Overview and Objective

The goal of this measurement experiment was to establish a baseline decomposition profile of food waste under controlled conditions, without any enzymatic enhancement. Since we cannot obtain the enzyme from our wet lab, this experiment provides the essential control dataset for future iGEM teams' comparison. By monitoring CO₂ emission from different substrates, we aimed to simulate how food waste behaves during storage or natural decomposition and illustrate the importance of enzyme-assisted degradation for future studies.

Soil control (no food waste) and three substrates were tested: apple flesh, apple peel, and banana peel. We chose apple as a representative substrate due to its clear compositional contrast. Its flesh is rich in sugars and water, while the peel contains more insoluble cellulose and phenolic compounds. The CO₂ release was measured using PASCO sensors in sealed flasks without oxygen supply, ensuring that the CO₂ change reflected microbial respiration.

Principle

Microbial respiration during decomposition produces CO₂, which can serve as an indirect indicator of substrate breakdown. By monitoring CO₂ emission patterns across different substrates, we can infer differences in degradability. The Gompertz model was used to fit the CO₂ release data and extract key parameters such as the asymptote, halfway point, time of maximum rate, and magnitude of maximum rate.

The Gompertz curve is a mathematical model that describes growth over time. It is a sigmoid-shaped function in which the growth starts slowly, increases rapidly in the middle phase, and slows again as it approaches its upper limit. Compared with the logistic function, which approaches both limits at a similar rate, the Gompertz curve reaches its upper asymptote much more gradually than its lower one. The model was first used to describe human mortality and later adapted in the biology field to explain population growth and other natural processes (Kirkwood).

Gompertz equation:

a = asymptote, b = lag-related constant, c = growth rate.
Derived parameters:
1. Halfway point (point when the curve reaches 50% of the total CO₂ that the sample can release)
2. Maximum rate of increase (the point of highest increase in CO2 production)

Measurement Protocols

Materials

• Substrates: apple flesh, apple peel, banana peel, and soil (control without food waste)
• Glass flasks with airtight seals
• Parafilm
• PASCO CO₂ sensor
• Distilled water

Procedure

Preparation of Substrates

1. Wash apple samples thoroughly with distilled water to remove surface impurities.
2. Cut all samples into small, uniform pieces of approximately 1 cm³ to ensure consistent surface area.
3. Divide the apple samples into two groups: flesh and peel.
4. Experiments were done separately: flesh on the first day, peel on the second day.
5. Sterilize soil using an autoclave and set aside as the control group (no food waste).
6. Weigh 10 g of each substrate and mix with 50 g of sterilized soil to maintain a consistent ratio.
7. Ensure all substrates are prepared under sterile conditions to minimize microbial contamination.

Experimental Setup

• Place each substrate mixture in a flask and seal with parafilm to simulate food-waste storage conditions.
• Maintain anaerobic conditions by avoiding any additional oxygen supply.
• Insert a PASCO CO₂ sensor through the seal to monitor gas concentration continuously.
• Calibrate the setup with room air before starting every measurement.

Measurement Process

• Maintain all flasks at room temperature (~25 °C) throughout the run.
• Log CO₂ accumulation automatically every 10 seconds using the PASCO sensor.
• Continue recording until the CO₂ trend plateaus or the 24-hour observation window ends.

Model Fitting and Analysis

• Fit CO₂ accumulation curves to the Gompertz function using `scipy.optimize.curve_fit`.
• Derive asymptote, lag-related constant, and growth rate parameters for each substrate.
• Calculate halfway point and maximum-rate metrics to compare decomposition kinetics.

Post-Experiment Handling

• Open flasks under a fume hood to release residual gases safely.
• Dispose of solid residues following institutional waste-management protocols.
• Rinse sensors and flasks with distilled water and dry before storage.

Measurement 1: Apple as substrate

Figure 1. Gompertz Model Fit and Residual Analysis for CO₂ Evolution in Apple Flesh, Apple Peel, and Control Soil Samples under Anaerobic Conditions

Table 1. Gompertz Model Parameters for Apple peel and flesh Substrates under Anaerobic Decomposition

Table 1.1 Comparison of Gompertz Parameters for Apple Flesh, Peel, and Control Soil under Anaerobic Conditions

The parameters provided a quantitative point of view into the decomposition behavior of each substrate.

The asymptote (a) represents the total CO₂ accumulated at the end of decomposition. This serves as an indicator of the overall degradation extent. The actual data showed a decrease that contradicts the asymptotic behavior of the Gompertz model. One peel sample reached the highest asymptote around 29,000 ppm, followed by both flesh samples, which ranged between 15,000 and 25,000 ppm, while the control soil sample remained much lower at approximately 3,000 ppm.

The lag time (b), which marks the period before CO₂ evolution begins, shows that control samples had the shortest lag (approximately 1.8 to 5 seconds). This shows that the substrate is less complex and has a faster microbial adaptation. In contrast, the flesh exhibited a longer lag phase (around 9 seconds), indicating that its more complex composition delayed microbial adaptation and required additional enzymatic activity before decomposition began.

The growth rate (c), expressed as 10⁻⁵ s⁻¹, showed the speed of CO₂ accumulation and therefore the rate of decomposition. Apple flesh had the fastest growth rates (12.3–12.6), showing that its higher sugar levels made it easier for microbes to break down. In contrast, apple peel decomposed more slowly (6.2–11.4), likely because its tougher cellulose fibers and phenolic compounds made it harder to digest.


Table 2. Kinetic Parameters Derived from CO₂ Evolution Curves for Apple Flesh, Apple Peel, and Control Soil Samples

Table 2.1 Comparison of Decomposition Kinetics between Apple Flesh, Peel, and Control Soil

The parameters in Table 2.1 reveal clear kinetic distinctions between substrate types. Apple peel exhibited the earliest onset of decomposition, reaching its HWP and max rate earlier. The HWP times for peel were 18889 s (Peel 054) and 19549 s (Peel 813), compared to 20669 s (Flesh 813) and 21314 s (Flesh 593) for the flesh. At these points, the corresponding CO₂ magnitudes ranged from 7487 ppm to 14283 ppm for peel and 11249 ppm to 12545 ppm for flesh. The time of maximum rate also occurred earlier for peel (15081 s – 16363 s) than for flesh (17730 s – 18435 s). The magnitude of the maximum rate was highest in Peel 054 (10508 ppm) and lowest in Peel 813 (5509 ppm), with flesh samples showing intermediate values (8277 – 9231 ppm).

Measurement 2: Banana peel as substrate

Figure 2. Gompertz Model Fit and Residual Analysis for CO₂ Evolution in Banana Peel and Control Soil Samples under Anaerobic Conditions

Table 3. Gompertz Model Parameters for Banana Peel Substrates under Anaerobic Decomposition

The asymptote values (20,000–22,000 ppm) indicate moderate CO₂ accumulation, suggesting measurable but uneven decomposition efficiency. Lag times remained short (around 2–3 seconds), implying that microbial activity began quickly after incubation, likely due to the simpler nutrient profile of banana peels. However, the growth rates (~5.5 × 10⁻⁵ s⁻¹) were noticeably lower than those observed in apple samples, showing slower CO₂ evolution and reduced microbial productivity.


Table 3. Kinetic Parameters Derived from CO₂ Evolution Curves for Banana Peel and Control Soil Samples

This table presents the kinetic behavior of banana peel samples compared to the soil control. The halfway point (approximately 27,000–28,000 seconds) indicates that CO₂ accumulation in banana trials reached equilibrium earlier than in the control group (~50,000 seconds), showing faster initial decomposition. Similarly, the halfway point for CO₂ accumulation (10,000–11,000 ppm) and the magnitude of the maximum rate (7,000–8,000 ppm) were several times higher than the control (1,600–2,000 ppm), confirming greater metabolic activity. The time of maximum CO₂ release (21,000 seconds) occurred much earlier in banana samples, aligning with their faster onset of microbial respiration. R² values above 0.91 suggest a relatively good model fit, though slightly less consistent than in apple trials.

A similar downward trend appeared in the later stages, supporting the idea that this decline results from experimental or instrumental limitations instead of substrate-specific microbial behavior. Therefore, while the Gompertz model effectively describes the overall growth phase, its predictive accuracy decreases over extended anaerobic incubation, especially when CO₂ stabilization is disrupted by physical factors.

Discussion

Using Gompertz allowed us to study parameters such as asymptote, lag time, and rate to be estimated across all samples in a consistent way. The model produced high R² and low residuals for the control sample, indicating that the Gompertz function accurately represented CO₂ accumulation under stable conditions. In contrast, the apple and banana substrates yielded lower R² and larger, irregular residuals. This shows deviations from the ideal sigmoidal shape of the function. Higher variability might have been caused by the introduction of high-cellulose, heterogeneous materials.

These results establish a baseline decomposition profile without enzymes, therefore reinforcing the need for enzymatic pretreatment in food waste management. The clear difference between apple flesh and peel samples demonstrates that cellulose-rich materials degrade more slowly, hence justifying the choice of cellulase in our project. Despite the discrepancies, we retained the Gompertz model to provide a consistent analytical framework applied to all samples. This provided a uniform basis for comparison of degradation behavior. It allowed us to consistently estimate the key parameters such as the asymptote (a), lag time (b), and growth rate (c).

We expected the apple peel to have a higher lag phase than the flesh due to its fiber-rich composition. However, due to the higher surface area, which leads to greater surface exposure compared to the flesh samples. The early spikes likely represent decomposition of readily available surfaces. Flesh samples showed a longer lag phase but had higher growth rates at later stages. The growth rate parameter is a more accurate indicator of degradation efficiency. Between the peel and flesh substrates, differences in lag time mainly describe the variations in activation speed, but the growth rate better represents the actual metabolic intensity once the decomposition begins.

As for the kinetic analysis, both peel samples reached their HWP and max rate points earlier than the flesh, which is consistent with the shorter lag time. This shows a faster onset of degradation, but it doesn't mean the peel samples have more complete degradation. This means that peel samples accelerate sooner, which is consistent with a higher surface accessibility since the peel samples have a higher surface area. Overall, peel samples might show earlier and sometimes stronger peaks, whereas the flesh samples show later initial kinetics, but the CO₂ decrease is much steadier.

On top of this, a noticeable decline after approximaely 150,000 seconds was observed in several trials. This deviates from the expected asymptotic plateau of the Gompertz model. This likely reflects sensor drift or gas leakage rather than a true biological decrease, illustrating one of the model's limitations when applied to long-term anaerobic measurements. To further verify whether this phenomenon was specific to apple samples, we conducted the same CO₂ measurement using banana peel as substrate.

Limitations

• The experiment was conducted without pH or oxygen monitoring, leaving certain decomposition factors uncontrolled.
• Fitting quality (low R² in some cases) reduced statistical strength.
• Only a limited number of replicates were performed.

Future Directions

• Repeat measurements with more replicates and better environmental control (pH, oxygen levels).
• Integrate enzyme-produced samples once secretion succeeds, to directly compare against this baseline.
• Explore multi-parameter monitoring (CO₂ + pH + O₂) to build a more complete picture of decomposition kinetics.

Kirkwood T. B. (2015). Deciphering death: a commentary on Gompertz (1825) 'On the nature of the function expressive of the law of human mortality, and on a new mode of determining the value of life contingencies'. Philosophical Transactions of the Royal Society of London. Series B, Biological sciences, 370(1666), 20140379. https://doi.org/10.1098/rstb.2014.0379