Evaluation the energy production potential of aggressively invasive Prosopis juliflora Sw (Fabacea): an emerging threat in the Bundala Ramsar Wetland

Prosopis juliflora is an invasive plant species rapidly expanding in the continents. Invasion of P. juliflora in Bundala Ramsar Wetland (BRW) in Sri Lanka has created a number of biodiversity and conservation issues. This study was conducted to assess the possibility to utilize this invasive plant as a fuel source for local industries. The moisture content, wood density, ash content, volatile matter, fixed carbon content, biomass/ash ration and calorific value of P. juliflora were measured and compared with Leucaena leucocephala, which is a widely used fuelwood source in Sri Lanka and elsewhere. P. juliflora , performed better than L. leucocephala for most of these parameters. Ash content was comparatively higher in P. juliflora than that of L. leucocephala ; however, biomass to ash ratio of P. juliflora was significantly lower ( at 0.05 level of significance) than that of L. leucocephala, suggesting its suitability as a fuelwood source. Further, the fuel value index of P. juliflora (3,276) was slightly lower than that of L. leucocephala (3,336), a non-significant difference. P. juliflora and L. leucocephala reached Fiber Saturation Point values within drying periods of 24 and 27 days, respectively. According to our estimates of energy properties, 1 Kg of P. juliflora could be used to substitute 0.5 L of diesel and furnace oil as well as 5 kWh (5 units) of electricity. As such, we recommend harvesting P. juliflora from BRW as a potential fuelwood energy source for local industries.

countries where it shows invasive characteristics [6].
The use of biomass especially for thermal energy in the industrial sector in Sri Lanka has been recognized in the country's Energy S ector Development Plan for 2015-2025. Several plant species including Gliricidia sepium, Acacia auriculiformis and Calliandra calothyrsus have been identified as common fuelwood species in Sri Lanka due to their high calorific values (4000-5000 kcal kg -1 ) [12]. Additional fuelwood species found in Sri Lanka are Eucalyptus grandis, Eucalyptus camaldulensis, Casuarina equisetifolia, Clusia rosea, Leucaena leucocephala and Paraserianthes falcataria [12].
Further, several species i n c l u d i n g Alstonia macrophylla and Acacia auriculiformis have been introduced by the Forest Department of Sri Lanka to fulfill timber and other wood requirements of the country [12,13].
The present production of biomass for energy generation requirements is only adequate to fulfill 55% of the total energy demand of the country. The rest is fulfilled by hydropower and by imported fossil fuels. Therefore, there is a need to identify alternative energy sources including biomass where extraction would not contribute to land degradation, deforestation and biodiversity loss.
In this context, the use of P. juliflora in energy production may be a good option.
However, studies in the region to evaluate the full energy profile of P. juliflora are lacking. Therefore, this study was conducted with the objective of estimating energy and mechanical properties of P. juliflora to understand its applicability and efficiency as a source of energy.
Study outcomes would be helpful in controlling the spread of invasive P. juliflora while determining its potential as an alternative energy source to partially fulfill the thermal energy demand of local industries in the region.

Sampling area
P. juliflora was sampled from BRW that lies in the south coast of Hambantota District, low country dry zone, Sri Lanka (6 0 08'-6 0 14'N, 81 0 08'-81 0 18'E). The total area of BRW is 3,698 ha. The mean annual rainfall in the area ranges from 900 mm to 1300 mm, with two peak rainfall periods during April-May and October-November. The mean annual temperature is about 27 0 C while relative humidity ranges between 76% and 81% [9].
BRW has diverse vegetation, which shows natural succession from low creeping plants to climax forests described as thorny, dry semi-evergreen and dry mixed evergreen. The forest canopy of the tropical semi-deciduous forest area of BRW consists of typical, native forest vegetation M. hexandra [9]. Canopy layers of other areas consist of species such as Salvadora persica, Limonia acidissima, Strychnos potatorum and Drypetes sepiaria. The lower tree layer consists of species such as Allophylus cobbe, Benkara malabarica, Capparis zeylanica, Erythroxylum monogynum, Memecylon umbellatum and Ochna lanceolata [14]. In addition, an array of vegetation occurs in low lying areas along Bundala Lagoon, as aquatic vegetation, salt marshes, mangroves, etc. Presently, the dry mixed evergreen forest area of BRW is dominated by invasive P. juliflora species. This invasion is more abundant along the lagoon as well as inland water bodies [9].

Selection of sampling sites
Sampling sites were selected using handheld GPS (Garmin eTrex Summit, Taipei/Taiwan).

Sampling of P. juliflora and L. leucocephala
All the P. juliflora individuals with Diameter Breast Height (DBH) > 5 cm in the sampling sites were considered for the study (Figure 1). Sixty individuals were categorized into different diameter classes; DBH class 01: 10 cm to 19.99 cm; DBH class 02: 20 cm to 29.99 cm and; DBH class 03: 30 cm to 39.99 cm. L. leucocephala is another widely used fuelwood species found in Hambantota District where P. juliflora is abundant. In order to compare the energy potential of P. juliflora with that of L. leucocephala, a site near BRW where L. leucocephala was abundant was selected. Optimum sample size to compare energy characteristics of P. julfiflora with those of L. leucocephala was determined according to Pearson et al. [15]. Altogether 40 individuals of P. juliflora and 20 individuals of L.
leucocephala were selected to obtain samples. GPS coordinates of the site were 81°16'41.526"E and 6°12'20.537"N. Study sites of P. juliflora and L. leucocepahala are illustrated in Figure 2.
In order to extract samples to estimate fuelwood properties of P. juliflora, sample wood disks were cut at breast height from the 40 stems found in the study plot. The disks were packed in airtight polyethylene bags. Later, in the laboratory, 5 cm x 5 cm x 5 cm cubes were cut from the disks to test for fuelwood properties such as, moisture content, density, specific gravity, ash content, volatile matter content, fixed carbon content and biomass/ash ratio. An additional set of specimens was prepared from 20 trees of L .leucocephala for comparison of fuelwood characteristics with P. juliflora. All specimens were prepared in triplicate from each individual selected tree for both species.

Drying profiles of sample specimens
In addition to the previously mentioned disks that were cut from sample trees, another set of samples with dimensions of 18 cm x 2 cm x 2 cm was prepared from both species for the construction of drying profiles. These specimens were prepared within 24 h of sample extractions and each specimen was labeled and packed in airtight bags to avoid moisture loss. The moisture content of P. juliflora was measured following t h e oven dry method [16].

Wood density:
Densities were determined using the volume and oven dry weight of samples. Volumes of the specimens were determined using the water displacement method as recommended in ASTM D2395-17.

Specific gravity:
Specific gravity was determined using the maximum moisture content method [17].

Ash content and volatile matter:
Ash content of each specimen was determined using the loss of ignition method [18]. The same samples used for moisture content testing were used to determine the percentage of volatile matter [19].

Fixed carbon content:
Fixed carbon content was determined by subtracting the total of volatile matter percentage and ash content percentage from 100. Each data point of every parameter considered here was estimated by taking the average of three replicate samples [16].

Biomass/ash ratio:
The biomass/ash ratio of two species was determined by dividing oven dry weights of the samples by ash weights [20].

Gross calorific values:
Gross calorific value of each sample was determined according to ASTM D 5865 standard procedures.

Fuelwood Value Index (FVI):
FVI was calculated for each individual sample using calorific value, wood density, ash content and moisture content and applying Eq. (1) [21]. In the FVI index, ash content and moisture contents were given in ratios (g g-1), density was given in g cm-3 and calorific value in kJ g -1 .

Equation (1)
Calorific value data and density data of diesel and furnace oil were provided by the Ceylon Petroleum Corporation [22]. The number of liters of diesel and furnace oil that can be replaced by 1 kg of each fuelwood type was calculated by the authors.

Statistical analysis
The values for moisture content, density, specific gravity, ash content, volatile matter and fixed carbon content of each DBH class were subjected to one-way Analysis of Variance (ANOVA) using MINITAB ® Version 14 statistical software after following the Anderson Darling Normality test. Percentage values of moisture content, ash content, volatile matter content and fixed carbon contents were subjected to arcsine transformation before doing Normality test. Tukey's pair-wise comparison was carried out to test for significant differences between the three DBH classes for each measured parameter. In addition, comparisons of energy potential of P. juliflora and L. leucocephala were subjected to the same statistical tests.

Comparison of fuelwood characteristics of P. juliflora in pre-selected diameter classes
Comparison of moisture content, density, specific gravity, ash content (%), volatile matter content (%) and fixed carbon content (%) among three diameter classes (10-19.99 cm, 20.00-29.99 cm and 30.00-39.99 cm), resulted in significant differences (at 0.05 level of significance) in volatile matter content (higher values) and fixed carbon content (lower values) for diameter class 3 (30 cm-39.99 cm) relative to the other two diameter classes. All other parameters tested did not show significant differences (at 0.05 level of significance) between the three diameter classes ( Table 1) Table 1: Mean values ± SEM of fuel wood characteristics in three DBH classes of P. juliflora For each characteristic, mean value indicated by different superscript letters are significantly different from each other (p<0.05). Since there was no significant difference (at 0.05 level of significance) among three diameter classes of P. juliflora in terms of its energy performance values obtained, all the diameter classes were pooled for further analysis. When comparing energy properties (moisture content, density, specific gravity, ash content (%), volatile matter content (%) and fixed carbon content (%), biomass/ash ratio and gross calorific value or GCV of P. juliflora and L. leucocephala, other than ash content, all other tested parameters had higher (better) values in P. juliflora compared to L. leucocephala. Although ash content of P. juliflora (1.70%) was slightly higher than that of  L. leucocephala (1.20%), biomass to ash ratio of P. juliflora was greatly lower than that of L. leucocephala which indicates a positive trait for use as fuelwood. However, considering overall performance estimated using FVI, L. leucocephala has a slightly higher value (3336) than P. juliflora (3276) ( Table 2).

Drying profile for P. juliflora and L. leucocephala
The drying profile of P. juliflora and L. leucocephala, constructed for a period of 35 days of drying, is shown in Figure 4.  Table 3). Table 3: Amount of fossil fuel replaced by 1 kg of P. juliflora wood compared to L. leucocephala P. juliflora samples had significantly lower (at 0.05 level of significance) moisture content than L .leucocephala samples; higher moisture content decreases wood calorific value [25].
In addition, species with low moisture content are favored as fuelwood due to their superior combustion characteristics and higher Net Calorific Values (NCVs) [18,26].
As trees with higher wood density contain more heat per unit volume [27], P. juliflora, with a significantly higher wood density, could be considered as a better option in energy production than L. leucocephala. Specific gravity of P. juliflora was found to be significantly higher (at 0.05 level of significance) than that of L. leucocephala, supporting the findings of Chavan et al. [25]. Ash content of P. juliflora (2.25%) was slightly higher than that of L. leucocephala (1.5%), as reported by Chavan et al. [25]. and Sterculia urens (1.4%) ( [18,28]. P. juliflora has significantly lower (at 0.05 level of significance) volatile matter content compared to L.leucocephala (at 0.05 level of significance). Typically, volatile matter content of biomass will volatilize and burn as a gas [27]. Results of a study by Oduor and Githiomi [29] support the volatile matter content obtained for P. juliflora in the present study. P. juliflora has significantly higher (at 0.05 level of significance) fixed carbon content (20.00) than L. leucocephala (14.56). Fixed carbon content is determined by the volatile matter content and ash content of the species. A study by Pasiecznik and Felker [11 ] suggested that P. juliflora has high heat of combustion due to a high amount of carbon content.
Natural air-drying is the most reliable method that could be used to dry fuelwood in a tropical country like Sri Lanka due to prevailing dry and sunny conditions. In addition, fuelwood pieces have to be small enough to facilitate rapid drying [ 3 0 ] . In the present study, initial drying rate was descending with the number of drying days and F S P was achieved by P. juliflora in 24 days and L. leucocephala within 27 days. The findings of the present study are supported by a study conducted by Groves et al. [30] for four Australian tropical species. However, in the Australian study, the four species reached FSP of 28%-30% within eight days of drying. Variations of results observed in the two studies may be due to climatic conditions in two regions, including air temperature, humidity, rainfall and evaporation. The FSP is generally considered as 28% to 30% moisture content. It is not a fixed point and it may vary in different species from 19% to 30% [31]. Both species in the present study achieved FSP in the 15%-20% range for moisture content. The FSP values could be affected by the drying rate of the species. Hence in the natural air-drying process it is practical to dry both species for a period of one month in order to achieve efficient drying, thereby increasing fuelwood performance including NCVs of the species due to less moisture content.
According to a study carried out in India by Pasiecznik and Felker [11], the calorific value of P. juliflora was 17.6 kJ Kg -1 . Comparison of calorific values of P. juliflora w i t h L.
leucocephala and other commonly used fuelwood species in Sri Lanka suggests that P.
juliflora could be utilized as an alternative fuelwood species in Sri Lanka. The fact that this is an invasive species warrants additional research into its use as fuelwood.
The current study shows that there is no statistically significant difference (at 0.05 level of significance) between the FVI of P. julilfora (3,276 ± 274) and that of L.
leucocephala (3,336±389). Therefore, FVI findings suggest that although there are significant differences (at 0.05 level of significance) between P. juliflora and L.
leucocephala for a few fuelwood characteristics, overall performance is better than or equal to L. leucocephala. Previous studies [ 2 0 ] showed that high calorific value, high wood density, low ash percentage, low wood moisture and a high biomass/ash ratio are highly desirable fuelwood properties. Fuelwood species with such properties have FVI over 2,000 and this includes P. juliflora (FVI value of 3,276).
According to Siri et al. [32], there is a considerable preference for P. juliflora due to its high heat capacity and availability. Industries which depend on fuelwood to fulfill their energy requirements are willing to use P. juliflora but they do not have access to obtain P. juliflora in BRW.
According to the results, there were no significant differences ((at 0.05 level of significance) among different DBH classes of P. juliflora for energy characteristics, such as moisture content, density, specific gravity and ash content. Complying with this result, Dias and Marenco [33] have indicated that there is no significant difference of moisture content and wood density of certain species with DBH ( [34,35]. That might be the possible reason for the current finding of no difference between densities in DBH classes. Results of specific gravity tests in three DBH classes in the present study support results found by Navarro [36] where the diameter of tropical species does not significantly affect nor is a determining factor of specific gravity. As such, P. juliflora could be harvested from BRW for fulfilling energy requirements in the area without considering different sizes or DBH classes in terms of energy performance.
Salt cedar and Russian olive are two species which show invasive characteristics including vigorous resprout after cutting. Both species have extensive root systems as they are sporophytes, hence complete removal of trees with the root system is necessary to prevent regrowth [37]. Since mesquite is generally similar in structure to salt cedar and Russian olive [37], cutting and uprooting could be considered as a potential solution in controlling the growth of mesquite, P. juliflora. Wood pellets, bio oil, and charcoal are promising sources of energy and P.juliflora can also be utilized to extract chemicals such as resins and polymers.
Further testing is recommended on such aspects of using P. juliflora.
Past studies have shown that managing P. juliflora infested lands for charcoal production with a four-year harvest cycle is more profitable than using P. juliflora pods for flour production. However, authors have recommended vigilant regulation in using P. juliflora to avoid exploitation of native plants. Management measures are necessary to prevent charcoal production sites from becoming potential seed sources [38].
A study conducted by Kurt et al. [39] concluded that there is a great potential for the usage of P. juliflora as a source for energy production. Further, the production of charcoal from P. juliflora could be identified as an additional income opportunity. Additionally, reclaiming degraded lands and remediating heavy metal-contaminated soils, and feeding lifestock can be identified as other benefits that can be obtained from P. juliflora [39].
Therefore, this study implies that the replacement of fossil fuel and electricity usage with P. juliflora fuelwood in certain industries in Hambantota region could be considered as a viable option in terms of economic and environmental sustainability for a developing country like Sri Lanka. At the same time, with the intervention of relevant government and non-governmental institutions, cutting and uprooting trees for energy would be an ideal solution to control aggressive invasiveness of P. juliflora on the biodiversity of BMR.

Conclusion
There were no significant differences among three DBH classes (10.00-19.99 cm, 20.00-29.99 cm, 30.00-39.99 cm) of P. juliflora for certain energy characteristics including moisture content, ash content, density and calorific value. Therefore, individuals of any diameter of P. juliflora could be harvested for energy production process. When considering calorific value as a measure of evaluating fuel wood, P. juliflora lie in the range of most common types of fuel wood species used in Sri Lanka. Although there were significant differences among certain energy characteristics, when considering overall performances, there is no significant difference between P. juliflora and L.leucocephala. Both P. juliflora and L.leucocephala should be air dried for a period of a month before utilizing it as a fuel wood. P. juliflora and L.leucocephala wood can be used to replace fossil fuel (diesel and furnace oil) and electricity to a certain extent.
Average size (DBH 23.16 cm) P. juliflora individual can replace 60 l fossil fuel and 620 kWh electricity units in kWh. Therefore, P. juliflora can be utilized as an alternative source of energy as a partial solution to foreign oil dependency for local industries in Sri Lanka.

Availability of data and materials
All data generated or analyzed during this study can be provided as a separate file in machine readable format upon request.