LEVY, Judge.

Appellants/Plaintiffs, Claire J. and Phillip Sidran, appeal from a Final Judgment of no liability in favor of the Appellees/Defendants, and from the trial court's denial of their Motion for New Trial. We reverse.

In 1973, Mr. and Mrs. Sidran converted Mrs. Sidran's orchid-raising hobby into a business. Between 1980 and 1985, they converted the business into a partnership, and formed a corporation in 1992. The business consisted of a small nursery located behind the Sidrans' home in the Suniland neighborhood of Miami-Dade County, Florida. By 1990, the Sidrans had five or six greenhouses on their property, and thousands of orchids in their inventory. In order to properly grow and cultivate their orchids, the Sidrans applied certain fungicides to the plants.

In the early 1980s, Mrs. Sidran applied Benlate WP, a product manufactured by Appellant/Defendant E.I. DuPont De Nemours Co. (DuPont), to her plants. When she ordered Benlate WP in the late 1980's, she received Benlate 50 DF ("Benlate DF"), the chemical at issue in the instant case. The Sidrans' supplier advised the Sidrans that Benlate DF was the new form of Benlate that was being produced by DuPont. The Sidrans spray records indicated that they used Benlate DF from 1988 through February 1991. After discontinuing their use of Benlate DF in February 1991, the Sidrans switched to a fungicide called Kocide.

According to Mrs. Sidran's testimony at trial, she noticed a change in her plants in 1989. The plants did not appear to be healthy, they were no longer growing, and there was chlorosis of the leaves. In early 1989, she sent several plants to the Florida Department of Agriculture and Consumer Services for testing. The reports issued by the Department of Agriculture and Consumer Services in May 1989 inquired whether chemical application processes were ruled out as the cause.

In May 1991, the Sidrans received a letter from DuPont which stated that Benlate 50 DF had been recalled. In June 1991, DuPont set up a claims settlement process. After attempts to settle the Sidrans' claim proved to be unsuccessful, the Sidrans initiated this litigation. This case was originally tried from March 15, 1995 to May 1, 1995, and the jury returned a verdict in favor of the Sidrans. However, eight days after the jury verdict, the trial court granted the Defendants' motion for a new trial. The retrial, which is the subject of this appeal, began on May 21, 2001.

At the retrial, the Defendants denied that Benlate DF was defective and denied that the product was the cause of the Sidrans' lost inventory. They presented two possible alternative causes — that the Sidrans' well water was contaminated with powerful organic solvents that were present in the septic tanks of dry cleaners located upstream from their wells; and that the Sidrans misused Kocide before reporting the widespread damage symptoms.

The testimony at trial revealed that the residents of the Suniland area used underground wells for their water. The Sidrans watered their orchids from 1987 through January 1992 with the ground water from their wells. In 1991, state and local authorities became aware of contamination in the ground water in the Suniland area. Further testing showed that there were various levels of contamination in the Suniland residents' ground water wells. At the Sidrans' home, testing for contaminants yielded negative tests for the presence of contamination on September 20, 1991, November 25, 1991, and January 29, 1992; and a positive test for the presence of contamination on November 5, 1991. The level of contamination discovered on November 5, 1991 was 1.7 micrograms per liter. Tests performed on the water of Ethyl Knapp, the Sidrans' neighbor, returned positive on October 28, 1991 and negative on February 5, 1992.

According to Walter Livingstone of the Miami-Dade County Health Department, the contaminant level at the Sidrans' home was high enough that he would not continuously drink water from the Sidrans' wells. Significantly, however, Livingstone testified that the purpose of the contamination investigation was for the benefit of the public health and not the health of plants, and that he had no knowledge of the effect of the contaminants on orchids. Harvey Kottke, a chemical engineer and Chief of Water and Wastewater Treatment for the Department of Environmental Resources Management (DERM), testified that the testing of the ground water showed contamination originating from dry cleaners, that were located upstream from the Suniland home wells. Mr. Kottke testified that the levels of contamination found in the wells changed on a daily basis, depending on the ebb and flow of the ground water. Kottke also testified that the Sidrans' contamination limit was considered safe for drinking purposes. Ricardo Fraxedas, an environmental engineer who worked for DERM during the Suniland contamination, testified that there were pulsations of contamination occurring in the Suniland area between 1989 and 1991. Fraxedas further testified that different properties in the area had varying test results on different days, and that the Sidrans' property was located in the middle of the plume of contamination. Like Livingstone, Fraxedas did not give an opinion as to whether the water was toxic to plants or humans. The Defendants also presented the videotaped testimony of Ethyl Knapp, who testified that she did not use Benlate DF but experienced similar problems with her orchids as the problems experienced by the Sidrans. Finally, Dr. Harold Coble testified that damage to orchid plants was more likely to occur in response to repeated applications of water contaminated with organic solvents than in response to Benlate DF as the level of exposure increases. Dr. Coble also testified that he never did any testing himself on the effect on plants of the dry cleaning contaminants dissolved in water.

At the conclusion of the evidence, the Plaintiffs moved to strike the defense of water contamination and moved for a directed verdict as to that issue. The trial court ruled that whether or not the contamination in the ground water was a cause of the Sidrans' orchid damage was a fact issue. The jury returned a verdict in favor of the Defendants. This appeal follows.

The admission of evidence is a matter within the sound discretion of the trial court. See Jimenez v. Gulf Western Manufacturing Co., 458 So.2d 58, 59 (Fla. 3d DCA 1984). The trial court's decision in that regard must be viewed in the context of the trial as a whole. See id. Relevant evidence is evidence which tends to prove or disprove a material fact. See § 90.401, Fla. Stat. (2001).

That brings us to the testimony of Dr. Coble, the Defendants' expert witness. The Sidrans contend that the testimony of Dr. Coble, that "the higher the level the contaminant, the more likely damage would occur", had no basis in fact or law and did not relate to the facts of the instant case. We agree. In order to admit expert testimony, the trial court should determine that the expert testimony will assist the trier of fact "in understanding the evidence or in determining a fact in issue." § 90.702, Fla. Stat. (1999). Moreover, an expert opinion is inadmissible when it is apparent that the opinion is based on insufficient data. See Young-Chin v. City of Homestead, 597 So.2d 879, 882 (Fla. 3d DCA 1992). In the instant case, Dr. Coble's testimony was not even based upon insufficient data — it was based upon no data at all. The testimony at trial established the Sidrans' water contamination level at 1.7 micrograms per liter. Dr. Coble, who testified that the higher level of contaminant, the more likely that damage would occur, performed no experiment and could cite to no study wherein the dry cleaning contaminant could be harmful to the orchids at 1.7 micrograms per liter. In the absence of such a predicate, admission of the water contamination testimony, while probative to some degree, was substantially outweighed by the potential of misleading the jury.

Accordingly, we conclude that the trial court abused its discretion in admitting the evidence regarding well water contamination absent any proper scientific predicate or evidence that it related, in fact, to the damages that the appellants claim was done to their orchids. This cause is reversed and remanded for a new trial.

Reversed and remanded.

LEVY, Judge.

On Rehearing Denied

PER CURIAM.

Dupont has moved for rehearing, contending that we misunderstood theories of defense and applied incorrect legal standards. We deny rehearing, but clarify our reasons for reversing the judgment.

The Sidrans used Benlate DF from 1988 through February 1991. As explained in the initial opinion, the plants' health declined during this period of time.

In May 1991, the Sidrans received a notice from Dupont indicating that Benlate DF had been recalled. The recall notice stated, in part:

The recall was initiated because trace levels of atrazine herbicide were found, by routine analysis, in some production batches. Through our investigations of reports of plant damage, we detected effects that we do not fully understand.

From the plant symptoms we have seen and our extensive analytical investigations, we believe the cause must be a herbicide at extremely low levels.

The plants generally show reduced vigor or stunted growth. If the herbicide was present at higher levels, it would be easily detectable and the plants would show more severe symptoms.

The symptoms suffered by the Sidrans' orchids corresponded to those outlined in the Dupont recall notice.

Dupont set up a claims process. The Sidrans' claim was not resolved. The Sidrans brought suit, alleging that Benlate DF was a defective product and Dupont was negligent. The Sidrans contended that Benlate had caused the loss of their orchids.

Dupont denied that there was any product defect and denied that it had been negligent. Dupont also offered evidence that the actual cause of the damage to the Sidrans' orchids lay elsewhere. Specifically, Dupont introduced evidence which it claimed would demonstrate that the orchid damage was caused by (a) groundwater contamination in the Suniland area, or (b) the Sidrans' misuse of another fungicide, Kocide, or (c) both. Dupont is, of course, allowed to defend in this way, denying any negligence or product defect and asserting that the actual cause of the damage to the orchids was the fault of something else — groundwater contamination or misuse of a different product. See Haas v. Zaccaria, 659 So.2d 1130, 1132 (Fla. 4th DCA 1995); Carnival Cruise Lines, Inc. v. Levalley, 786 So.2d 18, 19 (Fla. 3d DCA 2001).

Regarding the issue of groundwater contamination, Dupont offered evidence at trial that the groundwater in the Suniland area had been contaminated by improper disposal of dry cleaning materials from a dry cleaning establishment. The chemicals had seeped into the groundwater and traveled in a plume of contamination moving from upstream to downstream in the flow of subterranean groundwater. According to the testimony it was reasonable to assume that when contamination was detected in the wells of the neighbors across the street, the Knapps, a comparable level of contamination likely also existed in the Sidrans' wells.

The Knapps' wells tested negative for contamination on February 5, 1991. The Sidrans' three wells were negative on September 20, 1991. Five weeks later, on October 28, 1991, the Knapps had a positive test of 8.0 parts per billion of percholoroethylene. A few days later, on November 5, 1991, there was a positive test of 1.7 parts per billion in two of the Sidrans' three wells. Soon thereafter, on November 25, 1991, and January 29, 1992, the Sidrans' wells tested negative for contamination.

Based on this testimony and the testimony of the experts cited in our original opinion, Dupont maintained that the level of contamination rose and fell so that the orchids were watered with contaminated water during times relevant here.

The obvious question, however, was whether the chemicals involved here would have an adverse effect on orchids at the observed level of 1.7 parts per billion on the Sidrans' property or the higher level of 8.0 parts per billion detected on the Knapps' property.

At trial, Dupont offered the testimony of Dr. Harold Coble, an agronomist employed by the United States Department of Agriculture. Dr. Coble was accepted as an expert and introduced as such to the jury. Dr. Coble had performed Benlate testing for Dupont. As stated earlier, Dupont's recall notice indicated Dupont's belief that the problems associated with Benlate DF were attributable to contamination of Benlate DF with atrazine. Dr. Coble performed tests on orchids to determine if atrazine could damage orchids, and if so, at what levels. Dr. Coble found no damage to orchids, even at three times the highest level of atrazine found in a Benlate box.

While Dr. Coble was on the stand, Dupont sought to ask Dr. Coble about the effect on orchids if they were irrigated with water contaminated with organic solvents. The Sidrans timely objected because Dr. Coble had performed no studies, and was aware of none, which would establish that water contamination at the observed levels would cause damage to orchids. The objection was overruled, and Dupont asked:

Q . . . Doctor Coble, do you have an opinion, within a reasonable degree of scientific certainty, as to whether repeated applications of water contaminated with organic solvents would damage orchid plants?

A. Yes.

Q. What is that opinion, sir?

A. If water contaminated with organic solvents were repeatedly used to water plants, that certainly would have an [e]ffect on the plant cuticle that we talked about yesterday. Obviously the higher the level —

[Plaintiff's Counsel]: Your Honor, let me object to anything further. That is not responsive to the specific question.

The Court: Overruled, go ahead.

The Witness: Obviously the higher the level [of] the contaminant, the more likely the damage would occur.

TR. 3275 (emphasis added).

This testimony clearly told the jury that the water contamination in Suniland was of sufficient magnitude to cause injury to the orchids. However, Dr. Coble had conducted no studies showing this to be the case at any of the observed levels. Dupont never introduced any such studies in the proceedings below.

Dupont's failure in this regard stands in marked contrast to Dupont's careful preparation of scientific testing regarding atrazine contamination in Benlate. As relates to atrazine, Dupont and Dr. Coble made the point that the mere presence of atrazine contamination was not enough to cause plant damage. Dr. Coble testified that (based on testing he conducted) the level of atrazine found in Benlate was not high enough to cause the observed harm to plants.

Exactly the same question applies when considering whether the observed level of water contamination could create the injury sustained by the Sidrans' orchids. Since there was no proper scientific basis for Dr. Coble's testimony on this point, it should have been excluded.

Dupont argues that this error was harmless. We cannot accept that contention. The Sidrans objected not only to Dr. Coble's testimony but also moved to strike the entire water contamination defense. Dr. Coble was introduced to the jury not only as an expert, but as an agronomist from the United States Department of Agriculture, which carries with it a particular aura of credibility. The cited testimony was the main support for Dupont's water contamination defense, but there simply was no scientific basis for it. Clearly the testimony was harmful.

In support of its position, Dupont points to the testimony of the Sidrans' neighbor, Ethyl Knapp, who was also a successful orchid grower. Ms. Knapp testified by video deposition that her orchids likewise had serious problems during this time period. She maintained in her videotaped deposition that she never used Benlate. She testified that the more she watered her orchids, worse they did. She attributed this problem to the groundwater contamination.

Ms. Knapp's testimony was impeached, however, because witness D'Armand Hull, a University of Florida agricultural agent in South Florida, testified that Ms. Knapp had Benlate on her premises and that her husband Charlie Knapp had applied it to the plants. Plaintiff Jill Sidran testified that she saw Ms. Knapp use Benlate. The Sidrans proffered that Ms. Knapp and her husband were at the time of her testimony plaintiffs in a pending lawsuit brought by Suniland residents against the dry cleaning establishment which had allegedly leaked the dry cleaning chemicals into the ground. In addition, Ms. Knapp's testimony suffered from the same deficiency as that of Dr. Coble: it reveals no scientific basis for believing that water contamination at the tested levels would be harmful to orchids, and as already stated, there was no other expert testimony in the case to establish that point.

For the stated reasons, we adhere to our earlier opinion of reversal and remand for a new trial.