The effect of a crude oil spill on cereals

Environmental Pollution (Series A ) 22 (1980) 187-196

THE EFFECT OF A CRUDE OIL SPILL ON CEREALSt

E. DE JONG

Department of Soil Science, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N OWO

ABSTRACT A break in an oil pipeline in mid-winter caused oil to travel underground over a distance of about 850m. Oil moved upwards through cracks in the frozen soil especially during recovery attempts. The contamination in the affected area varied considerably both horizontally and vertically. Oil contamination and damage to the soil structure due to the clean-up efforts caused serious yield decreases during the next summer. In subsequent growing seasons oil was the major factor in reducing yields. Total above-ground dry matter and grain yield were affected similarly by oil pollution; even very smaH amounts t?f oi! (less thqn_O~2 ~ by weight ) in th e Oz30 cm or 0-90 cm depth reduced yields considerably. Oil reduced the available N content of the soil and markedly reduced water uptake by wheat from contaminated layers or from below such layers. Problems in reclaiming soils with oil contamination below the topsoil are discussed.

INTRODUCTION

Oil spills on agricultural land generally reduce plant growth for some time (Plice, 1948), although low levels of oil contamination may stimulate growth (Carr, 1919). Suggested reasons for the reduced plant growth in oil-contaminated soils range from direct toxic effects of oil on plants (Baker, 1970), lack of germination due to lack of viable seeds (Rowell, 1977) or reduced germination (e.g. Udo & Fayemi, 1975), to unsatisfactory soil conditions. Soil conditions may be unsatisfactory because of insufficient aeration due to a decrease in air-filled pore space and an increased tContribution R241 of the Saskatchewan Institute of Pedology.

187 Environ. Pottut.Ser.A. 0143-1471/80/0022-0187/$02-25© AppliedSciencePublishersLtd, England, 1980 Printed in Great Britain

188

E. DE JONG

demand for oxygen caused by oil-decomposing micro-organisms (Gudin & Syratt, 1975), a reduction in the level of available plant nutrients or toxic levels of certain elements such as Mn (Udo & Fayemi, 1975), possibly because of anaerobic conditions (Rowell, 1977), and interference with the uptake of soil water by the root system (Plice, 1948). The last phenomenon is not well documented. Most studies on the effect ofoil spills on plants have been conducted on artificially contaminated soils with the oil either uniformly mixed through the soil or applied at a uniform rate to the soil surface. This paper describes some observations on a field where the oil spread mainly underground and where contamination varied with depth and location. Thus, the usual reclamation procedures of improving aeration by cultivation and ensuring an adequate nutrient supply (Gudin & Syratt, 1975; McGill, 1976; Rowell, 1977) were only partly applicable.

SITE DESCRIPTION

In late January 1974, a crude-oil carrying pipeline running east-west broke just north of Moose Jaw, Saskatchewan, Canada. The pipeline has a diameter of 40 cm and is buried at 1 m depth. Subsequent inspection of the broken segment showed a crack of about 2 m long and up to 5 cm wide along the south side of the pipe. The asphalt-base oil (23 API gravity) has a density of 0.92 g/cm 3 at 15 °C and at the site of the break would be under about 20 kg/cm 2 pressure and have a temperature of approximately 4 °C. The soil in the area is described as a Dark Brown Chernozemic (Canada Soil Survey Committee, 1978) heavy lacustrine clay and slopes gently (about 0-5 ~o) to the south. By the end of January the top 1 m of the soil would usually be frozen in this area (Ouelle et al., 1975) and the surface covered with snow. After the first indications that oil was being lost from the pipeline, several days elapsed before the break was found. Oil was first detected in a ditch approximately 850 m south of the pipeline and subsequently also as a small pool on the soil surface about 25 m south of the line. To collect as much oil as possible, trenches and boreholes were dug to 1.5 m depth and oil collecting in them was removed with a vacuum truck. The oil appeared to be largely present as a lens of several millimetres thick just below the frost line (Fig. 1) and was under pressure, thus causing a'gusher' effect as it was being tapped or when the frozen soil cracked due to traffic. To prevent further spreading of the oil, a trench was dug around the contaminated area; the total area enclosed was just under 16 ha, with 15 ha to the south of the pipeline (Fig. 2). Approximately 2500 m a of oil were spilled and about 1600 m 3 recovered in the trenches. However, no exact estimate of oil remaining at the site can be made as an unknown amount of oil was removed from the site in the form of contaminated snow. Some oil would also have been lost by volatilisation.

189

EFFECT OF CRUDE OIL ON CEREALS

soil

m thick)

~ping down

Fig. 1.

Oil seeping down the wall of a trench from a thin lens just below the frozen soil.

Pipeline Snow covered soil Contaminated area, snow removed

A

Fig. 2.

2 ~

m

:

Perspective view looking north across the contaminated area from which the snow has been cleared.

METHODS

In the spring of 1974 it was apparent that contamination in the 16ha area was extremely variable. Surface contamination could be easily assessed but this was not the case for subsurface contamination. As the area had been fallowed in 1973, it was decided to seed the area to avoid problems with wind erosion and to use the crop's performance to delineate contaminated areas for future monitoring. Since it was

190

E. DE JONG

impractical to vary fertiliser rates with oil level, the maximum rate of N fertiliser to be used was determined by the need to avoid 'burning' of the crop in noncontaminated areas. Details of cropping and fertiliser history during 1974 and later are given in Table 1. TABLE 1 CROPS, FERTILISER APPLICATIONS AND YIELDS ON THE OIL-CONTAMINATED AREA

1974

1975

Crop Barley Oats Fertiliser ( k g / h a ) N 44 65 P 13 16 Yields ( k g / h a ) All s a m p l i n g sites, range Total 0-6300 34-5980 Grain --N o n - c o n t a m i n a t e d sites, mean and s t a n d a r d d e v i a t i o n Total 3060 + 2 2 7 0 4 9 0 0 + 1680 Grain --As reported by farmer Grain -1520

1976

1977

1978

Oats

Fallow

Wheat

3 x 63 3 x 27

6 13

65 a 16°

359-10810 12-3880

---

695-8990 276-3690

8730 + 1740 3000 + 6 4 0

---

6 2 3 0 + 1370 2490 5:59 I

3270

--

2630

" A little m o r e a p p l i e d o n a r e a s t h a t w e r e o b v i o u s l y contaminated.

In the autumn of 1974, 30 sampling sites were established in areas ranging from zero to normal growth. At each site the above-ground plant material from a I m 2 area was harvested and a soil core collected to 120cm depth. In some oilcontaminated areas sampling to 120 cm was impossible due to the wetness of the soil. The soil cores were divided into 30cm segments, air-dried and the oil content determined by extracting a i0 g sample for 17 h with dry ether in a Goldfisch fat extraction apparatus, evaporating the ether and weighing the oily residue. In noncontaminated samples this yielded less than 0.05% by weight of extractable material. Selected soil samples were analysed by the Saskatchewan Soil Testing Laboratory for pH and salt content (electrical conductivity in mmhos/cm) using a 1:1 soil-water mixture and for plant-available N, P and K using the sodium bicarbonate extraction technique described by H a m m e t aL (1970). In 1978 the moisture content and bulk density of the core segments were also determined. Some of the sampling sites located in 1974 in areas of poor growth were not contaminated by oil (see 'Results') and recovered by 1975. To keep a good representation of oilcontaminated soils, these sites were replaced in later years by new sites in areas that still showed depressed growth.

RESULTS

Due to the late seeding in 1974, the crop did not mature and only total above-ground yields were obtained. These yields varied considerably but there was no clear

19J

EFFECT OF CRUDE OIL ON CEREALS

relationship between yield and oil content (Fig. 3). Oil contents o f over 1% were associated with zero growth, but on some ofthe non-contaminated sites there was no growth due to soil disturbance (poorly filled trenches and boreholes, subsoil on the

1974

191'5 5OOO

5000

~ y -

35~-1025 n

,"o',';

%

:: ."

Jr °

i

.

; OiL % WT, O-30cm

OIL% WT, O- ] 0 ¢m

5OOO

SOOT

• %

~ y ~

NI3-1667x w* "C 8 S e e e

•

--. I OIL % WT, O - 9 0 c m

Fig. 3.

O I L % WT, 0 - 9 0 c m

Relationship between above-g.round dry matter production and oil content in the first two growing seasons after the oil spill.

surface, compaction) caused by the clean-up attempt and the resultant poor seedbed and poor germination. As damage to soil structure appeared to be a major problem, oats were seeded in 1975 as green manure crop. In July, it was apparent that all areas of poor growth were associated with oil contamination and that soil structure was no longer a problem, and it was decided to grow the oats to maturity. As the crop was sampled well before maturity, only total above-ground dry matter was determined. The final grain yield reported by the farmer (Table 1) was much lower than expected from the good stand in July, probably due to the dry spell in the last half of the growing season. The relatively low standard deviation in dry matter yields on the noncontaminated sites (Table 1) confirms that soil structure was not as important as in 1974. Oil content and yield were negatively correlated (Fig. 3), but the data showed considerable scatter, doubtlessly partly due to the non-uniform distribution of oil in the root zone. In 1976 and 1978 the contaminated area was cropped and the relationships between yield and oil content were similar to those observed in 1975. In both years grain yields were also determined and their response to oil contamination was

192

E. DE JONG

similar to that o f the total a b o v e - g r o u n d production. The standard deviation o f the yields on the n o n - c o n t a m i n a t e d sampling sites (Table 1) in 1976 and 1978 was similar to that encountered in normal fields, indicating that soil structure had returned to normal. The yield data for each year were expressed as a percentage o f the average yield o f the n o n - c o n t a m i n a t e d sites in that year, and these relative yields c o m p a r e d with the average oil content o f the 0-30 cm or 0-90 cm soil depths (Table 2). Relative, rather TABLE 2 EFFECT OF SOIL OIL CONTENT ON ABOVE-GROUND DRY MATTER PRODUCTION

Ether-extractable* material, air-dry soil

Relative productionb (%) Total above-ground

Grain

Mean

SD

n

Mean

SD

n

0-30 cm <0.05 0.05--0-25 0.26-0.50 0' 51- 1.00

99 70 67 43

23 21 41 20

32 11 6 5

101 63 66 34

23 26 46 33

28 7 4 2

1.01-2.00

29

!8

14

31

20

10

2.01-4.00 >4.00 0-90 cm' < 0.05 0.05--0.25 0-26-0.50 0.51-1.00

17 13

16 15

21 4

22 16

16 19

10 3

100 81 64 32 26 9 25

22 26 37 18 20 12 --

29 13 5 11

100 87 35 29

22 35 22 20

26 9 2 9

19

26

19

11

I0 1

19 24

16

3

--

1

1.01-2.00

2.01-4.00 > 4.00

*Uncontaminated soils contain less than 0-05~ wt of ether-extractable material. bThe average production of the uncontaminated sites in any year is assigned a value of 100~. cAverage of the ether-extractable material in the 0-30, 30-60 and 60--90cmcore segments at each site. SD, standard deviation. than absolute, yields were used to minimise differences caused by growing season climatic conditions and c r o p species. Table 2 indicates that even very small a m o u n t s o f oil (less than 0.25 ~ ) caused considerable yield depression. Carr (1919) f o u n d that small a m o u n t s o f oil do not affect, or even stimulate, crop growth. The different conclusion reached here is p r o b a b l y due to the n o n - u n i f o r m distribution o f the oil in the soil. Thus, within the root zone, concentrations m u c h higher than the averages reported in Table 2 would occur. This n o n - u n i f o r m distribution would also explain the large standard deviations o f the relative yields. Both total above-ground production and grain production are affected similarly by oil content. The soil analysis showed p H values from 7-2 to 8" I and electrical conductivities o f generally less than 4 m m h o s / c m on contaminated and n o n - c o n t a m i n a t e d areas, indicating that these two soil properties were not responsible for the yield depressions associated with oil contamination. T o illustrate the effect o f oil on available nutrient levels at harvest time, the 1975, 1976 and 1978 d a t a were grouped

EFFECT OF CRUDE OIL ON CEREALS

193

according to oil c o n t e n t (Table 3). Available N was low a n d tended to decrease with increasing oil c o n t e n t , p r e s u m a b l y due to i m m o b i l i s a t i o n b y oil d e c o m p o s i n g microo r g a n i s m s (Plice, 1948; U d o & F a y e m i , 1975; Rowell, 1977). T h e decreases in TABLE AVAILABLE

Depth (cm)

0--30

30-60

60-90

NUTRIENTS

(IN

PPM)

AT

3

HARVEST

TIME

IN

1975, 1976

NO 3- N

1978

AND

Ether extract (%)

Number of samples

P

K

Mean

SD

Mean

SD

Mean

SD

<0.05 0.05-0.25 0.25-2.0 >2.0 <0.05 0.05-0.25 0.25-2-0 >2.0 <0.05 0.05-0.25 0:25-2.0 >2.0

9 17 10 13 16 16 11 5 II 10 26 1

4.8 3.1 2.3 1.7 2.8 2.0 2.4 1.3 3.0 2.5 1'6 l

2.3 1.6 0'8 0.7 1.1 0.9 2.3 0.4 1.1 1.0 0'6 --

8.2 7.8 10.7 13.4 2.1 2.4 6.1 5.9 1.5 1.6 2"0 13

4.9 4.0 3.9 6-3 1.1 1.3 5-0 0-7 0-7 0.6 1.2 --

219 263 264 251 182 189 183 176 226 202 189 170

18 63 29 34 19 30 31 28 29 31 31 --

SD, standard deviation. available N are n o t large e n o u g h to a c c o u n t fully for the large decreases in yield (Table 2) with increasing oil c o n t e n t , especially as the crops were heavily fertilised at seeding (Table 1). Available P increased with increasing oil c o n t a m i n a t i o n , indicating that P is n o t a limiting factor in p l a n t growth or microbial d e g r a d a t i o n of the oil at this site. Oil had no consistent effect on the high level of available K. The soil cores collected in the a u t u m n of 1978 were divided into groups representing various patterns a n d degrees of c o n t a m i n a t i o n (Fig. 4). T h e n u m b e r o f °/. vol I

O

:°] I 120

JJ

50 % ~c4

•,

I.:/.'.L.':-I

• , i':::;'/:"1 •" l";'_;':._t

% vo~ i

O

50 % ~ol

"-!:4.,

P-_'.--':,!

I:t.,

t£:'.-,.':.'l

I -,, ,,,.,I.;":1

~ °1~.~1

.

!

C *9

l:::':

.,,

1:.:-:.:.:.:;:;1

.,, .,,

I:.:.:.:.:.:.:.:-I

.

.

. 5Q */o

vol

Fig. 4. Typical soil water, air and oil contents at the end of the fifth growingseason after the oil spill(the numbers in Figs 4(B) to 4(I) indicate differences in water content compared with Fig. 4(A)).

194

E. DE JONG

cores in each group varied from ! 3 in Fig. 4(A) to 1 in each of Figs 4(D) and 4(I). The oil and water contents were expressed on a per cent by volume basis, allowing calculation of the air-filled pore space of the soil. Figure 4(A) shows the soil water distribution in soil cores with no measurable oil (average ether extract less than 0.03 % weight) in the 0 to 120cm depth. The standard deviations in the moisture contents increased from 1 ~ voi. in the 0°30 cm depth to 4 % vol. in the 90-120 cm depth. The moisture contents between 30 and 120 cm depth probably represent the lower limit of the available soil water, but there has been some slight moisture recharge between the harvesting of the crop and the collection of the soil samples. Data for a similar soil (de Jong & MacDonald, 1975) suggest that the lower limit of the available water in the 0030 cm depth is about 20 ~ vol. and that the upper limit increases from about 40 ~ vol. in the 0030 cm depth to about 42 ~ vol. at 900120 cm depth. The air-filled pore space would be 1 0 ~ voi. or more when these soils are at field capacity. Figure 4(B) illustrates a case of very slight contamination (average oil contents of less than 0.05 ~ vol.) and the soil water contents are identical to those f~ii" the uncontaminated cores (Fig. 4(A)). Neither the slightly higher levels of contamination (about 0-3 ~ vol. oil) at the 900120cm depth (Fig. 4(C)) nor at the 0-30cm depth (Fig. 4(D)) caused significant differences in soil water contents. Average yields were very similar for the sites represented in Figs 4(A), 4(B) and 4(C), but the single site represented in Fig. 4(D) had a considerably lower yield, possibly due to a decrease in available N (Table 3). Oil contamination over 0-5 ~o vol. (Figs. 4(E) and 4(F)) clearly affected soil water uptake from the contaminated layer and from greater depths, the latter presumably due to reduced root penetration through the contaminated layer. Figures 4(G), 4(H) and 4(I) represent severely contaminated sites and all show considerably reduced soil water uptake relative to Fig. 4(A). In all cases the contamination of the 0030 cm depth was severe enough to have reduced water uptake by plants, yet the soil water contents are lower than in the uncontaminated sites. During the growing season the 0030cm depth would have lost water by evaporation and the low soil water contents of this layer could be due to reduced soil moisture recharge caused by the hydrophobic nature of the contaminated soil (RoweU, 1977; Myers & Frasier, 1969). A comparison of the grain yield and soil moisture data of the contaminated soil cores in 1978 showed that yields decreased by about 180 kg/ha for each centimetre of soil water not used. This decrease is about twice that calculated by Lehane & Staple (1965) from long-term data for clay soils in this area and could indicate a nitrogen-soil water interaction (i.e. as oil contamination increased nitrogen level and soil moisture uptake both decreased). By comparison, the yield of the noncontaminated sampling sites decreased by about 80kg/ha for each centimetre decrease in soil water uptake. From 1974 to 1978 there was a definite decrease in oil content, particularly during

EFFECT OF CRUDE OIL ON CEREALS

195

the 1977 fallow year when fairly heavy fertiliser applications were made (Table 1). The improvement was more pronounced for the 0-30 cm depth than at a greater depth. Unfortunately, the exact rate of disappearance of the oil cannot be calculated as the contamination was extremely variable and the method used to locate the sampling sites from year to year was not sufficiently accurate.

DISCUSSION

In most pipeline breaks oil saturates the soil around the break and gathers in pools on the soil surface. The highest oil content measured in 1974 was 13 ~ by weight and, using this as an estimate of the maximum oil-holding capacity of the soil, the escaped oil could, at most, have saturated an area of about 0.5 ha. The spill described here is different as the oil spread over a much larger area due to the frozen condition of the top 1 m of the soil. The average oil concentration was much less than 13 ~ byweight, but contamination was extremely variable, both horizontally and vertically. In Western Canada, soils are frozen for several months of the year and more research is needed on the spread of oil under these conditions. More research is also needed on movement ofoil in the soil after the spill: during 1975 and 1976 several new areas of surface contamination were found. These areas were generally located in small depressions and it is suspected that changes in soil moisture content, possibly the creation of periods of temporary saturation of the soil during or immediately following the spring thaw, caused the oil to migrate upwards. Oil in the subsoil presents a problem during reclamation. The common recommendation of ensuring good aeration and applying nitrogenous fertilisers (Gudin & Syratt, 1975; McGill, 1976; Rowell, 1977) cannot be implemented in this situation. Figure 4 suggests, at first glance, that even the most contaminated sites had adequate aeration as the air-filled pore space was close to, or greater than, 10 ~o vol. However, anaerobic microsites may still be present as the oil is not uniformly distributed and may cause enhanced anaerobiosis by microbial breakdown (Gudin & Syratt, 1975; Rowell, 1977). For example, in some of the cores taken in 1978 oil was still present in distinct layers. Whenever oil is present, the moisture content of the soil is in the available soil water range and should be adequate for microbial activity (Clark & Kemper, 1967). Available N in the subsoil (Table 3) is well below the level recommended by Kincannon (1972) for optimum microbial degradation of oily materials but it is impossible to incorporate fertilisers to these depths. To get maximum movement of NOa-N to the subsoil, the fertiliser should be incorporated in the topsoil immediately following harvest so that the NOa-N would be leached downward as soil moisture is recharged during the autumn and early spring. During the growing season soil moisture recharge is small in this area, even when the soil is fallowed (Staple & Lehane, 1952) and the fertiliser would remain in the topsoil. An additional factor that limits microbial activity in the subsoil is the temperature.

196

E. DE JONG

Ouellet et al. (1975) show that at 5 0 c m depth the soil is at, or below, 5°C for 6 m o n t h s a n d it has been estimated (Clayton et al., 1977) that the n u m b e r of Celsius degree-days over 5°C is between 1250 a n d 1400 at this depth (2250 to 2500 F a h r e n h e i t degree-days over 41°F). At greater depths the n u m b e r of degree days will be even lower.

ACKNOWLEDGEMENT This study was s u p p o r t e d by the C a n a d i a n Petroleum Association. M r L. A. Richards, M o b i l Oil, Calgary, provided the a u t h o r with details o n the initial c l e a n u p of the spill a n d Figs 1 a n d 2.

REFERENCES BAr.~R, J. M. (1970). The effects of oils on plants. Environ. Pollut., 1, 27-44. CA~4ADASOlESffRVEYCOMMITTEE(1978). The Canadian systems of soil classification. Pubis Can. Dep. Agric., 1646, Ottawa. CARR,R. H. (1919). Vegetative growth in soils containing crude petroleum. Soil Sci., 8, 67-68. CLARK,F. E. & KEMP~R,W. D. (1967). Microbial activity in relation to soil water and soil aeration. In Irrigation of agricultural lands, ed. by R. H. Hagan, 472-80. Madison, American Society of Agronomy. CLAYTON,J. S., EHRLICH,W. A., CANN,D. B., DAY,J. H. & MARSHALL,I. B. (1977).Soils of Canada, Vol. I, Soil Report. Ottawa, Research Branch, Canada Department of Agriculture. DEJol~G,E. & MACDONALD,K. B. (1975). The soil moisture regime under native grassland. Geoderma, 14, 207-21. GUDIN, C. & SVRAYr,W. J. (1975). Biological aspects of land rehabilitation following hydrocarbon contamination. Environ. Pollut., 8, 107-12. HAMM,J. W., RADFORD,F. G. & HAl.STEAD,E. H. (1970). The simultaneous determination of nitrogen, phosphorus and potassium in sodium bicarbonate extracts of soils, Proc. Technicon International Congress, Advances in Automated Analysis, 2, 65-9. Miami, Fla, Thurman Associates. KINCANNON,C. B. (1972). Oily waste disposal by soil cultivation process. U.S. Environ. Protection Agency, Env. Prot: Tech. Set. EPA-R2-72-110. LEHANE,J. J. & STAPLE,W. J. (1965). lnfluenceof soil texture, depth of soil moisture storage, and rainfall distribution on wheat yields in southwestern Saskatchewan. Can. J. Soil Sci., 45, 207-19. MCG1LL,W. B. 0976). An introduction for field/personnelito the effects of oil spills on soil and some general restoration and cleanup procedures. Edmonton, Alberta Institute of Pedology. MYERS,L. E. & FRASmR,G. W. 0969). Creating hydrophobic soil for water harvesting. J. Irrig. Drain. Dip., Am. Soe. cir. Engrs, 94 (IRI), 43-54. OUELt~T,C. E., SHARP,R. & Cto~rtJT, D. (1975). Estimated monthly normals of soil temperature in Canada. Agric. Can. Tech. Bull., 85. PLICE, M. J. (1948). Some effects of crude petroleum on soil fertility. Proc. Soil Sci. Soc. Amer., 13, 413-16. ROWELL,M. J. (1977). The effect of crude oil spills on soils--A reviewof literature. In The reclamation of agricultural soils after oil spills. Part 1: Research, ed. by J.A. Toogood, 1-33. Edmonton, Department of Soil Science, University of Alberta. STAPLE,W. J. ~' LEt-lANE,J. J. (1952).The conservation of soil moisture in southern Saskatchewan.Scient. Agric., 32, 36-47. Uvo, E. J. & FAWMI,A. A. A. (1975). The effect of oil pollution of soil on germination, growth and nutrient uptake of corn. J. environ. Qual., 4, 537-40.