BLACKMAR, Judge.

Plaintiff Missouri Farmers Association (MFA) sued on an account and some notes. Defendant Kempker filed a counterclaim alleging that plaintiff had supplied him with defective feed. The trial court entered judgment for MFA on its claim, and for Kempker on the counterclaim for $20,000. Kempker alone appealed, claiming that the trial court improperly excluded evidence of loss of milk production and of the loss of calves through failure of cows to conceive. The court of appeals reversed on the authority of Hoover's Dairy, Inc. v. Mid-America Dairymen, Inc., 700 S.W.2d 426 (Mo. banc 1985), which it read as allowing damages for loss of milk production and calves, in addition to diminution of value of the individual cows. We granted transfer to consider the proper application of Hoover's Dairy, and, taking the case as on original appeal, affirm the judgment because Kempker has not laid a proper foundation for the admission of the excluded evidence. In reaching our conclusion on the only issue presented we consider the evidence from Kempker's point of view, and give no attention to MFA's witnesses.

Kempker had operated a dairy farm for many years, with approximately 60 milking cows. In 1980 he decided to expand to about 200 head, and purchased pregnant and recently freshened heifers for the expansion.

In April of 1981 he began feeding the milk cows RPS-16 feed, supplied by MFA, in addition to corn and wheat silage and alfalfa produced on his farm. The use of RPS-16 was discontinued in February of 1982, after milk production was found to be disappointing.

A single sample of RPS-16 feed preserved by Kempker was found on analysis to contain approximately 1% urea, which was not an ingredient listed on the label. The time of taking of the sample is not shown. There is a disagreement among authorities as to whether urea is a harmful element in cattle feed. Kempker produced the testimony of Dr. Hersel Robertson, an expert in veterinary nutrition, who expressed the opinion that it should not be used for feeding lactating cattle. The evidence of causation of damages is scanty, but Kempker had a verdict on the counterclaim for $20,000 and MFA did not appeal from this verdict.

Kempker's evidence showed that the cows which had been fed RPS-16 showed symptoms of malnutrition and that their milk production declined. Problems with these cows continued so long as they remained with the herd. Twenty-five cows were sold off in 1982, 48 in 1983, 78 in 1984 and 7 in 1985, because they were no longer useful for milk production. The purchase price of these cows was $144,260 and the selling price $65,886, indicating a claimed loss in value of $78,374. Ten cows, of a value of $1,400 each, died. (There is no evidence as to the cause of death). Kempker was also allowed to prove without objection that he lost 189,807 pounds of milk production in 1981 and 179,468 pounds of production of a value of $210,000 in 1982.

His sole complaint on appeal is that he was not allowed to introduce evidence of lost calf production after 1982 amounting to $10,000, and was not allowed to show loss of milk production in 1983 in the amount of $204,037, in 1984 in the amount of $253,215, and in 1985 in the amount of $60,790. He made an offer to prove these figures, conceding that, because of the fewer number of milk cows, expenses would have been reduced by $52,888 in 1983, and $92,553 in 1984.

MFA argues that the measure of damages for injury to a milk cow is the difference in the value immediately before and immediately after the injury, citing several cases. It goes on to argue that the offered and rejected evidence would permit the duplication of damage, to the extent that it allows recovery for the loss of value of a cow and also for milk and calf production after that same cow has been sold or disposed of. Kempker argues that, under Hoover's Dairy, he may recover both for diminution in value, and for loss of milk and calf production.

We do not accept MFA's broad contention. The rule it cites would be appropriate for injury to a mule or a dog. When cows in a milk producing herd are injured in such a way that their production of calves and milk is interfered with, there may be recovery for loss of milk and calf production until a cow is sold or disposed of, and for loss of subsequent production until a satisfactory replacement cow could be obtained. See Divelbiss v. Phillips Petroleum Co., 272 S.W.2d 839 (Mo.App. 1954). The proprietor need not liquidate a cow simply because her milk production is declining, or has stopped. He may wait to see whether there will be improvement. Any questions as to claimant's having taken reasonable steps to mitigate damages are for the jury. The limiting circumstance is that there may be no recovery for future milk and calf production of a cow which has been disposed of, after a replacement of comparable capacity has been or could have been acquired. State v. Morison, 148 Colo. 79, 365 P.2d 266, 272-73 (1961); Snyder v. Bio-Lab, Inc., 94 Misc. 816, 405 N.Y.S.2d 596 (N.Y.Sup.Ct. 1978); Kintner v. Claverack Rural Electric Co-operative, Inc., 329 Pa. Super. 417, 478 A.2d 858, 861-62 (1984).

Our holding on this proposition does not conclude the case. Kempker is entitled to reversal only if he can demonstrate error prejudicial to his rights. When the alleged error consists of the rejection of evidence, the appellant is required to make an offer of proof showing that the rejected evidence is admissible and that any required foundation has been laid. We will not reverse unless persuaded that the trial court erred in rejecting the evidence. We are not confined to the objections made. The trial court's action will be upheld if there is any recognized ground on which the trial judge could have rejected the evidence.

Tested by this standard, Kempker has failed to demonstrate error. He made a lengthy offer to prove additional items of damage, asserting that the figures adduced were supported by his records. The problem, however, is not with the accuracy of the figures but with their relevance and materiality. The appellant must show that the figures represent damages attributable to the feed furnished by MFA which the jury found to be defective. The trial judge is charged with scrutinizing the evidence for relevancy and materiality. Without a substantial evidentiary linkage between the feed and the damage there is no basis for admission.

The first part of the offer of proof sought $10,000 in damages attributed to loss of calf production during the first recycling of the newly purchased heifers. The only testimony which could possibly support a connection between feed and reproduction was that of Dr. Hersel Robertson. He does not recommend that urea be included in a ration for dairy cows because it is a non-protein nitrogen which the cow converts into protein by expending energy, which might otherwise be used in milk production and in building up resistance to disease. If a cow's diet does not provide adequate energy, there can be problems in reproduction. The doctor was not asked to, and did not, express an opinion that the ingestion of RPS-16 feed caused a decrease in the reproduction rate of Kempker's cows. Nor did he express an opinion as to when normal reproduction would resume, after correction of any diet deficiencies.

Dr. Dennis Markway, Kempker's regular veterinarian, was called by Kempker to examine the cows in August of 1981. Some cows were not coming into heat regularly. Some of them were thin but others appeared healthy. He initially attributed the problems to a disease known as bovine viral diarrhea (BVD), and recommended vaccination for this condition. There was also a high incidence of mastitis, which can also affect reproduction. Markway was unable to point to urea as a causal agent in the diminished rate of reproduction. He said that respected authorities on animal nutrition believed that urea was a proper ingredient in the ration of lactating cattle, and that a level of one percent in feed was not improper.

Kempker has failed to lay a sufficient evidentiary foundation for the admission of his computation of reproductive losses. Neither of his experts expressed an opinion on causation, in terms of reasonable veterinary certainty or even reasonable probability. The evidence fails to exclude other possible causative factors, or to provide any basis for apportionment among the several hypotheses. Without such basis the evidence is speculative. See Kane v. Chicago, B. Q. R.R. Co., 271 S.W.2d 518, 521 (Mo. 1954), cert. denied 348 U.S. 943, 75 S.Ct. 365, 99 L.Ed. 738 (1955); Cato v. Modglin, 545 S.W.2d 307, 311 (Mo.App. 1977). Nor is there a scintilla of evidence that problems with a feed which was abandoned early in 1982 could affect calf production in 1983 and later years. Because of these deficiencies, the trial court did not err in rejecting this portion of the offer of proof.

The balance of the offer of proof related to diminished milk production in 1983, 1984, and 1985. Evidence of loss of production in pounds for 1981 and 1982, and for dollar value of loss of production in 1982, had previously been received. The record is barren of evidence that deficiencies in a feed ration which was not fed after February of 1982 could have any effect on milk production ten months later. Dr. Robertson testified that it would take a "considerable length of time" for milk production to start back up after undernourished cows were restored to an adequate feed ration — "It didn't start up the same day." He gave no further particulars. Neither of Kempker's experts, furthermore, expressed any opinion going beyond suggestion that deficiencies in the feed supplied by MFA caused any decline in milk production, at any time. Because of these inadequacies in the record, there was no error in rejecting the proffered evidence of loss of future production.

Hoover's Dairy does not mandate a contrary result. That case did not deal with admissibility of evidence of damages. It simply held that remittitur was not available at the appellate level to correct a verdict which was claimed to be excessive because it was based on duplicating items. There was no purpose of changing the law governing damages. It was not intended to hold that there could be recovery of the full loss of value of a cow and also for milk and calf production after that cow could have been replaced.

We need not discuss other inadequacies in the offer at length. Kempker admitted that some of the cows whose production is covered in the offered figures had been sold or had died, but made no effort to show that replacements were not available on the market. He was not experienced in the operation of a 200 milking-cow facility, and the evidence of loss of profits, which was based on the projected production of a herd of this size, is speculative. There is also a question whether a single sampling justifies the conclusion that all of the RPS-16 supplied to Kempker contained urea. Under the totality of the record, we are unable to say that the trial judge erred.

It is also appropriate to consider the jury's verdict of $20,000. Kempker was allowed to introduce evidence of items of claimed damage in an amount exceeding $200,000. The jury necessarily concluded that only a small proportion of his total losses were attributable to the feed supplied by MFA. It is highly doubtful that the award would have been higher if more remote items of damage had been admitted. Counsel complains that he was not permitted to argue loss of production, but admits that nothing in the record shows any limitation of what might be argued, and asserts that the limiting directions were given during an off-the-record conference in chambers. We look only to the record. Counsel who want to preserve claims of error must make sure that any rulings or directions about which they complain are transcribed so that they may be included in the record on appeal.

We reverse only for demonstrated error prejudicial to the appealing party. This record shows none, and the judgment is affirmed.

All concur.