delivered the opinion of the court:
This case concerns the alleged pollution of a farm drainage ditch by a dairy cattle operation conducted by Harold Craig and Robert Craig on a combined dairy and grain farm adjacent to the ditch. The principal issue is whether circumstantial evidence presented before the State Pollution Control Board (P.C.B.) was sufficient to support a determination by that board that manure flowing from a storage pit on the Craigs’ farm into the ditch had killed a substantial number of fish in that ditch.
Proceedings arising from the spillage of the manure were brought by the State Environmental Protection Agency (E.P.A.) against the Craigs before the P.C.B. The Craigs appeal from an order of the P.C.B. which: (1) found them guilty of violating certain Water Pollution Regulations and section 12(a) of the Environmental Protection Act (Ill. Rev. Stat. 1973, ch. 111½, par. 1012(a)) which resulted in a fish kill, (2) ordered them to cease and desist all further violations, (3) ordered them to pay *1,345.40 to the State Game and Fish Fund as reimbursement for the fish killed, and (4) assessed a penalty of *500 against them.
The Craigs contend that: (1) the decision that they were responsible for the fish kill was improper as a matter of law and not based upon an inference that might be drawn from facts in evidence, (2) the foregoing decision was contrary to the manifest weight of the evidence, (3) the requirement that they pay for the fish kill was improper, (4) the finding that they violated Rule 203(a) of the State’s Water Pollution Regulations was improper as a matter of law, (5) P.C.B. admitted improper exhibits into evidence, and (6) the cease and desist order was illegal and improper.
For some time prior to July 18, 1975, the Craigs had operated an 830-acre dairy and grain farm on the northeast edge of Indianola in Vermilion County. At the time about 130 head of dairy cattle and 100 beef cattle were on the farm. The drainage ditch known as Swank Creek flowed through the farm in a southeasterly direction. About 100 yards from the creek, manure was stored in a pit which was behind a bam. On or prior to July 18, 1975, some manure from that pit had flowed into the creek. On that date Harold Craig attempted to stop the flow of the manure into the creek by the erection of an earth dam. However, some runoff from the manure continued to get into the ditch.
Also, on July 18,1975, Tom Smith, an employee of the E.P.A. inspected four portions of Swank Creek and took samples of the water from each *67place. He testified that at a road bridge near the northeast corner of Indianola, he found dead fish and found the water to have a slightly brownish cast and a slight manure odor. Southeast of that point at the last road bridge before the creek entered the Little Vermilion River, he made the same observations as to the water but observed no dead fish. He then returned upstream to the first bridge upstream from the first sampling point. He found the water there to be clear and odorless and observed no dead fish. He then went to a point between the first and third sampling points. There he found manure in the stream and discharge coming from a tile. He described the manure as being 25 yards upstream from the tile and “approximately 100 yards east of the dairy farm.” Smith described the stream as having a very low water mark. He did not testify to finding any dead fish here but his notes, admitted into evidence, state that he found some dead minnows at that point.
On July 19, 1975, Smith returned to the scene with Richard Ryczeck also of the E.P.A. They went to about the same points observed by Smith the day before and in general made the same observations. Ryczeck testified that the discharge from the drain near the place where the manure was in the stream was brown in color and had a manure odor. He stated that they found dead fish and an accumulation of manure solids below the drain. Ryczeck also testified to finding a drain discharging raw sewage into the creek at the point where Smith had taken his first water samples and observed dead fish the day before and a further raw sewage discharge some 50 yards downstream from that point. Evidence indicated that these drains came from inside Indianola.
Kim Brummett, a fishery biologist for the State Department of Conservation, testified that on July 19, 1975, he made an estimate of the number of fish killed by counting the dead fish in three 100-foot sections of the ditch, all of which were downstream from the place where the manure entered the stream. These points were (1) a pasture between the point first mentioned and the road bridge on the northeast corner of Indianola where Tom Smith had taken his first water samples, (2) an area further downstream near the school, and (3) still further downstream near the last bridge before the creek empties into the river. He testified that nitrogen and bacteria from the manure would have the effect of reducing the oxygen content in the water and that fish would die if the oxygen content got too low. He said that he thought that the fish had been killed within the 24-hour period previous to his count. He said he could not form an opinion as to whether the manure from the Craig farm had caused the fish kill and said that it could have been caused by municipal sewage from Indianola.
Dr. Peños Kokoropoulos, professor of urban and environmental engineering at Southern Illinois University, Edwardsville Campus, was *68the principal witness for the respondents. He had seen the ditch a few days before testifying and had examined Smith’s analysis sheets and notes admitted into evidence. He stated that the presence of 4 milligrams of oxygen per liter of water was the minimum amount of oxygen to sustain fish life in water. He noted that of the samples of water taken by Smith on July 18, 1975, none were shown on Smith’s analysis sheets as having less than 4.1 milligrams per liter. Those notes, however, showed no results for the sample taken at the place where the manure was found. The witness also noted that these documents also set forth that the samples were taken between about 6:45 and 7:45 on July 18, 1975. He testified that in his opinion it was very unlikely that the oxygen count would have been as low as 4 milligrams per liter at any earlier time that day.
Dr. Kokoropoulos also examined the analysis of the fecal bacteria found in the water and concluded that those figures showed that the majority of this bacteria in the water at the points of the manure and the nearby drain had come from animal waste but that the majority of this bacteria in the water further downstream including the place where Smith first observed dead fish came from human waste. He, like Mr. Brummett, was unable to form an opinion as to the cause of the fish kill from the information before him.
The theory of the E.P.A. and the P.C.B. order is that the fish died due to lack of oxygen in the water which was, in turn, caused by the flow of manure into the stream. The order places much emphasis upon the speculative and conjectural nature of the other possible explanations given by the Craigs. The finding of the Board was supported by neither direct evidence nor expert opinion. No attempt was made to examine any of the dead fish to determine the cause of death. No evidence was presented as to the oxygen count at the point of manure discharge on July 18, 1975.
Actually, the strongest support for the finding that the Craigs were responsible for the fish kill came from Dr. Kokoropoulos’ testimony that Smith’s records indicated that the oxygen levels at the sample points downstream from the manure flow were at levels only slightly above that necessary to support aquatic life. There was also evidence that the concentration of manure, nitrogen and bacteria was higher at the point of the manure entry than downstream.
The conclusion reached by the P.C.B. must have been based upon an inference from the foregoing circumstances that the oxygen level at the point of the manure entry was enough lower than that elsewhere that the dead fish later found downstream were killed by the lack of oxygen at that point or that the oxygen level of a larger portion of the stream had earlier been below that required to support aquatic life. Neither Dr. *69Kokoropoulos nor Mr. Brummett were able to base an opinion upon the facts presented. It is beyond our technical knowledge to determine if such an inference would be logical. The P.C.B. has the expertise to make such a determination but cannot do so in the absence of expert testimony in the record. See our recent decision in Farney v. Anderson (1978), 56 Ill. App. 3d 677, 372 N.E.2d 151, citing Smith v. Department of Registration & Education (1952), 412 Ill. 332, 106 N.E.2d 722, and McCormick, Evidence §353, at 849 (2d ed. 1972).
We agree with the Craigs that the determination that they were responsible for the fish kill was not based upon an inference that can be drawn from the circumstantial evidence. We conclude that the P.C.B. erred in so ruling.
We need not consider other arguments made by the Craigs with reference to the fish kill.
P.C.B. Rule 203 requires waters to be free “from unnatural sludge or bottom deposits, floating debris, visible oil, odor, unnatural plant or algal growth, unnatural color and turbidity, or matter in concentrations or combinations toxic or harmful to human, animal, plant or aquatic life of other than natural origin.” The Craigs contend that in the absence of a finding that they caused the fish kill, the record does not support a finding that the manure that they permitted to flow into the river was shown to be “toxic or harmful to human, animal, plant or aquatic fife.” We are not so persuaded, however, and conclude that evidence abounds that the lowered oxygen count and increased bacteria count of the water made the water toxic and harmful even though we are not satisfied as to the proof that the resultant condition of the water killed the fish.
We see no merit to the Craigs’ other contentions. Any error in admission of exhibits would have been prejudicial only on the question of the fish kill. The evidence showed that some manure continued to flow into the stream. The cease and desist order was proper.
The portion of the P.C.B. order finding the Craigs responsible for the fish kill and requiring them to pay for the same is reversed. The balance of the order including the imposition of the $500 fine is affirmed.
Affirmed in part and reversed in part.