Abstract
Fire intensity and spatial variability along slopes and their affect on bioturbation rates are investigated in this study at Blue Gum Creek in the Nattai national park. Many studies (mainly focusing on soil hydrophobicity) have been done to determine the effects of fire on soil erosion rates but none have attempted to determine the effect of fires of different severity on bioturbation rates. High intensity fires seem to increase the importance of ant mounding, while low intensity fires tend to increase the rates of small vertebrate scrapings.
The position in the landscape also has an effect on bioturbation rates with ant mounding greatest at the shallower slopes with a lower gravel percentage and small vertebrate scrapings unaffected by gravel percentage and being greatest on the steeper middle slope where ant-mounding rates are at the lowest.
The long-term bioturbation rates of this area where also determined so that the length of time needed to form the observed gravel layers could be calculated.
Introduction: Landscape location, bioturbation and bushfire intensity are three separate natural processes that can all affect soil erosion rates. Whether there is a connection between these processes is unknown. Burnt sites are more prone to accelerated erosion than unburnt/undisturbed sites (Zierholz et al., 1995, Blong et al., 1982). Ant mounding is likely to contribute to an increased downslope sediment transfer by providing material that can be detached by runoff (Dragovich & Morris, 2002).
Dragovich and Morris (2002) determined that bio-transfer of sediment which was mainly due to mounding was more than 10 times greater on moderately burnt areas than areas that suffered high or low intensity burns (however different studies have different definitions of high, moderate and low intensity fires). Dragovich and Morris (2002) also found that Slopewash movement was greatest on high intensity burn areas but the rate of ant mounding was lower the moderate and low intensity burn areas.
Most studies focusing on the effects of bushfires on soil are in relation to fires affecting the hydrophobic nature of soils (Shakesby et al., 2002, Zierholz et al., 1995, Prosser & Williams, 1998). This increased runoff rate (caused by the hydrophobic soil) means that hillslope erosion can be orders of magnitude greater after a fire than before it (Prosser & Williams, 1998).
Bioturbation is an important soil process but rates of bioturbation have not been studied in relation to landscape position or in relation to bush fire intensity. Within two weeks of a fire passing through Prosser and Williams (1998) recorded the first change to the soil surface as ant mounding. The funnel ant (Aphaenogaster longiceps) is an important bioturbator with mounding rates between 5.45 - 68.38 t ha-1yr-1 (Patton et al., 1995). Aphaenogaster longiceps has been recorded as able to turn over the upper 30cm of soil in 430yrs (Humphreys, 1981) and Aphaenogaster barbigula (a different species of funnel ant) has been estimated to remove 92% of the total volume of the soil within 100 years (Eldridge & Pickard, 1994) making Aphaenogaster an important agent in soil turnover. The fact that ants are unable to move larger material can also lead to texture development in soil. Other larger and relatively less common species of ant (Iridomyrmex purpureus and Camponotus intrepidus) however have been found to be far less effective agents of soil turnover (Cowan et al, 1985).
There have however been no studies on the affect that landscape position has on bioturbation rates, or the affect of different intensities of bushfire have on bioturbation rates. This study intends to investigate the spatial variability of ant mounds and small vertebrate scrapings, the effect of fire severity on bioturbation and the longer term impacts of bioturbation on the soil/sediment characteristics of Blue Gum Creek.
Study Area
The position in the landscape also has an effect on bioturbation rates with ant mounding greatest at the shallower slopes with a lower gravel percentage and small vertebrate scrapings unaffected by gravel percentage and being greatest on the steeper middle slope where ant-mounding rates are at the lowest.
The long-term bioturbation rates of this area where also determined so that the length of time needed to form the observed gravel layers could be calculated.
Introduction: Landscape location, bioturbation and bushfire intensity are three separate natural processes that can all affect soil erosion rates. Whether there is a connection between these processes is unknown. Burnt sites are more prone to accelerated erosion than unburnt/undisturbed sites (Zierholz et al., 1995, Blong et al., 1982). Ant mounding is likely to contribute to an increased downslope sediment transfer by providing material that can be detached by runoff (Dragovich & Morris, 2002).
Dragovich and Morris (2002) determined that bio-transfer of sediment which was mainly due to mounding was more than 10 times greater on moderately burnt areas than areas that suffered high or low intensity burns (however different studies have different definitions of high, moderate and low intensity fires). Dragovich and Morris (2002) also found that Slopewash movement was greatest on high intensity burn areas but the rate of ant mounding was lower the moderate and low intensity burn areas.
Most studies focusing on the effects of bushfires on soil are in relation to fires affecting the hydrophobic nature of soils (Shakesby et al., 2002, Zierholz et al., 1995, Prosser & Williams, 1998). This increased runoff rate (caused by the hydrophobic soil) means that hillslope erosion can be orders of magnitude greater after a fire than before it (Prosser & Williams, 1998).
Bioturbation is an important soil process but rates of bioturbation have not been studied in relation to landscape position or in relation to bush fire intensity. Within two weeks of a fire passing through Prosser and Williams (1998) recorded the first change to the soil surface as ant mounding. The funnel ant (Aphaenogaster longiceps) is an important bioturbator with mounding rates between 5.45 - 68.38 t ha-1yr-1 (Patton et al., 1995). Aphaenogaster longiceps has been recorded as able to turn over the upper 30cm of soil in 430yrs (Humphreys, 1981) and Aphaenogaster barbigula (a different species of funnel ant) has been estimated to remove 92% of the total volume of the soil within 100 years (Eldridge & Pickard, 1994) making Aphaenogaster an important agent in soil turnover. The fact that ants are unable to move larger material can also lead to texture development in soil. Other larger and relatively less common species of ant (Iridomyrmex purpureus and Camponotus intrepidus) however have been found to be far less effective agents of soil turnover (Cowan et al, 1985).
There have however been no studies on the affect that landscape position has on bioturbation rates, or the affect of different intensities of bushfire have on bioturbation rates. This study intends to investigate the spatial variability of ant mounds and small vertebrate scrapings, the effect of fire severity on bioturbation and the longer term impacts of bioturbation on the soil/sediment characteristics of Blue Gum Creek.
Study Area
The study area for this study was the same area as used by Shakesby et al. (in press). Blue Gum creek is found in Nattai national park (150°29.5’E, 34°13.3’S, Shakesby et al. unpublished). The fire passed through the area around Christmas 2001. The high intensity and low intensity burn sub-catchments used were also the same as those used in Shakesby et al. (unpublished). Both study catchments where found on the Western side of Blue Gum Creek. The high intensity burn area was characterised by moderate to extreme fire intensity according to GIS remote sensing imagery analysis. The low intensity burn area was characterised by low to moderate fire intensity.
The bedrock is Hawkesbury sandstone with soils ranging in texture over the catchment from loamy sands to sandy loams over the slopes. Sandy clay loams could be found in sheltered locations. The textures in the profiles recorded range from sandy loams to clayey sands (on the flat base of the high intensity burn site), medium sandy loam to medium sandy clay to fine sandy clay loam (at the mid slope position of the high intensity burn site) and sandy clay loam to sandy clay (at the flat base of the low intensity burn site) (see profiles 1, 2 and 3 in the appendix).
The bushfire that affected the area began on the 3rd of December 2001 and first affected the study site on the 24th and 25th of December (Shakesby et al., in press).
Method
The bedrock is Hawkesbury sandstone with soils ranging in texture over the catchment from loamy sands to sandy loams over the slopes. Sandy clay loams could be found in sheltered locations. The textures in the profiles recorded range from sandy loams to clayey sands (on the flat base of the high intensity burn site), medium sandy loam to medium sandy clay to fine sandy clay loam (at the mid slope position of the high intensity burn site) and sandy clay loam to sandy clay (at the flat base of the low intensity burn site) (see profiles 1, 2 and 3 in the appendix).
The bushfire that affected the area began on the 3rd of December 2001 and first affected the study site on the 24th and 25th of December (Shakesby et al., in press).
Method
Three plots of 5x1m where set up at different sites along the catchment slope in the aftermath of the Christmas 2001 bushfires. The sites where the hill top (T), the upper mid-slope (MU), the lower mid-slope (ML), the bottom of the slope (B) (on the high intensity site the bottom slope was divided into upper and lower (BU and BL) but if a comparison was made between high burn intensity and low burn intensity an average of HBU and HBL was used) and the flat at the base of the slope (F). The letter H or L in front of the site name indicated whether the site was from the high burn site area or the low burn site area (thus HT indicate the top of the high burn slope and LBU indicates the upper portion of the low burn sites bottom slope.
After 14 months the mounds of the funnel ant (Aphaenogaster longiceps) where collected and taken back to the lab. The size of small mammal scrapings was also measured, and surface bulk density was taken at a selection of sites. Three soil profiles where taken (HF, HM and LF, see profiles 1,2 and 3). The surface stone percentage was recorded. Bulk density was also recorded at a series of locations and for ant mounds and animal scrapes.
After 14 months the mounds of the funnel ant (Aphaenogaster longiceps) where collected and taken back to the lab. The size of small mammal scrapings was also measured, and surface bulk density was taken at a selection of sites. Three soil profiles where taken (HF, HM and LF, see profiles 1,2 and 3). The surface stone percentage was recorded. Bulk density was also recorded at a series of locations and for ant mounds and animal scrapes.
High Intensity Burn Flat | Notes | ||
| | 5yr 2.5/1 | | |
| | Sandy Loam | | |
| | | | |
| 30cm | 5yr 3/2 | | Gravel Layer 30-59cm (<1cm to 15cm, sub-rounded) |
| Clear | | | |
| | Sandy Loam | ||
| | 7.5yr 3/3 | | |
| 59cm | | | |
| Diffuse | 7.5yr ¾ | | |
| | Clayey Sand to | | |
| | 5yr 4/4 | | |
| 110cm | Sandy Loam | | |
Profile One
| High Intensity Burn Mid-Slope | Notes | ||
| | Organics and Charcoal | Rainsplash Features | |
| | Medium Sandy Loam | ||
| | 5y 2.5/1 | | |
| 15cm | | | |
| Clear | | | Gravel Layer 40cm (up to 40cm sub rounded) |
| | Medium Sandy Clay Loam | ||
| | 5yr 2.5/1 | | |
| 55cm | | | |
| Diffuse | | | |
| | Fine sandy clay loam | | |
| | 10yr 3/2 | | |
| | | | |
Profile Two
| Low Intensity Burn Flat | Notes | ||
| | 7.5yr 3/3 | | |
| | Sandy Clay Loam | | |
| | pH 5 | | |
| 12cm | | | |
| | 7.5yr ¾ | | |
| | Sandy Clay to | | |
| | pH 5 | | |
| 51cm | 5yr 3/3 | | |
| | 5yr 4/4 | | |
| | Sandy Clay | | |
| | pH 5.5 | | |
| 80cm | | | |
| | 7.5yr 4/6 | | Gravel Layer 105cm, rounded s/sl pebbles |
| | Sandy Clay Loam | ||
| | (faint mottles) | ||
| | pH 5 | | |
Profile Three
In the lab calculations where made on the oven dry weight of all the samples taken. Samples of ant mounds were sieved first and then dried to remove moisture. The comparison of the mounding amounts can be found in the appendix. Bulk density and gravimetric moisture content was calculated for the samples taken (although gravimetric moisture content was not used in this study). The percentage of the total sample weight that was gravel was then calculated.
The data also had to be prepared to enable a comparison between ant mounding and animal scrapes. First an average ant mounding amount and animal scrape volume was calculated from the three plots at each sample location. To enable comparison between the ant mounding (which was measured in grams) and animal scrapes (which was measured in cm3) the volume of the scrape was multiplied by the average bulk density of the site (either high intensity burn or low intensity burn). This gave both ant mounding and animal scrapes in g/5m2/14 months which when divided by 5 gave g/m2/14months.
Results
The data also had to be prepared to enable a comparison between ant mounding and animal scrapes. First an average ant mounding amount and animal scrape volume was calculated from the three plots at each sample location. To enable comparison between the ant mounding (which was measured in grams) and animal scrapes (which was measured in cm3) the volume of the scrape was multiplied by the average bulk density of the site (either high intensity burn or low intensity burn). This gave both ant mounding and animal scrapes in g/5m2/14 months which when divided by 5 gave g/m2/14months.
Results
Fire intensity did seem to play a role in the amount of bioturbation. The high intensity burn site had a much greater total amount of ant mounding and average amount of ant mounding (measure in g) (See table one). There where more sites sampled in the high intensity burn location so the upper bottom slope (HBU) and lower bottom slope (HBL) where averaged together to be compared to the low intensity burn bottom slope (LB).
| | Total Ant Mounding | Average Ant Mounding |
| High Burn | 12739.69 | 4246.563333 |
| Low Burn | 9346.22 | 2924.803333 |
| Difference (High-Low) | 3393.47 | 1321.76 |
Table One- Comparison of ant mounding between high intensity and
low intensity burn locations (in g).
However a comparison in the rates of small vertebrate scraping shows that the is an even greater difference between low intensity burn sites and high intensity burn sites, but in this case the low intensity burn location had a greater total amount and average amount of small vertebrate scrapings (measured in cm3) (See table two).
| | Total Scrapes | Average Scrapes |
| Low Burn | 36714 | 2447.6 |
| High Burn | 15640 | 1042.666667 |
| Difference (Low-High) | 21074 | 1404.933333 |
Table Two- Comparison of animal scrapes between high intensity and low intensity burn locations (in cm3).
There was approximately 36% more total ant mounding at the high intensity burn location then the low intensity burn location and 45% greater average ant mounding. The was over double (approximately 135%) the total volume and average volume of small vertebrate scrapes at the low intensity burn site compared to the high intensity burn site. There is no observed mechanism relating to fire intensity that can be used to explain this difference simply, although ant activity is related to temperature and moisture (Porter & Tschinkel, 1987 in Eldridge & Pickard, 1993) and animal scrapings could be related to searching for food.
A different picture emerges when landscape location and ant mounding rates and small vertebrate scraping rates are compared. Overall (both high intensity and low intensity burn sites added together) ant mounding is greatest on the bottom slope, followed by the flat, which is then followed by the top of the slope and then the upper mid slope and the lower mid slope. For the High burn location specifically ant mounding is greatest at the bottom slope, followed by the top of the slope, then the slope flat, followed by the upper middle slope and the lower middle slope. This is different again for the low intensity burn catchment. Here the flat has the highest amount of ant mounding, followed by the top of the slope, then the bottom of the slope, and followed by the lower middle slope then the upper middle slope. This is all summarised in table three.
A different picture emerges when landscape location and ant mounding rates and small vertebrate scraping rates are compared. Overall (both high intensity and low intensity burn sites added together) ant mounding is greatest on the bottom slope, followed by the flat, which is then followed by the top of the slope and then the upper mid slope and the lower mid slope. For the High burn location specifically ant mounding is greatest at the bottom slope, followed by the top of the slope, then the slope flat, followed by the upper middle slope and the lower middle slope. This is different again for the low intensity burn catchment. Here the flat has the highest amount of ant mounding, followed by the top of the slope, then the bottom of the slope, and followed by the lower middle slope then the upper middle slope. This is all summarised in table three.
| | Total (g) | Average (g) | Slope |
| T | 5792.4 | 870.0983333 | 8o |
| MU | 1428.03 | 238.005 | 32o |
| ML | 699.81 | 116.635 | 20o |
| B | 7204.7 | 1200.783333 | 5o-15o |
| F | 6960.97 | 1160.161667 | <5o |
| HT | 3505.16 | 1168.386667 | |
| HMU | 1127.54 | 375.8466667 | |
| HML | 73 | 24.33333333 | |
| HBU | 5133.67 | 1711.223333 | |
| HBL | 7397.97 | 2465.99 | |
| HF | 1768.17 | 589.39 | |
| LT | 2287.24 | 571.81 | |
| LMU | 300.49 | 100.1633333 | |
| LML | 626.81 | 208.9366667 | |
| LB | 938.88 | 312.96 | |
| LF | 5192.8 | 1730.933333 | |
Table Three- Spatial Variation of Ant Mounding
The animal scrapes are different to ant mounding across the landscape also. Overall the upper middle slope has the greatest volume of scrapings, followed by the lower middle slope, then the bottom slope, then the flat and finally the slope top. At the high intensity burn location the most animal scrapings occurred at the upper middle slope, followed by the bottom slope and then the slope top, followed by the lower middle slope and then the flat. The low intensity burn location differs again with the greatest volume of animal scrapes at the lower middle slope, then the bottom slope and the flat, followed by the upper middle slope and then the slope top (Shown in table four).
| | Total (cm3) | Average (cm3) | Slope |
| T | 2356 | 392.6666667 | 8o |
| MU | 15773 | 2628.833333 | 32o |
| ML | 14200 | 2366.666667 | 20o |
| B | 10954.5 | 1825.75 | 5o-15o |
| F | 7479 | 1246.5 | <5o |
| HT | 988 | 329.3333333 | |
| HMU | 10679 | 3559.666667 | |
| HML | 864 | 288 | |
| HBU | 3183 | 1061 | |
| HBL | 2515 | 838.3333333 | |
| HF | 260 | 86.66666667 | |
| LT | 1368 | 456 | |
| LMU | 5094 | 1698 | |
| LML | 13336 | 4445.333333 | |
| LB | 9697 | 3232.333333 | |
| LF | 7219 | 2406.333333 | |
Table Four- Spatial Variation of animal scrapes
Graph One compares the total ant mounding (in g) to the total animal scrapes (in g) over the landscape. This shows that animal scrapings where far greater than ant mounding at the middle slope locations and that ant mounding was only greater than animal scrapings on the tops of the slope and the bottom slope and the flats of the high intensity burn location.
Graph One (Not included here)- Ant Mounding and Animal Scrapes spatial variation
There is a much greater relationship (with an exponential trendline) between the amount of ant mounding and the gravel percent than the rate of animal scraping and gravel percent. The correlation coefficient of low burn ant mounding rates is approximately 0.66 and 0.53 for high burn ant mounding. When comparing this to animal scrapings and gravel percentage you get a r2 value of 0.0004 (for the low intensity burn location) and 0.24 for the high intensity burn location (See graphs two, three, four and five).
Graph Two (not included here)- Ant mounding compared to gravel % (low intensity burn site)
Graph Three (not included here)- Ant mounding compared to gravel % (high intensity burn site)
Graph Four (not included here)- Animal scrapings compared to gravel % (low intensity burn site)
Graph Five (not included here)- Animal scrapings compared to gravel % (high intensity burn site)
Graph Three (not included here)- Ant mounding compared to gravel % (high intensity burn site)
Graph Four (not included here)- Animal scrapings compared to gravel % (low intensity burn site)
Graph Five (not included here)- Animal scrapings compared to gravel % (high intensity burn site)
Using the ant mounding amounts, animal scrapings volume and bulk densities sample an average bioturbation rate can be calculated in g/m2/14mths (table 5). This data can be used to calculate soil turnover rates to determine the longer-term impacts of bioturbation on soil characteristics.
Site | Average Bioturbation (g/m2/14mths) |
| LT | 205.9090894 |
| LMU | 360.9251182 |
| LML | 934.2376147 |
| LB | 711.518993 |
| LF | 829.2849406 |
| HT | 291.47458 |
| HMU | 699.8826909 |
| HML | 55.41000789 |
| HBU | 528.4477397 |
| HBL | 640.3235824 |
| HF | 133.0878018 |
Table Five- Average bioturbation rates
The results of this study can be interpreted in a variety of ways. The effect of fire severity on bioturbation show that after an intense fire ant mounding becomes the most important form of bioturbation but after a less intense fire animal scrapings are the most important form of bioturbation (see tables one and two). This could be due to a variety of reasons, after an intense fire the might be less food available to small vertebrates so they are not found as commonly in intensely burnt locations, which would affect animal scraping rates. Or possibly Aphaenogaster longiceps return to a burnt location faster then other animals and therefore get a ‘head start’ on other bioturbators. Prosser and Williams (1998) recorded ant activity two weeks after a fire passed through a location. Neither of these hypotheses can be backed up with evidence from this study however.
The spatial variability of ant mounding seems to be related to gravel percent (shown in graphs two and three). This would be because ants are unable to move gravel-sized material. This corresponds with the bulge in gravel percent at the middle slope location where ant mounding is at its lowest rate (see graphs one and six).
Graph Six (not included here)- Gravel % compared to landscape location
Animal scraping rates are unaffected by gravel percentage as small vertebrates are able to move gravel sized material (see graphs four and five). However this does not explain why animal scraping rates are greatest at the middle slope locations or why the animal scraping rates are lowest by far at the slope top locations (see table four).
The longer-term impacts of these rates of bioturbation can be determined. Profile one (at the high intensity burn flat location) had a gravel layer between 30 and 59cm deep, by taking an average bioturbation rate (for the high intensity burn flat) of 133.0878018 g/m2/14mth and converting in to cm3 (by dividing it by the bulk density) you get an annual soil turn over rate of 1.05mm/yr or 1.3mm/yr (depending on whether you use the average high intensity burn location bulk density or the high intensity burn flat bulk density). This means that if the gravel layer is 30 cm deep it will take either approximately 280 or 230 years to form. If the gravel layer is 59cm deep it would take either approximately 560 or 450 years to develop. This gives soil turnover rates very similar to the rates calculated by Humphreys (1981) of 30cm in 430yrs.
At profile two (the high intensity burn middle slope location) a gravel layer was found at 40cms deep. There were large differences in bioturbation rates between the upper middle slope and lower middle slope so the time to form this gravel layer will vary depending on which sites bioturbation rates where used or if an average bioturbation rate is used. If an average bioturbation rate is used it will take either approximately 110 or 120 years to form (once again depending on using a site specific bulk density or a high intensity burn location bulk density). If the bioturbation rates for the upper middle slope are used the layer would take approximately 60 years to form (using either bulk density) and if the lower middle slopes bioturbation rates are used it will take either 600 or 740 years (depending on which bulk density is used). Both the upper middle slope and average middle slope rates of soil turnover are very high and this would have a large effect on soil profile development.
The third profile (at the low intensity burn flat) had a much deeper gravel layer (at 105cm). But the site had a high rate of bioturbation meaning that the top 105cm of soil is turned over in approximately 150yrs (whether you use the bulk density of the low burn flat or the average bulk density of the low intensity burn location. This high rate of turnover would have a large role to play in soil development.
Conclusion
The effect of fire severity on bioturbation according to the results is that ant mounding is most important after a high intensity fire and animal scraping is more important after a low intensity fire. There was no mechanism determined to be the cause of this however.
Spatial variability of ant mounding rates along a hillslope seems to be related to the amount of gravel in the soil, as ants are unable to move gravel sized material. This relates back to hill slope location, as the gravel percentage seems to relate to the spatial location along the slope. Small vertebrate scrapes however are unaffected by the percentage of gravel in the soil. There is no explanation relating to landscape position or fire severity however for the distribution of small vertebrate scrapings.
The rates of soil turnover calculated mean that the gravel layers found in the profiles could have formed in 10’s of or 100’s of years, similar to other studies on Aphaenogaster longiceps (Paton et al., 1995). These high rates would have a significant impact on soil profile development.
References
Blong, R.J., Riley, S.J. & Crozier, P.J. 1982. Sediment Yield from Runoff Plots Following Bushfire near Narrabeen Lagoon, NSW. Search. Vol. 13 No. 1-2
Cowan, J.A., Humphreys, G.S., Mitchell, P.B. & Murphy, C.L. 1985. An Assessment of Pedoturbation by two species of Mound-building Ants, Camponotus intrepidus (Kirby) and Iridomyrmex purpureus (F. Smith). Australian Journal of Soil Research. 22, 95-107
Dragovich, D. & Morris, R. 2002. Fire Intensity, Slopewash and Bio-transfer of sediment in Eucalypt Forest, Australia. Earth Surface Processes and Landforms. 27, 1309-1319
Eldridge, D.J. & Pickard, J. 1994. Effects of Ants on Sandy Soils in Semi-arid Eastern Australia: II. Relocation of nest entrances and consequences for bioturbation. Australian Journal of Soil Research, 32, 323-333
Humphreys G.S. 1981. The Rate of Ant Mounding and Earthworm Casting near Sydney, New South Wales. Search, Volume 12 no.5
Paton, T.R., Humphreys, G.S. & Mitchell, P.B. 1995. Soils: A new global view, UCL Press. London
Prosser, I.P. & Williams, L. 1998. The effect of wildfire on runoff and erosion in native Eucalyptus forest. Hydrological Processes. Volume 12, 251-265
Shakesby, R.A., Chafer, C., Doerr, S.H., Blake, W.H., Wallbrink, P., Humphreys, G.S. & Harrington, B.A. 2003. Fire intensity, water repellency characteristics and hydrogeomorphological changes following the Christmas 2001 Sydney forest fires. Unpublished
Zierholz, C., Hairsine, P. & Booker, F. 1995. Runoff and soil erosion in Bushland following the Sydney bushfires. Australian Journal of Soil and Water Conservation. Volume 8, No.4
No comments:
Post a Comment