Bioresource Technology 126 (2012) 375382

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

An evaluation of aerobic and anaerobic composting of banana peels treated with different inoculums for soil nutrient replenishment
Frank Kalemelawa a, Eiji Nishihara b, Tsuneyoshi Endo b, Zahoor Ahmad b, Rumana Yeasmin a, Moses M. Tenywa c, Sadahiro Yamamoto b,
a b c

The United Graduate School of Agricultural Sciences, Tottori University, Koyama-minami 4-101, Tottori 680-8553, Japan Faculty of Agriculture, Tottori University, Koyama-minami 4-101, Tottori 680-8553, Japan Department of Agricultural Production, Makerere University Kampala, P.O. Box 7062, Kampala, Uganda

a r t i c l e

i n f o

a b s t r a c t
This study sought to evaluate the efcacy of aerobic and anaerobic composting of inoculated banana peels, and assess the agronomic value of banana peel-based compost. Changes in the chemical composition under aerobic and anaerobic conditions were examined for four formulations of banana peel-based wastes over a period of 12 weeks. The formulations i.e. plain banana peel (B), and a mixture with either cow dung (BC), poultry litter (BP) or earthworm (BE) were separately composted under aerobic and anaerobic conditions under laboratory conditions. Inoculation with either cow dung or poultry litter signicantly facilitated mineralization in the order: BP > BC > B. The rate of decomposition was signicantly faster under aerobic than in anaerobic composting conditions. The nal composts contained high K (>100 g kg1) and TN (>2%), indicating high potential as a source of K and N fertilizer. 2012 Elsevier Ltd. All rights reserved.

Article history: Available online 19 April 2012 Keywords: Banana peels Composting Aerobic Anaerobic Inoculation

1. Introduction Uganda, sometimes referred to as Banana Republic, is among the Worlds 20 leading producers of banana, with an annual production of 615,000 metric tonnes (FAO, 2007). Ninety percent of what is produced is consumed locally, generating large volumes of wastes that result from harvesting and processing such as: market-rejected bananas, banana peels, stems, and leaves. Sentongo and Munyagwa (2004) estimated 390,550 metric tonnes of waste peelings to have come off from an annual banana production of 610,000 metric tonnes. These wastes are currently dumped in landlls, rivers, lakes and other unregulated dumping grounds where upon decomposition, pose serious environmental hazards. Though sanitary landll is the most common disposal method used, it is proving highly unsustainable due to land scarcity and pollution risks, especially in urban and peri-urban centers. Attempts to utilize these materials as fodder for peri-urban zerograzing cattle have been inadequate in addressing the menace being caused. Therefore, there is need to develop appropriate disposal practices to mitigate negative impact on the environment. Composting is among the waste management strategies that are gaining interest as a suitable option for organic waste disposal with economic and environmental prots, since the process leads to a
Corresponding author. Address: 4-101 Minami, Koyama-cho, Tottori 680-0941, Japan. Tel./fax: +81 857 31 5367. E-mail address: [email protected] (S. Yamamoto).
0960-8524/$ - see front matter 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2012.04.030

stabilized nal product, which can be used to improve and maintain soil quality and fertility (Larney and Hao, 2007). Composting of banana wastes to produce a stable fertilizer resource needs to be investigated as an alternative waste management pathway to landlling, for potential benet to the resource poor farmers of East Africa. Application of crop residue compost to soil is considered as a good management practice because it stimulates soil microbial growth and activity with the subsequent mineralization of plant nutrients (Eriksen, 2005). The effect of particular plant residue composts on soil properties, however, depends on its dominant nutrient component (Chaves et al., 2004). Currently, smallholder farmers in Uganda use simple local composting techniques such as pit or aboveground piling methods to compost animal and household wastes. Pit composting creates anaerobic conditions within the pit that may result into slow stabilization rate and increases the potential risk to human health and the environment through groundwater contamination by leachate (Hudgins and March, 1998). On the contrary, aboveground piling method, usually done under shade, presents aerobic composting conditions since air currents constantly blow through the pile. The efcacy of pit (anaerobic) and heap (aerobic) composting of banana residue-based wastes is however, still largely unknown. Several materials have been used as starters to enhance decomposition in the past; the actions of earthworms in enhancing composting have been extensively studied and include: substrate aeration, mixing, grinding, fragmentation, enzymatic digestion and microbial decomposition of substrate in intestine of earthworms

376

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382

(Sharma et al., 2005). Addition of urea for reducing the C/N ratio of crop waste has been practiced in the past (Gaind and Gaur, 2000), but the recent inclination towards organic farming requires the use of nitrogen rich organic wastes that can substitute the chemical source of nitrogen. Poultry litter and cow dung are such wastes that may t in the category of organic nitrogen supplements (low C/N ratio <12); moreover, these are widely available on most Ugandan farms. The effect of inoculation on the quality of banana residue compost has, however, not been claried. This study aimed to evaluate the efcacy of aerobic and anaerobic composting methods for banana peels and the effect of poultry litter, cow dung and earthworm as starter inoculums on composting, as well as to determine the dominant mineral nutrient components of banana residue-based compost. 2. Methods 2.1. Composting materials Fresh banana peels were collected from Yamazaki Baking Co. Ltd, Okayama, Japan, and used as the major constituents of the compost mass. Partially composted cow dung and poultry litter were procured from the Field Science Centre, Faculty of Agriculture-Tottori University, and a local poultry farm in Tottori prefecture, Japan respectively, while the earthworms Lumbricus terrestris, a commercial product (MARUNI CHI 21 Inc., Japan) were purchased from a local market. The cow dung and poultry litter were air-dried and ground prior to use in this experiment. A summary of selected physico-chemical characteristics of the original materials prior to composting is presented in Table 1. 2.2. Experimental apparatus Mini- ten-liter (14 30 24 cm) styrene foam boxes were used to simulate aerobic and anaerobic composting conditions under laboratory conditions. For the aerobic setup, boxes were tted with a 10 mm-diameter vent pipe at the top and a 3 mm air-supply tube connected to a 1.5 L min1 air pump (Chikara a 1500, Nisso-int. Co, Japan) at the bottom. Air supply rate in the compost mass was maintained at 0.3 L min1 kg1 organic matter as recommended by Li et al. (2008), while the boxes under anaerobic composting were completely sealed. Both setups were tted with portable automatic thermal recorders (TR-71U, T&D Co., Japan) mid-way into the compost mass. 2.3. Composting procedures Fresh banana peels were withered to 60% moisture content under natural atmospheric conditions for about 4 weeks, chopped into small pieces of 25 cm in length and 5 kg aliquots placed in the two different sets of compost boxes. These were then inoculated and thoroughly mixed with 0.5 kg of either cow dung (C), poultry litter (P) or earthworm (E) at 1.6 kg/m3 density to obtain the following formulations: (i) plain banana peels (B) serving as control, (ii) banana peel + cow dung (BC), (iii) banana peel + poultry litter (BP) and

(iv) banana peel + earthworm (BE). The two composting conditions, three formulations along with the control were then replicated three times, to give twenty-four sampling units. These were arranged in a randomized complete block design (RCBD) in a temperature-controlled room set at 40 C for a period of 12 weeks. In the aerobic setup, materials were turned weekly to allow uniform airow throughout the composting period. Moisture content was periodically replenished at 60% using the weighing method. 2.4. Chemical analysis Prior to composting, the original materials were analyzed for physicochemical characteristics: pH, electrical conductivity (EC), total nitrogen (TN), total carbon (TC) and total mineral nutrients: (Phosphorus (P), Potassium (K), Calcium (Ca) and Magnesium (Mg). The banana peels were oven dried and separately milled along with the cow dung and poultry litter. The pH and EC of samples were measured in 1:10 compost-water extracts using pH and EC meters (Models HM-30S and CM-30R-DKK, TOA electronics Ltd., Japan), respectively. Total C and N were determined by dry combustion method using an automated C-N coder (Model MT 700 Yonaco, Japan). Total mineral nutrients: P, K, Ca and Mg were determined after digestion with a H2O2H2SO4 mixture. Total P in the digest was determined colorimetrically on a spectrophotometer (Model U-2001, Hitachi Co., Japan) using phosphomolybdate blue method (Murphy and Riley, 1962); while total K, Ca and Mg were measured using an atomic absorption spectrophotometer (Model Z-2300 Hitachi Co., Japan). To monitor the composting progress, the following qualitative parameters were measured: compost temperature, C/N ratio, pH, EC, available mineral nutrients (ammonium nitrogen (NH4+N), nitrate nitrogen (NO3N), P, K, Ca, and Mg). Temperature in the compost masses was recorded using automatic thermal recorders (TR-71U, T&D corp., Japan). To analyze other parameters, composite subsamples were collected from ve symmetrical locations of each box at day 0 and then bi-weekly until the end of the twelve-week composting period. Collected samples were divided into two parts: one was oven dried, milled and analyzed for TC and TN by the method described earlier, while the second part was stored at 4 C and used for analysis of pH, EC, available mineral nutrients: NH4+-N, NO3N, P, K, Ca, and Mg. The pH and EC were determined using the same procedure already described above; while NH4+-N and NO3N were determined colorimetrically after extraction with KCl solution (1:10 w:v) using spectrophotometer (Spec model U-2001, Hitachi Co., Japan) at 635 nm and 410 nm, respectively. Available P was determined colorimetrically in NaHCO3 at 710 nm according to Olsen and Sommers (1982), while available K, Ca and Mg were quantied on atomic absorption spectrophotometer (Model Z-2300 Hitachi Co., Japan) after leaching with neutral CH3COONH4 (Alburquerque et al., 2004). 2.5. Statistical analysis Experimental data presented are the means of three replicates; Statistical analyses were executed using Stat view software.

Table 1 Selected physiochemical properties of original materials prior to composting. Parameter (Compost-water mixture, 1:10, w/v) pH Banana peel Cow dung Poultry litter 9.2 0.20 6.6 0.12 7.7 0.10 EC (dS m1) 0.51 0.08 11.10 0.05 10.02 0.03 TC (%) 31.52 1.10 31.03 1.18 32.11 0.98 TN (%) 1.05 0.29 2.31 0.23 4.25 0.18 C/N Ratio 30.1 0.01 13.4 0.01 7.6 0.02 Total mineral nutrients (g/ kg air dried) P 0.22 0.02 2.41 0.07 14.80 0.73 K 66.40 3.12 38.61 2.12 17.62 1.13 Ca 2.11 0.50 22.63 0.38 51.91 0.81 Mg 1.02 0.03 6.91 0.15 8.63 0.09

Values are means SD of three replicates.

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382

377

One-way analysis of variance (ANOVA) was used to test for treatment signicance and mean differences were compared using Fishers protected least signicant difference (LSD) at 5%. In addition, principal component analysis (PCA) was performed to elucidate total data variability with respect to correlation-ships between formulations, different physicochemical characteristics and composting conditions with composting time using XLSTAT 2011. Each formulation was treated as categorical data, converted to dummy variables and PCA performed along with other variables. Due to very low cumulative contribution ratio, some dummy variables were omitted.

unsuccessful, as total mortality in all vermi-beds occurred just two days into composting; likely due to the unfavourable environmental conditions such as the high temperatures of >40 C, pH >7 and moisture content of >45%. (Nagavallemma et al., 2004). Other lethal substances might have been produced at the start of composting, owing to the strong smell observed, likely from volatile ester compounds; however, this needs further study to obtain conclusive evidence. 3.2. pH and EC Generally, pH varied within a narrow range of 9.29.3 at the start to 9.39.7 at the end of the composting period. A precipitous rate of increase occurred in the rst 2 weeks, then a gentle increase with composting progress. Initial increase in pH was likely due to ammonia production following protein degradation as was explained by Alburquerque et al. (2006). Decline in the rate of pH increase later on was likely due to ammonia volatilization and increase in concentration of water-soluble base cations arising from organic matter decomposition. All 24 samples showed high alkalinity, (pH >9), likely due to the inherently high K content (6.64%) in the banana peels. Available K+ in water-soluble form could have combined with bi-carbonic acids (HCO3) produced during the organic matter mineralisation to form potassium hydroxide (KOH), a strong base. Similarly, biodegradation of the organic acids and mineralization of organic compounds also favour pH rise (Paredes et al., 2000). Aerobic composting showed slightly higher pH than anaerobic process, likely due to higher K release (Fig. 2) that resulted from the higher bio-oxidative activity. Among the inoculum treatments, B formulation showed higher pH as compared to BC and BP inoculum treatments, and this might have been due to the higher K content in B. The nal compost pH of >9 is slightly higher than the recommended range of 68, but typical in the composting of organic wastes, (Barreira et al., 2008). The strongly alkaline banana peel-based compost could be potentially benecial as an acidity control agent for the increasingly acidic soils of Uganda. Electrical conductivity increased from 0.53 to 0.77 dS/m in B, 0.58 to 0.79 dSm1 in BC and 0.56 to 0.83 dSm1 in BP during composting which was probably due to chemical transformation into more soluble inorganic salt forms during organic matter decomposition, and the concentration effect due to the net loss of dry mass (Silva et al., 2009). The effect of poultry litter inoculation was highly signicant (p < 0.05) as compared to cow dung and plain peels under both aerobic and anaerobic conditions. Though the trend under anaerobic composting was not

3. Results and discussion 3.1. Temperature of composting piles Temperature proles for the different composting treatments are illustrated in Fig. 1. Initial composting temperature was about 18 C, and rapidly increased to a maximum of 69.5 and 58.5 C under aerobic and anaerobic conditions (in BP), respectively, during the rst 2 weeks, dropping slightly to 54 C in week 6 and nally to 37 C during the last 2 weeks in the aerobic setup. Under anaerobic conditions, temperature dropped to between 45 and 52 C after week 6 and eventually to about 44 C during the last 2 weeks. This trend conforms to an earlier three-stage composting cycle reported by Inbar et al. (1993) in which temperature increases to about 45 C during the rst 3 days (mesophilic stage), reaches a maximum of 72 C within 4 weeks (thermophilic stage) and then the ambient temperature phase, which lasts for 8 weeks. The faster and higher temperature surge under aerobic as compared to anaerobic was likely due to increased bio-oxidative activity. Bernal et al. (2008) had found that during the different stages of biodegradation, organic compounds are degraded to CO2 and NH3, with consumption of copious amount of oxygen. Among the inoculum treatments, temperature build-up was in the order BP > BC > B, suggesting that inoculation promoted temperature build-up as compared to plain banana peels. This could be attributed to increased microbial activity, as a diverse population of organic matter degrading microorganisms, including Pseudomonas had been reported in animal wastes (Nodar et al., 1990). Additionally, mineral nutrients N and P in the inoculum formulations (BP and BC) might have promoted prolic growth and activity of the microbes. These are believed to be essential in the biodegradation of organic compounds (Gupta and Rao, 1998). Earthworm inoculation was

Aer obic
70

Anaer obic

7 0

Temperature (C)

50

5 0

30

3 0

10

1 0

10

12

10

12

Weeks
Fig. 1. Temperature variation of composting piles. bars means SE of triplicate data analysis. Plain banana peel (B), Banana peel + Cow dung (BC), Banana peel + Poultry litter (BP), small vertical

378

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382

Aer obic
8000
8 0 0 0

Anaer obic (b)

NH4+-N, mg/ kg dw

(a )
6000

4000

4 0 0 0

2000

2 0 0 0

0
0 2 4 6 8 1 0 1 2

1 0

1 2

16.0

1 6

(C)
NO3 --N, mg/ kg dw
12.0
1 2

(d)

8.0

4.0

0.0
0 2 4 6 8 1 0 1 2

1 0

1 2

4.0

4 . 0

P, g/ kg dw

(e)
3.0
3 . 0

(f)

2.0

2 . 0

1.0

1 . 0

0
0 2 4 6 8 1 0 1 2

1 0

1 2

150

1 5 0

(g)
K, g/ kg dw
135
1 3 5

(h)

120

1 2 0

105

1 0 5

90

9 0

10

12

10

12

Weeks
Fig. 2. Available nutrient patterns during composing. vertical bars means SE of triplicate data analysis. Plain banana peel (B), Banana peel + Cow dung (BC), Banana peel + Poultry litter (BP), small

consistent, mean EC values were signicantly lower (p < 0.05) than those obtained under aerobic composting. Final EC values were

well below the upper limit of 4.0 dS m1 which is considered tolerable by plants of medium salinity sensitivity (Lasaridi et al., 2006).

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382 Table 2 Selected physiochemical characteristics of the nal compost. Available nutrients (g kg1 dwt) pH Composting condition Aerobic EC (dS m1) TC % 22.44 0.19 21.95 0.26 22.24 0.21 26.75 0.33 23.18 0.19 26.70 0.36 2.04 0.09 2.09 0.11 2.18 0.13 2.09 0.19 1.84 0.18 2.07 0.15 11.0 0.71 10.5 0.67 10.2 0.67 12.8 0.31 12.6 0.29 12.9 1.10 TN C/N ratio NH4+N 1.27 0.69 0.77 0.13 1.07 0.09 2.23 0.72 1.22 0.64 4.37 0.10 NO3N 0.012 0.01 0.012 0.01 0.013 0.03 0.001 0.01 0.002 0.05 0.002 0.01 CH3COONH4-extractable P 2.01 0.39 2.43 0.29 3.90 0.03 1.71 0.06 1.92 0.09 3.44 0.84 K 139 5.96 122 3.36 123 4.27 118 6.36 111 7.75 104 4.27 CA 1.62 0.31 2.85 0.90 3.71 0.50 2.43 0.38 3.72 0.41 4.61 0.12 Mg

379

Formulation Compost-Water mixture 1:10 w/v B BC BP B BC BP 9.7 0.10 9.5 0.12 9.6 0.04 9.4 0.01 9.5 0.05 9.3 0.03 0.77 0.03 0.79 0.08 0.80 0.08 0.59 0.01 0.57 0.01 0.65 0.03

Anaerobic

1.10 0.06 1.52 0.16 1.41 0.01 2.43 0.27 3.91 0.53 3.32 0.29

Values are means SD of three replicates.

3.3. Available nutrient contents Knowledge of the mineral nutrient content of compost is of prime importance as it varies widely, and is an important determinant of the end use for the compost (Zethner et al., 2000). Variation in the concentration of available primary (N (NH4+-N and NO3N), P, K) and secondary (Ca and Mg) mineral nutrients during the composting process is shown in Fig. 2. 3.3.1. Nitrogen Most of the mineral N in the composts was in ammonium nitrogen (NH4+-N) form, with very low detectable nitrate nitrogen (NO3-N). NH4+-N rapidly increased and attained its peak (2.28 3.90 g kg1 under aerobic and 2.476.52 g kg1 under anaerobic conditions) in the rst 24 weeks and gradually decreased to between 0.771.27 g kg1 (aerobic) and 1.654.37 g kg1 (anaerobic) toward the end of composting period (Fig. 2a and b). The initial rapid rise was due to thermophilic decomposition of organic forms of nitrogen such as proteins and amino acids into ammonia (Paredes et al., 2000), however, the decline in the latter part was not proportional to the increase in NO3N concentration, which suggested that some of the NH4+N might have been lost through NH3 volatilization and/or microbial denitrication. Tiquia et al. (2002) had reported high NH3 volatilisation losses and immobilisation by microorganisms under strong alkaline conditions (pH >9). Throughout the composting period and in all inoculum treatments, NH4+N concentration was signicantly higher (p < 0.05) under anaerobic conditions as compared to the aerobic process, probably as a result of volatilization losses due to pressurised air circulation (Solano et al., 2001) and the higher rate of nitrication under aerobic composting that might have resulted from relatively more favourable conditions especially oxygen for survival of the nitrifying bacteria (Fig. 2a). Nitrication is a bio-oxidative process, requiring oxygen for normal functioning of the nitrifying bacteria (Tiquia and Tam, 2002). Inoculated treatments also showed higher NH4+N concentration than plain peels, with BP showing the highest effect for most of the composting period, but at the end, concentration was not signicantly different from BC and B. Higher NH4+N under inoculated conditions was likely due to additive effect and higher bioactivity, especially in BP, while the subsequent rapid decline in BP was due to higher ammonia volatilization arising from easily mineralizable compounds such as uric acid in the poultry litter. NO3N exhibited a clear trend only under aerobic composting conditions. Generally, a meagre increase occurred during week-2, and remained stable until week-6. A rapid increase was then observed after week-10, with the peak occurring in week-12 (Fig. 2c and d). The low NO3N accumulation in the early part was probably caused by the high amount of ammonia and temperature during that period, which could have inhibited nitrication through

microbial inactivation and/or de-naturation (Stentiford and de Bertoldi, 2010). Among the different inoculums in the aerobic setup, NO3N concentration was signicantly higher (p < 0.05) in BP during the rst 28 weeks as compared to BC and B, but rate of increase reduced towards the end of the composting period. In the anaerobic condition, NO3N concentration was also highest in BP treatments but the differences with BC and B were not statistically signicant. Summarily and in all treatments however, NO3N concentration remained very low (<0.02 g kg1) throughout the composting period, probably due to the high pH conditions (9.29.9). The nitrifying bacteria responsible for nitrate formation are largely inactivated by the pH value of >7.5 and temperature >40 C (Tiquia et al., 2002).

3.3.2. Phosphorus and potassium P and K concentration in all treatments increased with composting time, with formulations under aerobic conditions showing signicantly (p < 0.05) higher concentration than those under anaerobic conditions (Fig. 2e and h). The increase in P and K during composting was possibly caused by concentration effect arising from the higher rate of carbon loss that occurs when organic matter is decomposed or mineralised into CH4 or CO2 (Chitsan, 2008). Among the inoculum treatments; P was in the order BP > BC > B, but K behaved exceptionally; it was signicantly higher in plain peels than inoculated treatments in the order: B > BC > BP under both composting conditions. Besides the higher microbial activity, higher P concentration in inoculated treatments was possibly also due to additive effect of P from the inoculums, especially poultry litter, which had the highest P content (14.8 g kg1). The unique K behaviour in the inoculated treatments might have been partly due to Potassium assimilation and immobilisation by microbes (Hankin et al., 1976). Overall, Potassium was notably and consistently highest in all treatments and of all other mineral nutrients ranging >9.0%, probably due to the high inherent content in banana peels, suggesting that banana peel compost might be a good source of K fertilizer. The K content of banana-peel compost is much higher when compared to that of other composted materials such as: municipal solid waste compost 7.1% (Mbarki et al., 2008), Olive-mill waste compost 2.3% (Altieri and Esposito, 2010), and bovine manure 2.1% (Gil et al., 2008).

3.3.3. Calcium and magnesium Available calcium (Ca) and magnesium (Mg) showed similar trend under both aerobic and anaerobic composting conditions, with the aerobic attaining higher concentrations than anaerobic setup. Signicantly higher Ca and Mg releases were observed in the inoculated than plain banana peels. At the end of the composting period, Ca and Mg nutrient concentration was in the order; BP > BC > B under both composting conditions (Table 2).

380

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382

Fig. 3. C/N ratio development during compositing. bars means SE of triplicate data analysis.

Plain banana peel (B),

Banana peel + Cow dung (BC),

Banana peel + Poultry litter (BP), small vertical

3.4. C/N ratio The initial C/N ratio of banana peels (30) reduced to just under 16 following a 4-week pre-composting treatment (withering) to adjust moisture content to about 60%. During composting, C/N ratio usually has a decreasing tendency and is an important parameter to index composting progress and maturity (Charest and Beauchamp, 2002). Fig. 3 shows C/N ratio variations of the different formulations under both aerobic and anaerobic composting conditions. At the start of composting, C/N ratio of the formulations averaged 15.7, 15.3 and 14.6 for B, BC and BP formulations, respectively, and by the end of composting, it had dropped to 11.0, 10.5 and 10.2 under aerobic and 12.8, 12.6 and 12.0 under anaerobic composting for B, BC and BP formulations respectively. This decline was likely due to mineralization of substrates present in the initial composting materials, or an increase in total N resulting from a concentration effect as C is biodegraded (Sanchez-Monedero et al., 2001). Aeration signicantly facilitated C/N ratio decline
Variables factor map (PCA)

(p < 0.05) as compared to anaerobic composting; for instance, BPformulation under aerobic conditions attained a C/N ratio of 12 in weeks 45 while it took 12 weeks to attain the same C/N under anaerobic conditions. Regardless of the aeration condition, inoculated treatments generally showed faster rate of C/N ratio decline than plain banana peels. Overall inoculum effect was in the order BP > BC > B, under both aerobic and anaerobic composting conditions. Faster C/N decline under aerobic and among the different inoculum treatments could be largely attributed to higher bio-oxidative activity, as earlier described. The nal compost C/N ratio of 1012 in all treatments was within the range of 1015, which is ideal for nal stabilized composts (Boldrin et al., 2010). Total Carbon (TC) degradation of the four formulations under both aerobic and anaerobic composting conditions showed only a slight difference, owing to earlier (36%) carbon losses during the withering treatment prior to composting. Initial TC concentrations ranged 33.038.5%, decreasing across all treatments as composting progressed to between 22.0% and 22.5% under aerobic and
Variables factor map (PCA)

Poor O2

(a)
1.0

(b)
Poor O2
1.0

BP.an

NH4.N Mg

Ca

0.5

Composting condition

0.0

0.0

BC.an

Dim 2 (26.3%)

BP.a

Composting condition

Dim 2 (22.84%)

0.5

CN

EC Weeks

B.an

BC.a

NO3.N

-0.5

B.a

-0.5

pH

Rich O2

-1.0

-1.0

Rich O2
-1.0 -0.5 0.0 Dim 1 (28.64%) 0.5 1.0

-1.0

-0.5

0.0 Dim 1 (32.79%)

0.5

1.0

Composting progress/Mineralization

Composting progress/Mineralization

Fig. 4. (a) Variable factor map (PCA) for total variation in the three compost formulations with time, under aerobic and anaerobic composting conditions. (b) Variable factor map (PCA) for total variation in physiochemical characteristics with composting time. B, BC and BP are the compost formulations are earlier described, while a and an denote aerobic and anaerobic conditions respectively.

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382

381

23.226.8% under anaerobic conditions, at the end of composting (Table 2). This decline could have resulted from degradation of labile C as composting progressed. Higher Carbon loss under aerobic conditions is reective of the higher bio-oxidative activity as compared to anaerobic conditions. Final compost TC content is similar to that from other crop biomass sources, such as wheat: 21.9% (Gaind and Gaur, 2000). 3.5. Overall physiochemical characteristics variability Principal component analysis (PCA) of the total variation in the three compost formulations under aerobic and anaerobic composting conditions is shown in Fig. 4a, and that among physiochemical characteristics in Fig. 4b. Weeks of composting and C/N ratio showed very high factor loading, and are denoted by Dim 1 in this analysis. Dim 2 denotes the various physicochemical characteristics monitored during the composting period. Inoculated formulations and aerobic composting conditions had a signicant positive effect on the composting process as compared to the un-inoculated formulations and anaerobic composting conditions. This reafrms the signicance of inoculation and aeration during the composting process as earlier described. Among the physiochemical interactions, K and pH, were strongly positively, but negatively correlated with NH4+N. The strong positive correlation between K and pH further suggests that K might have been the precursor for the high pH recorded throughout the composting period. Such conditions favour NH4+N volatilization, hence the negative correlation observed. 3.6. Final compost characteristics The nal compost formulations had high N and K mineral nutrient concentrations ranging 2.042.18% and 12.213.9% under aerobic and 1.842.09 and 10.4411.86% under anaerobic composting conditions respectively. Other nutrients were quite low: P (<0.39%), Ca (<0.46%), and Mg (<0.39%) under both composting conditions (Table 2). The compost was also highly alkaline (pH > 9).These parameters are within the reported range for compost materials (Sanchez-Arias et al., 2008).

References
Alburquerque, J.A., Gonzalvez, J., Garca, D., Cegarra, J., 2004. Agrochemical characterization of Alperujo, a solid by product of the two phase centrifugation method for olive oil extraction. Bioresour. Technol. 91, 195200. Alburquerque, J.A., Gonzalvez, J., Garca, D., Cegarra, J., 2006. Composting of a solid Olive-mill by-product (alperujo) and the potential of the resulting compost for cultivating pepper under commercial conditions. Waste Manage. 26, 620 626. Altieri, R., Esposito, A., 2010. Evaluation of the fertilizing effect of olive mill waste compost in short-term crops. Int. Biodeter. Biodegr. 64, 124128. Barreira, L.P., Philippi, J.A., Rodrigues, M.S., Tenorio, J.A.S., 2008. Physical analyses of compost from composting plants in Brazil. Waste Manage. 28, 14171422. Bernal, M.P., Alburquerque, J.A., Moral, R., 2008. Composting of animal manures and chemical criteria for compost maturity assessment. A review. Bioresour. Technol. 100, 54445453. Boldrin, A., Korner, I., Krogmann, U., Christensen, T.H., 2010. Composting: mass balances and product quality. Solid Waste Managem, 9781-405-17517-3. Charest, M.H., Beauchamp, C.J., 2002. Composting of de-inking paper sludge with poultry manure at three nitrogen levels using mechanical turning: behavior of physico-chemical parameters. Bioresour. Technol. 81, 717. Chaves, B., DeNeve, S., Hofman, G., Boeckx, P., Van Cleemput, O., 2004. Nitrogen mineralization of vegetable root residues and green manures as related to their biochemical composition. Eur. J. Agron. 21, 161170. Chitsan, L., 2008. A negative-pressure aeration system for composting food wastes. Bioresour. Technol. 99, 76517656. Eriksen, J., 2005. Gross sulphur mineralisationimmobilisation turnover in soil amended with plant residues. Soil Biol. Biochem. 37, 22162224. FAO, 2007. In the Food and Agriculture Organization of the United Nations Website. <http://faostat.fao.org/site/567/default.aspx>. Gaind, S., Gaur, A.C., 2000. Quality assessment of compost prepared from y ash and crop residue. Bioresour. Technol. 87, 125127. Gil, V.M., Calvo, F.L., Blanco, D., Sanchez, E.M., 2008. Assessing the agronomic and environmental effects of the application of cattle manure compost on soil by multivariate methods. Bioresour. Technol. 99, 57635772. Gupta, G., Rao, V., 1998. Bioremediation of phenol with poultry litter microorganisms. J. Environ. Sci. Heal. A. 33, 23952407. Hankin, L.H., Poincelot, R.P., Anagnostakis, S.L., 1976. Microorganisms from composting leaves: ability to produce extracellular enzymes. Microb. Ecol. 2, 296308. Hudgins, M., March, J., 1998. In situ Municipal Solid Waste Composting Using an Aerobic Landll System. American Technologies Inc., pp. 144157. Inbar, Y., Hadar, Y., Chen, Y., 1993. Recycling of cattle manure: the composting process and characterization of maturity. J. Environ. Qual. 22, 857863. Larney, F.J., Hao, X., 2007. A review of composting as a management alternative for beef cattle feedlot manure in southern Alberta. Canada Bioresour. Technol. 98, 32213227. Lasaridi, K., Protopapa, I., Kotsou, M., Pilidis, G., Manios, T., Kyriacou, A., 2006. Quality assessment of composts in the Greek market: the need for standards and quality assurance. J. Environ. Manage. 80, 5865. Li, X., Zhang, R., Pang, Y., 2008. Characteristics of dairy manure composting with rice straw. Bioresour. Technol. 99, 359367. Mbarki, S., Labidi, N., Mahmoudi, H., Jedidi, N., Abdelly, C., 2008. Contrasting effects of municipal compost on alfalfa growth in clay and in sandy soils: N, P, K, content and heavy metal toxicity. Bioresour. Technol. 99, 67456750. Murphy, J., Riley, J.P., 1962. A modied single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27, 3136. Nagavallemma, K.P., Wani, S.P., Lacroix, S., Padmaja, V.V., Vineela, C., Babu Rao, M., Sahrawat, K.L., 2004. Vermicomposting: recycling wastes into valuable organic fertilizer. Global Theme on Agroecosystems Report no. 8. International Crops Research Institute for Semi-Arid Tropics, Andhra Pradesh (India), p. 20. Norder, R., Acea, M.J., Carballas, T., 1990. Microbial composition of poultry excreta. Biol. Wastes 33, 95105. Olsen, S.L., Sommers, L.E., 1982. In: Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis. Part 2: Chemical and Microbiological Properties. Am. Soc. Agron., Madison, WI, pp. 403430. Paredes, C., Roig, A., Bernal, M.P., Snchez-Monedero, M.A., Cegarra, J., 2000. Evolution of organic matter and nitrogen during co-composting of olive mill wastewater with solid organic wastes. Biol. Fertil. Soils. 32, 222227. Sanchez-Monedero, M.A., Roig, A., Paredes, C., Bernal, M.P., 2001. Nitrogen transformation during organic waste composting by the Rutgers system and its effects on pH, EC and maturity of the compost mixtures. Bioresour. Technol. 78, 301308. Sanchez-Arias, V., Fernndez, F.J., Villaseor, J., Rodrguez, L., 2008. Enhancing the co-composting of olive mill wastes and sewage sludge by the addition of an industrial waste. Bioresour. Technol. 99, 63466353. Sentongo, M., Munyagwa, W., 2004. Uganda Youth Voluntary Effort in Advancement and Environment Protection. In: An NGO making briquettes from banana peelings: opportunities for utilizing waste biomass for energy in Uganda, pp. 6769. Sharma, S., Pradham, K., Satya, S., Vasudevan, P., 2005. Chemical changes of spinach waste during composting and vermicomposting. AJB 10 (16), 31243127. Silva, M.E., Lemos, L.T., Cunha-Queda, A.C., Nunes, O.C., 2009. Co-composting of poultry manure with low quantities of carbon-rich materials. Waste Manage. Res. 27, 119128.

4. Conclusions Banana peel decomposition and nutrient mineralisation were faster and higher under aerobic and inoculated (especially poultry litter) conditions as compared to anaerobic and un-inoculated composting conditions, probably due to the oxygen and supplementary nutrient supply that could have promoted prolic growth of microorganisms essential for mineralisation of organic substances. High potassium and nitrogen content in the composts, presents a high potential for banana peel compost to be used as a K and N nutrient source. The highly alkaline pH of banana peel compost suggests it could also be useful in acidity remediation in the acidic parts of Uganda. Acknowledgements This work was funded by the Global Centre of Excellence in Dry land Sciences Program, and the Japan International Cooperation Agency. The authors would also like to express their deep appreciation to: Management of Yamazaki Bread Factory, Okayama for the timely provision of specimen banana wastes. Members of the Environmental Soil Science and Plant nutrition Laboratories, Tottori University, Japan 2008/2011, for assistance in laboratory analyses and data compilation.

382

F. Kalemelawa et al. / Bioresource Technology 126 (2012) 375382 Zethner, G., Gotz, B., Amlinger, F., 2000. Quality of Austrian Compost from Derived Waste Collection Summary. Available at: <http://www.ubavie.gv.at/ publikationen/Mono/M133s.HTM>.

Stentiford, E., de Bertoldi M., 2010. Composting process. Solid Waste Manage. 978 1-405-17517-3. Tiquia, S.M., Richard, T.L., Honeyman, M.S., 2002. Carbon, nutrient, and mass loss during composting. Nutr. Cycl. Agroecosys. 62, 1524. Tiquia, S., Tam, N., 2002. Characterization and composting of poultry litter in forced aeration piles. Process Biochem. 37, 869880.