zhang, X., Friedl, M.A., Schaaf, C.B., Strahler, A.H., Hodges, J.C., Gao, F., Reed, B.C., Huete, A., 2003. Monitoring vegetation phenology using MODIS. Remote Sens. Environ. 84, 471–475.