Modeling Temperature Effects on Decomposition Summary Objectives

14 downloads 0 Views 1MB Size Report
where Ct (kg/ha)is the modeled mass of cumulatively evolved carbon at time t, ... at time t (not cumulative), Φ is the set of replications under consideration for the.

2001-2006 Mission Kearney Foundation of Soil Science: Soil Carbon and California's Terrestrial Ecosystems Final Report: 2002009, 1/1/2003-12/31/2004

Modeling Temperature Effects on Decomposition David M. Crohn1

Summary In irrigated landscapes and agricultural systems, temperature is the most significant environmental factor affecting carbon decomposition. Climatic variation therefore makes it difficult to apply decay rates determined in one environment to another. Decay rates are critically important, however, as they will determine the efficiency of carbon sequestration efforts, the useful life of mulches and composts, as well as the rate at which nutrients are released from organic fertilizers. This project explored an approach modeling decay by using the Arrhenius equation to adjust time. The resulting decay constants are independent of temperature and can therefore be applied to similar materials in other climates. Dissimilar materials can be considered by predicting temperature-independent decay constants from proximate carbon analysis assays. An incubation experiment was completed including seven different soil amendments decomposing under four different climatic conditions. Carbon dioxide evolution was measured within the incubation vessels, and samples have been collected for total carbon analysis. Results suggest that the approach has practical promise, but that the Arrhenius relationship itself did not adequately represent the effects of heat on decomposition. Decomposition rates increased steadily as temperature rose from 25°C to 40°C and fell dramatically between 25°C and 5°C, but the pattern did fit the Arrhenius pattern well. Overall, a Q10 of 2 was reasonable, but at lower temperatures Q10 values appears to be between 3 and 4.

Objectives 1. To evaluate methods for optimizing fitting procedures for determining temperature independent decay parameters from field and laboratory data using temperature adjusted time. 2. To begin the process of collecting decomposition data on California litter materials as well as proximate carbon analysis data. To develop temperature-independent decay parameter values for the tested materials and to test a model for predicting decay model parameters from proximate carbon analysis. To make available frozen samples of these data to other researchers interested in developing instrumentation alternatives for predicting temperature-independent decay parameters. 3. To use web-based methods for extending the temperature-adjusted-time and Arrhenius approach to California clientele with an eye toward extending the approach toward nitrogen management. To date, Objective 2, the incubation experiment has concluded. All samples have been prepared for analysis but additional sample measurements will be necessary before a full analysis of results will be possible. Objective 2 was the principal goal of the study. Objective 1 was a small part of the overall project. Results suggest that parameterization results indeed sensitive to the length of a study. The objective function used in Objective 2, which normalizes the sum-ofsquare-errors was derived from this work. Work on Objective 3, the web site, has commenced 1

Department of Environmental Sciences, University of California, Riverside

Modeling Temperature Effects on Decomposition—Crohn

using data from the CIMIS system but it remains premature to promote the approach based upon the data generated during this study.

Approach and Procedures We studied the influence of temperature on the decomposition of seven organic amendments commonly used in California agriculture, including old (MO) and fresh (MF) dairy manure, dried poultry manure (MP), anaerobically digested biosolids (BA), a greenwaste compost (CG), a greenwaste compost amended with biosolids (BC), and a greenwaste compost amended with dairy manure (CC).

Laboratory Procedures Each material was mixed into a 50/50 blend of sandy loam soil and quartz sand. The amendments were introduced to compose 2 percent of the total amendment/soil/sand dry mass. A total of sixteen 750 g (dry weight) replicates were prepared for each amendment and placed into 2 L canning jars fitted with airtight septa. Water was added to each amended mixture to bring it to field capacity as determined with pressure plates. To study the influence of temperature on decomposition, four replicates of each mixture was decomposed under four different conditions including cool (~6.4°C) treatment, moderate (~20°C), and warm (~41°C) conditions as well as a treatment that varies diurnally under greenhouse conditions. The headspaces above the soils were sampled through the septa with a 20 mL syringe and the evolved CO2 is then measured with a PP Systems EGM4 infrared gas chromatograph. The experiment began on Aug. 1, 2003, and continued until Jan. 26, 2004.

Modeling The traditional approach is to use the Arrhenius equation to increase decay rates under warm conditions and to reduce them under cool conditions. Decay rates must be estimated repeatedly each time temperatures change. Our approach is similar except, instead of altering the decay rates, we used a constant decay rate for all conditions that modifies time such that time expands under warm conditions and contracts under cool conditions. This approach makes parameterization and use of first-order models much more convenient as it permits the development of decay parameters suitable for all climates. For this procedure, temperatureadjusted time is first determined as t° =

! Q10

) Tr & )' Tr ' 1+ $ ' 1* ( 10 % ( Tt

& $$ % #t

(1)

#t "t

where Tr (K) is a reference temperature, 298.15 K, Tt (K) is the soil temperature at time t, Δt (days) is the time step, and Q10 is the relative change in the decay rates expected after a 10ºC (18ºF) increase in temperature from the reference temperature. The temperature-adjusted time series were then used to model decomposition. The model assumes that each added amendment contains two compartments, one labile, the other recalcitrant. This can be written as

2

Modeling Temperature Effects on Decomposition—Crohn

[

C t = M o 1 ! Ae ! k Rt ° ! (1 ! A)e ! k Lt °

]

(2)

where Ct (kg/ha) is the modeled mass of cumulatively evolved carbon at time t, Mo (kg/ha) is the initial measured carbon mass in the added amendment, A is the fraction of the initial carbon mass which is recalcitrant, kL and kR (day-1) are constants independent of temperature, and tº (days) is temperature-adjusted time. Two approaches were taken toward fitting parameter values to the model. Both involved minimizing the normalized sum-of-squared error (NSSE) between observations and model predictions. NSSE was determined as 2

( C tˆ ) M tˆ % & # !! & M # " tˆ ' ˆt $ NSSE = (3) R where Mt is the measured evolved carbon during time interval tˆ , C tˆ is the modeled evolved carbon at time t (not cumulative), Φ is the set of replications under consideration for the parameterization, and R is the total number of replications in this set. An independent approach (IA) calculated separate Q10 values for each amendment. A common approach (CA) determined shared Q10 values for all amendments. In addition, four set of environmental conditions (EC) were considered for each approach. These were EC1: (CR, LA), EC2: (LA, IN), EC3: (CR, LA, IN), EC4: (CR, LA, IN, GH). Including both IA and CA for each EC resulted in a total of eight parameterization alternatives.

Results Laboratory Procedures Figure 1 presents temperature data collected in each of the four temperature environments included in this study. Incubator (IN) temperatures remained fairly consistent at about 41°C. Periodic drops in IN temperatures reflect periods when samples were removed for sampling as temperature transducers remained with the samples at all times. Cold room (CR) temperatures fluctuated around 5°C. Again, periodic spikes in CR temperatures represent sampling events. Laboratory temperatures varied somewhat due to heating and cooling construction activity in the building until Sept. 23, when temperatures stabilized around 20°C. Prior to Sept. 23, temperatures slowly drifted between 15˚C and 23˚C. GH indicates a greenhouse environment. Temperatures within the greenhouse fluctuate diurnally, particularly during the summer when the sun is most intense. During the first few days of the study, temperatures exceeded 70°C, near the upper limit tolerated by thermophilic soil organisms (Haug 1993). In general, greenhouse temperatures varied from 20°C to 26°C in the morning to between 40°C and 72°C in the evening during the first five weeks of the experiment. The very elevated temperatures during the afternoon were not desirable because they extended well above the 42°C temperature of the incubated samples. Samples experiencing the elevated greenhouse temperatures also tended to dry out and required the addition of moisture during the

3

Modeling Temperature Effects on Decomposition—Crohn

second, sixth, twelfth, and sixteenth weeks of the study. None of the other temperature environments required the addition of moisture.

Figure 1. Temperatures measured for the incubator (IN), laboratory (LA), cold room (CR), and greenhouse (GH) treatments.

Figure 2. Daily incubator (IN) treatment CO2 evolution.

Figures 2, 3, 4, and 5 represent the concentrations of CO2, above the control concentration, as it accumulated daily over the course of the experiment. Figure 2 shows the CO2 generated within the incubator (IN). Initially, IN generated between 61,100 ppm/day for the anaerobically digested biosolids (BA) to 6,840 ppm/day for the manure/greenwaste compost (CC) and 1,880 ppm/day for the control (ST). BA remained the most labile material throughout the experiment 4

Modeling Temperature Effects on Decomposition—Crohn

and CC remained the most recalcitrant. After four weeks of decomposition, the most labile fractions of these materials were exhausted and BA was releasing 7,330 ppm/day, an 88 percent reduction from the initial CO2 generation rate. CC was releasing 2,260 ppm/day, a 67 percent reduction. In decreasing order of CO2 generation were BA > MP > MF ≈ CG > MO ≈ BC > CC > ST, although MF was initially greater than CG and MO was initially greater than BC. Figure 3 reports results from the cold room (CR). In the CR, activity fell off very rapidly during the first three days, a factor likely to the suppression of microbial growth and development under cool conditions. Once again, BA was the most labile material releasing 33,700 ppm/day initially. CC was the most recalcitrant amendment releasing 920 ppm/day. These values were 55 percent and 13 percent of the corresponding incubated (IN) values. Excluding the anaerobically digested biosolids, initial activity fell off by an average 22.5±15.2 percent (mean ± one standard error). By day three, activity in the cold room was much depressed having fallen to 2.5±2.1 percent of the IN values. Activity declined steadily during the first nine days of the experiment and then slowly increased during the following four after which CO2 release steadily declined. The decrease and subsequent increase in CO2 release rates may have been due to an adjustment in the microbial population to organisms favoring the low temperatures in the CR. As the population became more cryophilic, its capacity to decompose organic material increased until the beginning of week three of the experiment, after which activity slowly declined. After a month of decay, CR CO2 evolutions rates were 20.0±9.6 percent as large as the IN rates. As in the IN, different materials displayed different level of decomposability. During the first two days of the experiment, a time of shifting microbial populations, BA >> MP ≈ MF > CG ≈ MO > BC > CC > ST. Later, BA > MP ≈ MF ≈ CG ≈ MO > BC > CC > ST, and MP, MF, CG, and MO shifted among each other in their prominence.

Figure 3. Daily cold room (CR) treatment CO2 evolution.

Figure 4 presents results for the laboratory (LA). Between Aug. 4 and Sept.23, temperature conditions within the laboratory varied somewhat, but the 8˚C range of temperatures was 5

Modeling Temperature Effects on Decomposition—Crohn

probably not enough to significantly alter the population of the microbes in the samples. In the LA, initially BA > MP > MF > CG > MO > BC > CC > ST. Within two weeks, MO and BC converged and within a month MF, MO, CG, and BC were all generating between 3,500 and 4,500 ppm/day CO2 while BA generated 7,680 and CC generated 390 ppm/day, respectively. BA was releasing 5 percent more CO2 at that time in the laboratory than in the incubator. LA CC, by contrast, was only 20 percent of IN CC. Because CC, which was composted and cured, contains a much greater percentage of recalcitrant carbon than BA, which is quite labile, this may suggest that the decay of recalcitrant materials is accelerated more than is the decay of labile fractions (Crohn and Valenzuela 2003). Note, however, that after one month, BC, a biosolids co-compost, also generated two percent more CO2 in the LA than in IN. Presumably, BA and BC, both biosolids products, contained labile materials that had decayed after a month in the IN, but that additional labile carbon remained in the lower temperature incubations. Apparently this was true even though BC was composted. For all the amendments, the ratio of LA:IN evolution is 77.0±24.9 percent after one month of decomposition.

Figure 4. Daily laboratory (LA) treatment CO2 evolution.

Figure 5 presents data from the greenhouse (GH). GH treatments were exposed to elevated diurnally varying temperatures (fig. 1). Drying was sufficient to slow decomposition on two occasions (Aug. 15 and 25) but microbial activity quickly recovered after the reintroduction of water. Initial decay rates for all of the materials were approximately the same as for IN; 101±14.3 percent of IN values. By the next day, however, ratios had fallen to 68.5±22.1 percent. By Aug. 15, decomposition had declined to 26.2±17.9 percent of IN, but after the addition of water, decomposition recovered to 90.5±62.7 percent of IN. GH temperatures averaged 35˚C during this period, somewhat less than IN. Temperatures were often higher than GH, however, probably accelerating decay above IN during warm afternoons periods.

6

Modeling Temperature Effects on Decomposition—Crohn

Figure 5. Daily greenhouse (GH) treatment CO2 evolution.

Figure 6. C evolution as a percentage of the initial added amendment: Incubator (IN).

7

Modeling Temperature Effects on Decomposition—Crohn

Figure 7. C evolution as a percentage of the initial added amendment: Cold room (CR).

It should be kept in mind that comparisons between different temperature environments are in terms of daily CO2 generation and are therefore memoryless. As time passes, comparisons between warmer and cooler treatments increasingly compare the decomposition of more recalcitrant materials in the warmer treatments to the decomposition of more labile materials in the cooler treatment.

Figure 8. C evolution as a percentage of the initial added amendment: Laboratory (LA).

8

Modeling Temperature Effects on Decomposition—Crohn

Figure 9. C evolution as a percentage of the initial added amendment: Greenhouse (GH).

An alternative mode of expression would be to report CO2 generation in cumulative terms. Figures 6 through 9 present results as the total C evolved as a percentage of the C in added amendments. Values are corrected for the contributions of the soil, which appears in the figures as a standard (ST). IN treatments decompose much more rapidly than LA and CR treatments exhausting labile materials more quickly. IN, LA, and GH treatments tend to converge as labile materials are exhausted. LA release is substantially greater than CR rates, significantly exceeding the difference between LA and IN rates. Total C evolution is only a small fraction of the amount added, however, ranging from 0.1% to 3.1% in the CR, 0.7% to 8.7% in the LA, 2.9 to 10.6% in the IN and 0.8 to 7.7% in the GH. In all cases, CC was the most recalcitrant compound and BA was the most labile. The order at which the materials expressed their relative evolution was remarkably consistent for all of the environments.

Model Testing Eight parameterization alternatives were included in this analysis. IA Q10 estimates were significantly greater for EC1 than for EC2. When Q10 was estimated separately for each amendment (table 1), values ranged from 3.0 (CC) to 4.1 (MO) when only CR and LA were considered. Labile materials showed Q10 in the vicinity of 4 while the two of the three composts were lower, near 3 (BC and CC). These two composts were the most recalcitrant of all of the tested amendments, suggesting that highly stabilized materials may express less temperature sensitivity than more decomposable materials. EC2 Q10 values ranged from 1.42 for BA to 2.11 for CC. No pattern emerged between the materials and the Q10 values, however, as the Q10 for BC was also low, at 1.52. EC3 values ranged between 2.0 and 2.4 while EC4 Q10’s were between 1.9 and 2.1 except for CC, which was 2.7. Other parameter estimates also appear in Table 1. Values for A range from 0.89 to 0.99, reflecting the relatively slow decomposition rates of the materials in the experimental 9

Modeling Temperature Effects on Decomposition—Crohn

environment. Recalcitrant fraction decomposition rates ranged from 4.9⋅107 to 3.13⋅104 d-1. Labile fraction values are between 6.7⋅104 and 0.10 d-1. It is difficult to comment on patterns between these values because they co-vary significantly. Goodness-of-fit terms also appear in the table. The best fits are associated with slowly decomposing materials, while the labile materials manifest more variability and are therefore more difficult to fit closely. The best values are associated with EC1, with NSSE terms ranging from 4.3 (CC) to 95 (BA). The worst fit is associated with EC4 with values ranging from 32 (CC) to 429 (BA) pointing to difficulties with the GH data. Fits for the EC2 and EC3 were comparable reasonably close to those for EC1. Table 1. Parameterization results and goodness-of-fit measure for seven amendments, each with its own Q10 value. Amendments MO

MF

MP

BA

BC

CC

CG

Parameters Q10 kR kL A NSSE Q10 kR kL A NSSE Q10 kR kL A NSSE Q10 kR kL A NSSE Q10 kR kL A NSSE Q10 kR kL A NSSE Q10 kR kL A NSSE

EC1 4.09 1.32e-4 0.104 0.97 26.9 3.72 3.13e-4 0.069 0.95 26.5 4.07 3.10e-4 0.067 0.94 36.5 3.90 1.05e-4 0.064 0.97 94.8 3.22 1.78e-4 0.034 0.97 29.0 3.00 9.62e-7 0.001 0.91 4.3 3.77 2.33e-4 0.052 0.96 23.1

10

Included Environments EC2 EC3 1.55 2.01 1.21e-4 8.93e-5 0.049 0.040 0.97 0.98 59.2 68.0 1.65 2.33 1.73e-4 9.89e-5 0.041 0.028 0.94 0.94 125.4 50.5 1.59 2.33 1.79e-4 1.01e-4 0.037 0.025 0.93 0.93 52.5 71.7 1.42 2.16 5.81e-5 3.11e-5 0.038 0.027 0.97 0.97 66.2 120.1 1.52 2.16 9.46e-5 5.19e-5 0.025 0.016 0.97 0.97 53.4 56.3 2.11 2.30 7.91e-7 4.85e-7 0.001 0.002 0.89 0.95 22.1 9.2 1.78 2.41 1.26e-4 7.91e-5 0.029 0.021 0.95 0.95 32.1 39.7

EC4 1.86 9.28e-5 0.045 0.98 255.2 2.11 1.15e-4 0.034 0.96 222.3 2.05 1.25e-4 0.028 0.95 265.1 1.82 4.69e-5 0.030 0.98 429.0 1.95 5.85e-5 0.021 0.98 275.7 2.67 2.71e-5 0.008 0.99 31.7 2.13 9.21e-5 0.026 0.96 198.3

Modeling Temperature Effects on Decomposition—Crohn

Table 2 presents an additional perspective on the goodness-of-fit of these IA ECs. The table was derived by regressing model predictions against measured data. The resulting data include r2 statistics, and slope and intercept parameters. An ideal r2 value is 1, slope should equal 1 and intercepts should equal zero. All must be considered when assessing fit. The cumulative nature of the model and aggregated data assists in relatively high r2 values. Significant differences emerge in the slope data, however. The r2 statistics range between 0.9 and 1.0 for all treatments in all ECs. The best are found in EC2, which ranges between 0.95 and 1.0. Slopes range from 0.90 - 1.02, 0.82 - 1.10, 0.76 - 1.27, and 0.59 - 2.32 for EC1, EC2, EC3, and EC4, respectively. Models clearly over- or under-predict more significantly at higher temperatures or when the entire range of temperatures is included when parameterizing models. Q10’s common to all amendments were also derived for each EC (table 3). CA values for EC1, EC2, EC3, and EC4 were 3.8, 1.7, 2.2, and 2.0, respectively. Model parameters A, kL, and kR were similar in magnitude to the IA approach. NSSE values ranged from 5.4 to 429 and again were comparable to IA values. Goodness-of-fit parameters (table 4) were also very similar to those in Table 2.

Discussion Decomposition of carbon in soils is clearly temperature dependent. Data from the incubation study associated with this research are in the final stages of collection and analysis. Nevertheless it is apparent that temperature is controlling the rates at which decay is occurring. The incubation results illustrate the importance of temperature in establishing rates of decay, an observation consistent with the scientific literature. The relative decomposability of different materials appears to be consistent from temperature treatment to temperature treatment offering support of the proposition that decay kinetics can be determined from the chemical constituents of the materials. The wide spectrum of temperatures included in this study are somewhat higher than those that occur in California’s diverse environments, ranging between 5˚C and 41˚C (41˚F and 106˚F), although frozen conditions have not been included. Temperatures in the greenhouse (GH) in the early part of the study were thermophilic, exceeding natural conditions, but were representative of composting conditions. CR, LA, and IN treatments have maintained their moisture, but GH treatments partially dried during the early part of the experiment. This will need to be considered during the analysis of results from the greenhouse treatment and will serve to inform as to the need to include moisture in temperature-independent models.

11

Modeling Temperature Effects on Decomposition—Crohn

Table 2. Linear regression statistics for seven amendments, each with its own Q10 value. Model predictions were regressed against cumulative measurements of evolved CO2. Amendments

Parameters

MO

r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept

MF

MP

BA

BC

CC

CG

2

EC1 CR 0.93 0.90 0.19 0.97 1.02 -3.37 0.97 0.97 -2.13 0.90 0.93 -6.75 0.92 0.93 1.84 0.96 0.99 -0.37 0.96 0.97 -1.39

EC2 LA 0.99 0.97 -1.16 1.00 1.02 -3.82 0.98 1.00 -3.75 1.00 1.00 -7.35 0.99 0.96 1.27 0.99 0.96 -0.47 0.99 1.01 -3.43

LA 0.98 1.10 -5.94 0.99 1.05 -3.95 0.98 1.06 -4.87 1.00 1.01 -5.49 1.00 0.93 2.98 0.99 1.03 -0.49 0.99 1.07 -4.62

EC3 IN 0.99 0.82 7.51 0.98 0.95 -0.89 0.99 0.94 -1.94 1.00 1.02 -7.47 0.95 0.99 -0.02 0.97 0.94 -4.66 1.00 0.97 -1.72

CR 0.94 1.25 0.92 0.97 1.24 -4.20 0.97 1.27 -2.84 0.90 1.25 -8.57 0.91 1.14 2.28 0.96 1.83 -0.69 0.96 1.21 -1.86

12

LA 0.97 0.87 -7.33 0.99 0.90 -9.58 0.97 0.87 -10.93 0.99 0.82 -13.81 0.99 0.76 -0.50 0.99 1.10 -0.45 0.99 0.89 -7.61

EC4 IN 0.98 0.82 6.31 0.97 0.95 5.64 0.99 0.93 5.64 0.99 0.98 5.47 0.95 1.03 5.11 0.99 0.96 -2.90 1.00 0.97 2.42

CR 0.94 1.14 1.48 0.97 1.23 -3.41 0.97 1.37 -2.39 0.90 1.49 -8.30 0.90 1.07 2.51 0.96 1.30 -0.48 0.97 1.25 -1.50

LA 0.97 0.74 -5.77 0.99 0.75 -6.14 0.97 0.77 -8.60 1.00 0.78 -11.25 0.99 0.59 0.47 1.00 0.89 -0.19 0.99 0.73 -4.88

IN 0.98 0.71 2.10 0.97 0.82 -1.07 0.99 0.85 -3.72 0.99 0.96 -8.52 0.95 0.80 0.67 0.99 0.81 -0.89 0.99 0.80 -2.80

GH 0.99 1.62 4.30 0.97 1.54 2.43 0.90 1.15 10.86 0.96 1.00 2.73 0.99 2.32 -0.62 1.00 1.57 -1.75 0.99 1.70 2.66

Modeling Temperature Effects on Decomposition—Crohn

Table 3. Parameterization results and goodness-of-fit measure for seven amendments, sharing common Q10 values. Amendments MO

MF

MP

BA

BC

CC

CG

Parameters Q10 kR kL A NSSE kR kL A NSSE kR kL A NSSE kR kL A NSSE kR kL A NSSE kR kL A NSSE kR kL A NSSE

EC1 3.75 1.26e-4 0.097 0.97 28.0 3.14e-4 0.070 0.95 26.6 2.91e-4 0.062 0.94 38.2 1.02e-4 0.062 0.97 95.2 2.12e-4 0.044 0.97 34.2 1.18e-6 0.007 0.98 5.4 2.32e-4 0.052 0.96 23.1

Included Environments EC2 EC3 1.66 1.13e-4 0.048 0.97 59.2 1.71e-4 0.040 0.94 125.4 1.68e-4 0.035 0.93 52.5 4.64e-5 0.034 0.97 66.2 8.13e-5 0.022 0.97 53.4 1.82e-7 0.002 0.95 22.1 1.40e-4 0.031 0.95 32.1

2.25 7.68e-5 0.038 0.97 68.0 1.04e-4 0.028 0.94 50.5 1.06e-4 0.025 0.94 71.7 2.96e-5 0.027 0.97 120.1 4.94e-5 0.017 0.97 56.3 6.83e-7 0.002 0.95 9.2 8.70e-5 0.022 0.95 39.7

13

EC4 2.01 9.28e-5 0.045 0.98 255.2 1.15e-4 0.034 0.96 222.3 1.25e-4 0.028 0.95 265.1 4.69e-5 0.030 0.98 429.0 5.85e-5 0.021 0.98 275.7 2.71e-5 0.008 0.99 31.7 9.21e-5 0.026 0.96 198.3

Modeling Temperature Effects on Decomposition—Crohn

Table 4. Linear regression statistics for seven amendments sharing common Q10 values. Model predictions were regressed against cumulative measurements of evolved CO2. Amendments

Parameters

MO

r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept 2 r slope intercept

MF

MP

BA

BC

CC

CG

2

EC1 CR 0.93 0.96 0.31 0.97 1.01 -3.36 0.97 1.03 -2.21 0.90 0.96 -6.91 0.92 0.83 1.66 0.96 0.87 -0.32 0.96 0.97 -1.39

EC2 LA 0.99 0.95 -1.53 1.00 1.02 -3.75 0.98 0.98 -4.44 1.00 0.99 -7.70 1.00 0.99 2.50 0.99 1.01 -0.20 0.99 1.00 -3.47

LA 0.98 1.06 -6.27 0.99 1.05 -3.99 0.98 1.05 -5.68 1.00 0.97 -9.07 1.00 0.90 2.08 1.00 1.65 -0.51 0.99 1.10 -3.70

EC3 IN 0.99 0.83 8.29 0.98 0.95 -0.72 0.99 0.95 -0.65 0.99 1.03 -1.11 0.95 1.01 1.46 0.99 0.81 -2.74 1.00 0.96 -3.18

CR 0.93 1.08 0.28 0.97 1.30 -4.28 0.97 1.34 -2.87 0.90 1.19 -8.35 0.91 1.08 2.12 0.96 1.98 -0.75 0.96 1.35 -1.95

14

LA 0.96 0.86 -7.83 0.99 0.90 -9.35 0.97 0.87 -10.74 0.99 0.82 -14.12 0.99 0.76 -0.52 0.99 1.13 -0.45 0.99 0.90 -7.34

EC4 IN 0.98 0.83 9.33 0.98 0.94 3.49 0.99 0.93 3.43 0.99 0.98 8.20 0.95 1.04 6.71 0.99 0.92 -2.55 1.00 0.96 -0.53

CR 0.94 1.05 0.98 0.97 1.32 -3.37 0.97 1.41 -2.36 0.90 1.31 -8.16 0.91 1.01 2.32 0.96 2.30 -0.84 0.97 1.36 -1.48

LA 0.97 0.73 -5.92 0.99 0.76 -5.93 0.98 0.78 -8.52 0.99 0.77 -11.84 0.99 0.59 0.39 0.99 0.99 -0.15 0.99 0.74 -4.72

IN 0.97 0.77 1.78 0.98 0.80 -1.30 0.99 0.84 -3.96 0.99 1.01 -6.75 0.95 0.82 0.88 0.99 0.59 -0.95 0.99 0.77 -3.14

GH 0.99 1.64 7.27 0.97 1.55 -1.13 0.91 1.15 9.22 0.95 0.99 12.52 0.98 2.33 0.58 0.99 1.22 -3.24 0.99 1.71 -0.52

2001-2006 Mission Kearney Foundation of Soil Science: Soil Carbon and California's Terrestrial Ecosystems Final Report: 2002009, 1/1/2003-12/31/2004

The Q10 values for EC1 were much greater than those for EC2. There are several possible explanations for this that deserve further investigation. First, it is possible that microbial populations cooling down behave differently from populations that are warming. The CR, LA, and GH environments differed significantly from each other, possibly enough to require a shift in the microbial population. It is possible that representative organisms necessary to thrive under the CR and GH conditions were not well represented in the samples, particularly the CR samples in which C evolution fell off so dramatically. There is also the possibility of experimental error due to gas leakage. This seems unlikely in the CR, LA, and IN jars as moisture was conserved in these systems. Moisture in the GH jars needed to be periodically replenished suggesting that CO2 was also likely to have escaped. The GH jars were subjected to a temperature regime that varied daily. The jar and lid would have expanded and contracted at different rates in response to these temperature changes, which may have led to small breaches during which air could have escaped. Variance in the samples of particular treatments did not differ significantly between the GH and other ECs, however. A more likely explanation is that the variable temperatures within the GH prevented the establishment of a dominant microbial community thereby reducing the overall efficiency of carbon oxidation. The resulting instability within the GH treatment microbial communities therefore prevented the full realization of the biochemical potential associated with the warmer greenhouse conditions. Such dramatic fluctuations are unlikely to occur within soils. It would therefore have been more appropriate to include a temperature-varying treatment in which changes were more gradual. Overall C evolution rates from the soils were low, compared with other studies. For example, Hartz et al. (2000) observed average cumulative losses of 35% and 14% of initial C for manures and composts, respectively, over a 24-week incubation period (25°C). Corresponding values for this study are 6% manures and 4% for composts (LA). It is not clear why decomposition was slower than expected in this study. Analysis of nitrogen mineralization data also collected during the study may help to explain why the estimates of CO2 evolution were unusually low. The pattern of gradual diminishing of CO2 levels is consistent with other studies and indicates exhaustion of available organic C. Future analysis of nitrogen mineralization data collected during the study may help explain whether evolution rates were, in fact, so low. This study did not find that C respiration rates did strict Arrhenius patterns between 5°C and 40°C, but the pattern was more suitable for narrower temperature ranges. This may have been due to the failure of the experimental approach to successfully capture patterns occurring in natural environments, or it may have been due to a failure of the Arrhenius relationship to capture environmental effects of microbial activity. Use of the equation and temperature-adjusted time does offer a means for estimating the fate of land-applied carbon, but its use over large temperature ranges is questionable. Similarly, temperature-adjusted time may be an effective approach for predicting C evolution where temperatures vary seasonally or diurnally, but not where variations are as great as occurred in the greenhouse. This experiment should be redeveloped to better represent temperatures and conditions present in the environment rather than the more extreme conditions included in this study.

1

Department of Environmental Sciences, University of California, Riverside

Modeling Temperature Effects on Decomposition—Crohn

References Crohn, D.M. and C. Valenzuela-Solano. 2003. Modeling temperature effects on decomposition. Journal of Environmental Engineering 129(12): 1149-1156. Haug, R.T. 1993. Practical handbook of compost engineering. Boca Raton, FL: Lewis Publishers. Leirós, M.C., C. Trasar-Cepeda, S. Seoane, and F. Gil-Sotres. 1999. Dependence of mineralization of soil organic matter on temperature and moisture. Soil Biology and Biochemistry 31: 327-335. Levenspiel, O. 1999. Chemical reaction engineering, 3rd Ed., New York: John Wiley & Sons. Neilsen, H, and L. Berthelsen. 2002. A model for temperature dependency of thermophilic composting process rate. Compost Science and Engineering 10: 249-257.

This research was funded by the Kearney Foundation of Soil Science: Soil Carbon and California's Terrestrial Ecosystems, 2001-2006 Mission (http://kearney.ucdavis.edu). The Kearney Foundation is an endowed research program created to encourage and support research in the fields of soil, plant nutrition, and water science within the Division of Agriculture and Natural Resources of the University of California.

16