Post-harvest cane deterioration in the South African sugar industry results in significant revenue loss, estimated to be in the region of ZAR 60 million per year. The experimental work described in this thesis was carried out in the laboratories of the South African Sugarcane Research Institute (SASRI) and the University of KwaZulu-Natal (UKZN, Pietermaritzburg) under the supervision of Dr. Derek Watt (SASRI and UKZN) and Mr. Hunter (UKZN).
LITERATURE REVIEW
More recently, several cane spoilage products have been identified to predict and control processing problems at the factory level (Eggleston, 2002; Lionnet and Gooch, 2002; Eggleston and Harper, 2006; Corcodel and Mullet, 2007). Therefore, this aspect of the larger project was designed to analyze the specific biological factors associated with cane post-harvest decay in order to gain support for more stringent control measures in the factory.
Ultimately, the responsibility to control cane post-harvest degradation and its negative effects on factory processing falls on the shoulders of producers as well as the millers. Such an approach would avoid the negative effects that weakened cane has on mill processing, making growers responsible for limiting the extent of post-harvest cane deterioration and millers responsible for maintaining exemplary factory hygiene.
Plant material
The extreme difficulty in ensuring uniform temperatures during different cane burning events precluded the use of burnt cane in this study, despite the widespread burning of cane immediately before harvest in the industrial Midland areas. In this context, "mature" or "mature" denotes sugarcane where sucrose concentrations had reached a maximum in the mature internodes.
Preparation of plant material
Pre-harvest burning of sugar cane, which is very common in the SA sugar industry (about 90% of SA sugar cane, (Davis and Archery, 2007)), is reported to be another primary cause of L. It is specifically designed to consistently simulate and regulate daily temperature changes in the field or on the mill room floor that are common during harvest to suppress delays in the operation.
Sampling and replication
The duration of the harvest-to-shredding delay used between the winter and summer simulations was 9 days (approximately 216 hours). Although the HTCD period of 9 days may be considered longer than those regularly seen in the industry, there are reports of longer delays, particularly due to unforeseen circumstances such as heavy frosts in the Midlands, prolonged spring rains and/or flash fires when the mill must simultaneously process large quantities of cane (Ravnö and Purchase, 2005).
Analyses: Microbiological, biochemical and molecular .1 Microbiological: Bacteria
Homogenized samples were clarified by centrifugation (10000 x g) for 15 min at 4ºC and the supernatant transferred to fresh tubes. Data were expressed according to kit specification and the analysis was performed using KC4 software (Biotek® Instrument, Inc.). The tubes were allowed to stand for 10 min, mixed again and then 200 l of each sample and/or standard was poured into the well of a microtiter plate and the absorbance read at 520 nm (Synergy HT Multi-Detection Microplate Reader (Biotek ®) Instrument, Inc., Vermont, USA)).
After gel polymerization, the comb was removed and the gel sandwich was released from the casting stand. The core apparatus containing the two gel sandwich assemblies was then placed in the running buffer and the power control module was replaced on top of the tank.
RESULTS
Diurnal temperature variation during a simulated HTCD
The summer profile reflects the temperature and humidity between days 0 and 4, and the winter profile between days 1 and 4. Data for the first day of the winter regime is not shown due to a data logger error. The incubator can regulate the temperature, but not the relative humidity, and the output data for the latter is only a reflection of the ambient humidity in the incubator chamber.
These moisture profiles show a similar characteristic decrease in relative humidity with time, with the summer profile falling to a slightly lower level on day 4 than the winter profile. Data for the first day of the winter regime is not presented due to a data logger error.
Evaluation of bacterial cell numbers during a simulated HTCD
Summer averages correspond to levels recorded between days 0 and 4, and winter averages correspond to levels recorded between days 1 and 4. Regardless of the source of contamination, the presence of these cells within tissue samples on the day of harvest indicates the aggressive nature of L. Error bars indicate the standard error of the mean values (n = 3); absence indicates that the standard error bar was smaller than the data symbol.
This trend was also evident in inoculated stems incubated under a simulated summer temperature regime, where the overall cell count profile (Figure 4.3 A) was similar to that of L. The total cell count in inoculated stems exposed to higher temperature showed less dramatic changes like those of L.
Characterisation of bacterial populations during a simulated HTCD
These were considered representative of the bacteria making up a significant portion of the remaining total cell count. Two of the isolates from the simulated winter temperature regime (W1 and W2) showed noticeable growth on the selective medium (PES), while the remaining three isolates did not. Notably, four of the six isolates were Gram-negative, with only one isolate (S2) being Gram-positive.
Of the six isolates that grew under the simulated summer temperature regime, four were coccoid in shape and the remaining two were rod-shaped. The letters 'W' and 'S' within the reference codes refer to the sampling of the winter and summer plates respectively, where each microbial colony is isolated in numerical order.
Characterisation of bacterial populations during a simulated HTCD
The data in the table indicate the microorganisms with the most similar sequences in the database, along with the corresponding inventory number and relative identities. Colonies isolated on selective medium (PES) were PCR-amplified with universal 16S rDNA primers and then subjected to nucleotide sequence analysis.
Variation in bacterial community composition during a simulated HTCD: PCR-DGGE
The inoculated winter samples (Figure 4.9) differed from the uninoculated winter samples (Figure 4.8), in that the band indicating the presence of L. A similar significant difference (p<0.05), although to a lesser extent, was observed in culms. subjected to the simulated summer HTCD (Figure 4.14 A). Under summer conditions, the levels of sucrose in both inoculated and uninoculated stems did not change significantly (p<0.05) over the initial 7 days of the simulated HTCD (Figure 4.14 A).
Significant differences (p<0.05) were observed in glucose and fructose concentrations in response to inoculation and temperature during simulated HTCD (Figure 4.14 C, D, E and F). Similarly, there were no significant differences (p<0.05) in glucose and fructose concentrations between inoculated and non-inoculated stems during simulated winter HTCD (Figure 4.14 D and F).
Concentration of microbial by-products during a simulated HTCD
Lactic acid (or L-lactate) concentrations within stalks at harvest (day 0) were in the range 8 – 14 g mg protein-1 across all treatments (Figure 4.15). In contrast to the observation made during the summer HTCD, lactic acid concentrations in the uninoculated winter treatments did not vary significantly (p<0.05) over the nine day attenuation period. Of interest is an apparent seasonal trend in lactic acid concentrations, with higher concentrations of lactic acid found in the summer treatment samples overall compared to those of the winter samples (Figure 4.15).
Ethanol concentrations increased between days 0 and 2 in both the uninoculated and inoculated treatments, with a more marked increase in the latter. At the end of the simulated summer HTCD, a significant (p<0.05) increase in ethanol concentration occurred in the uninoculated samples between days 7 and 9, reaching a concentration of approx. 410 g mg protein-1.
Mill room scale assessment of sugarcane deterioration
Consequently, mannitol concentrations were assessed in inoculated and non-inoculated stems exposed to simulated winter and summer HTCD. Using current technologies, no mannitol is expected to be detected even at SMRI (PG Morel du Boil, 2007, Sugar Milling Research Institute (SMRI), University of KwaZulu-Natal, Durban, South Africa, personal communication). However, investigations are underway into the use of mannitol as a reliable indicator of post-harvest spoilage in sugarcane due to its relative ease of use and convenience (Eggleston and Harper, 2006).
DISCUSSION
Post-harvest sugarcane deterioration in the South African sugarcane industry
On the contrary, sucrose concentrations appeared to increase in the infected sugarcane samples between days 7 and 9. In contrast, the lactic acid concentrations in the winter samples measured between days 4 and 9 did not change significantly from those measured during harvest (day 0). (Figure 4.15 B). The effect of summer temperatures on ethanol accumulation was evident between day 0 and 2, with increases in the concentration of the alcohol occurring in both inoculated and uninoculated treatments (Figure 4.16 A).
In summer, lactic acid and ethanol concentrations in inoculated samples reached higher values compared to uninoculated samples, indicating L. Although not investigated in the current study, reports in the literature suggest that cultivar may contribute to the extent of deterioration (Ivin and Bevan, 1973 ; Wood, 1973; Wood, 1976; Irvine and Legendre, 1977).
Harvesting techniques and their effect on deterioration indicators
Microbiology of post-harvest sugarcane
For example, Curtobacterium flaccumfaciens is known to be pathogenic to soybean (Huang et al., 2007) and has a variety of other hosts. Each of the other bacterial bands migrated further down the denaturing gradient gel compared to L. Optimizing the concentrations of denaturants used in the polyacrylamide gel gradient to provide narrower gradients may allow increased migration distances and higher resolution between fragments of similar GC content ( Muyzer et et al., 1993).
Culture-based studies in this project focused primarily on determining bacterial infection levels and detecting the dominant bacterial species residing in the sugarcane stalks throughout the lag period. Optimizing the concentrations of denaturants used in the polyacrylamide gel gradient, or as mentioned, introducing and optimizing a second gradient (Double Gradient-DGGE) may also be worthwhile to improve the resolution of the DGGE profile (Haruta et al., 2002) ) .
Recommendations and conclusion
It can be concluded that the structure of the microbial community within sugarcane after harvest is very complex and a complete microbial characterization of this environment is needed for a better understanding of the processes occurring there. Eighty-second Annual Review of the Southern African Milling Season Proceedings of the Association of Sugar Technologists of South Africa, Workshop: Introducing Quality in the Sugar Industry. Separation of species of the genus Leuconostoc and differentiation of leuconostocs from other lactic acid bacteria.
Proceedings of the Association of Sugar Technologists of South Africa, Workshop: Introducing Quality in the Sugar Industry. An assessment of synthetic landfill leachate mitigation in soil and the spatial and temporal implications of leachate on bacterial community diversity.
