ON REMAND FROM THE SUPREME COURT OF FLORIDA
ALTENBERND, Judge.Upon remand from the Florida Supreme Court, we reconsider Leo Richard Be-rube’s conviction for second-degree murder in light of the subsequent decision in Daniels v. State, 121 So.3d 409 (Fla.2013). Because the giving of the erroneous instruction on the lesser-included offense of manslaughter by act unquestionably did not pertain to a disputed element of that offense, we conclude that Mr. Berube has not established that the error in the instruction was harmful. Having failed to establish this first prong of a fundamental error analysis, he is not entitled to any relief. Thus, we once again affirm the judgment and sentence on appeal.
L THE PRIOR PROCEEDINGS IN THIS CASE
In 2006, Mr. Berube was convicted of first-degree murder for the strangulation *1167of a prostitute with whom he admitted having a sexual encounter on the evening of her death in 2003. He appealed his judgment and sentence of life imprisonment to this court.
In Mr. Berube’s first appeal, this court wrote a lengthy opinion detailing the facts of the case. Berube v. State, 5 So.3d 734 (Fla. 2d DCA 2009) (Berube I). To avoid repetition, we rely on that recitation of the facts and provide supplemental factual information only as needed in this opinion. In Berube I, this court held that the evidence was sufficient to support a prima facie case of premeditated murder. Id. at 745. However, we reversed and remanded for a new trial.
The new trial was required because the trial court erroneously admitted some Williams1 rule evidence. Id. at 736. Pri- or to trial, the State had sought leave to introduce into evidence two prior instances of alleged rape by Mr. Berube and an unsolved murder for which Mr. Berube was a suspect. Id. at 739. The trial court denied the State’s request to admit evidence of the unsolved murder — a homicide that occurred shortly after the victim had engaged in sexual intercourse with Mr. Berube. But the trial court granted the State’s request to admit evidence of the two instances of alleged rape. Thereafter, that evidence was admitted during the trial. In Berube I, we held that the trial court erred in admitting this evidence and that the error required a new trial. Id. at 745.
In the second trial, Mr. Berube was convicted of second-degree murder. On appeal, he argued that the instruction concerning intent given on the next-lesser offense of manslaughter was fundamentally erroneous for the reasons discussed in the First District’s opinion in Montgomery v. State, 70 So.3d 603 (Fla. 1st DCA 2009). This court concluded that the instruction did not constitute fundamental error. Berube v. State, 84 So.3d 436, 436 (Fla. 2d DCA 2012) (Berube II).
On review, the supreme court quashed our decision in Berube II and remanded the case for reconsideration in light of the supreme court’s decision in Daniels, 121 So.3d 409. See Berube v. State, 137 So.3d 1019 (Fla.2014).
II. THE STANDARD OF REVIEW
This case is very similar to Nieves v. State, 144 So.3d 649 (Fla. 2d DCA 2014). Likewise it is similar to Griffin v. State, 128 So.3d 88, 90 (Fla. 2d DCA 2013) (“Because there was no dispute regarding the element of intent, the erroneous jury instruction on the intent element of the lesser included offense of manslaughter did not constitute fundamental error.”), review granted, 143 So.3d 918 (Fla.2014). We reach a similar outcome for similar reasons.
That said, we have considered what legal standard should be applied to our reconsideration. It seems that the supreme court, in its reliance on dicta from Pena v. State, 901 So.2d 781, 787 (Fla.2005), in both State v. Montgomery, 39 So.3d 252, 259 (Fla.2010) (Montgomery II), and Daniels, 121 So.3d at 415 & n. 5, implies that this error — an error in the instruction for a lesser-included offense that is one step removed from the conviction — should be treated as a “per se” error for which this court should conduct no examination of the record to determine harmfulness. We do not believe that is the supreme court’s intent. Thus, we describe with some care the standard of review and the analysis that we have used to decide this case on remand. In section III, we explain why *1168we believe the dicta from Pena should not govern the question of whether this erroneous instruction entitles Mr. Berube to a third trial.
Under the shift in the analysis for fundamental error that was announced in Reed v. State, 837 So.2d 366, 369-70 (Fla.2002), the defendant has the burden of persuasion on direct appeal to establish two requirements. First, the defendant must show that an unpreserved error is harmful or prejudicial. If that is demonstrated, the defendant must then establish that the error reaches the level of a fundamental error. As we read Montgomery II, Daniels, and Haygood v. State, 109 So.3d 735 (Fla.2013), it is clear that the supreme court has already determined the second requirement as a matter of law; an error in the instruction on intent for the offense of manslaughter in the case of a conviction for second-degree murder reaches the level of a fundamental error if that error is actually harmful. Thus, to prevail in this appeal, Mr. Berube is required to establish only that the error was harmful or prejudicial. See Reed, 837 So.2d at 370; see also Haygood, 109 So.3d at 741 (“[Fjundamen-tal error occurs in a jury instruction where the instruction pertains to a disputed element of the offense and the error is pertinent or material to what the jury must consider to convict.”).
In the context of this issue, the key question seems to be: Putting the elements of second-degree murder aside and assuming that the jury gave serious consideration to the lesser offense of manslaughter, has Mr. Berube demonstrated there was a dispute either of fact or of the jury’s application of fact to law that involved the instruction explaining the intent required where the offense of manslaughter is committed by act? At least in the usual case, to establish this burden the defendant must show that, the issue of intent was “a material element that [was] disputed at trial.” Daniels, 121 So.3d at 418. We believe that the defendant can either show that this element of intent was an issue in dispute in the case on appeal or that, in a new trial with correct instructions, it would become an issue in dispute.
Because harmfulness is the first prong of the fundamental error analysis, if ' the defendant establishes both prongs of the fundamental error test, we do not need to further subject the fundamental error to any of the tests for harmfulness or harmlessness used in cases of preserved error; such a consideration simply would be redundant. The fact that the error occurred in the next-lesser offense, making it one step removed from the offense of conviction, does not invoke any concept of per se reversible error when the error is unpre-served. Instead, it is a circumstance that gives Mr. Berube the right to require this court to consider in this decision-making process, as a matter of logic, that the jury might have given serious consideration to this lesser offense during deliberations. The fact that it is one step removed contributes to the decision to categorize the error as one reaching the level of a fundamental error; it does not alter the requirement that Mr. Berube first establish that the error was harmful.
Although we believe that the law requires Mr. Berube to demonstrate from the record a reasonable probability2 of *1169such harm before this court is authorized to declare that the error is fundamental, in this case we have carefully reviewed the record and conclude beyond a reasonable doubt that the giving of this instruction was not harmful. Accordingly, we conclude that the error in the manslaughter by act instruction is not a fundamental error in this case.
III. PENA’S RELIANCE ON ABREAU IN A POST -REED FUNDAMENTAL ERROR ANALYSIS
In order to understand our concern with the portion of Pena that is cited in Montgomery II and Daniels, it is necessary to review a little legal history. In State v. Abreau, 363 So.2d 1063, 1064 (Fla.1978), the supreme court held:
Only the failure to instruct on the next immediate lesser-included offense (one step removed) constitutes error that is per se reversible. Where the omitted instruction relates to an offense two or more steps removed, DeLaine [v. State, 262 So.2d 655 (Fla.1972),] continues to have vitality, and reviewing courts may properly find such error to be harmless.
This holding occurred in the context of a preserved error. The trial court had completely omitted the requested instruction. Because the error did not involve the next immediate lesser-included offense, the supreme court quashed the district court’s reversal, and the district court affirmed the judgment on remand. See Abreau v. State, 365 So.2d 201 (Fla. 3d DCA 1978). Likewise, in Rojas v. State, 552 So.2d 914, 916 & n. 1 (Fla.1989), the court applied Abreau where the mandatory instruction of justifiable and excusable homicide was not given and the error was preserved.
The “per se reversible” error described in Abreau is an error that is not tested for harmlessness or harmfulness. Essentially, it is an error for which the extent or degree of the resulting harm cannot be determined from the record. Accordingly, the party on appeal with the burden to prove that such an error is harmful or harmless can never do so from the record. Because the State is the party with that burden in a direct criminal appeal,3 per se errors result in a reversal against the State without any review of the record for harm. See, e.g., Hazuri v. State, 91 So.3d 836, 846-47 (Fla.2012). Thus, after Abreau, the total omission of an instruction on the next-lesser offense always required a reversal of a judgment and a new trial so long as the issue was preserved.
In 2005, the supreme court considered two unpreserved issues for potential fundamental errors in Pena, 901 So.2d 781. It held that the omission of an instruction concerning the age of the defendant for an offense that required the defendant to be eighteen years of age or older and the omission of that same allegation from the indictment were not fundamental errors when the defendant was twenty-eight *1170years old and his age was not a matter of dispute. Id. at 785-86. Concerning the omission of an instruction on excusable and justifiable homicide, the court determined that the error occurred in an instruction for a lesser offense that was more than two steps removed from the conviction. As a result, the supreme court correctly did not regard the error as fundamental. Id. at 787.
But in its reasoning in Pena, the supreme court referred back to Abreau and Rojas. It stated:
The lesser offense of manslaughter was three steps removed from the conviction of first-degree murder. Thus, because the lesser offense in this case was more than two steps removed from the conviction of first-degree murder, under Rojas. and Abreau, the district court properly conducted a harmless error analysis.
Pena, 901 So.2d at 787. It further explained:
If the jury is not properly instructed on the next lower crime, then it is impossible to determine whether, having been properly instructed, it would have found the defendant guilty of the next lesser offense. However, when the trial court fails to properly instruct on a crime two or more degrees removed from the crime for which the defendant is convicted, the error is not per se reversible, but instead is subject to a harmless error analysis.
Id. The holding in Pena seems completely correct, but the above-quoted analysis, which is dicta, appears to be flawed. Pena involved an unpreserved error. Thus, Abreau and Rojas applied at most by comparison or analogy in Pena. As we explain next, this dicta in Pena did not take into account the new approach adopted for fundamental error in Reed.
In State v. Clark, 614 So.2d 453, 454 (Fla.1992), receded from in Reed, 837 So.2d at 370 n. 3,4 the supreme court held that an unpreserved error involving the use of a deposition at trial where the defendant had not had the opportunity of confrontation at the taking of the deposition was a fundamental error. It then proceeded to decide that the test for harmless error established in State v. DiGuilio, 491 So.2d 1129 (Fla.1986), should apply to fundamental error, concluding that the State had not established that the reading of the deposition was harmless. Clark, 614 So.2d at 454-55. Although Clark involved a case in which the trial court was reversed, its reasoning led to cases where the district courts affirmed matters of fundamental error because they were “harmless.” See, e.g., Mincey v. State, 684 So.2d 236, 238-39 (Fla. 1st DCA 1996); Sigmon v. State, 622 So.2d 57, 59-60 (Fla. 1st DCA 1993).
Clark quickly engendered issues relating to the review of fundamental errors that would be treated as per se errors if these errors had been preserved. In State v. Lucas, 645 So.2d 425, 426-27 (Fla.1994), the supreme court wrote a short opinion in which it answered the following certified question in the affirmative without any qualification:
When a defendant has been convicted of either manslaughter or a greater offense not more than one step removed, does failure to explain justifiable'and excusable homicide as part of the manslaughter instruction always constitute both “fundamental” and per se reversible error, which may be raised for the first time on appeal and may not be subjected to a *1171harmless-error analysis, regardless of whether the evidence could support a finding of either justifiable or excusable homicide?
Id. at 426.
Simply put, the concept of a fundamental per se error, which required reversal in the absence of any demonstration of harm, did not pass the test of time. The supreme court receded from Clark in Reed. There, the court held:
Furthermore, we take this occasion to clarify that fundamental error is not subject to harmless error review. By its very nature, fundamental error has to be considered harmful. If the error was not harmful, it would not meet our requirement for being fundamental.
Reed, 887 So.2d at 370 (footnote omitted).5
Although this sentence can perhaps be read to suggest that fundamental error
is inherently harmful, that does not seem to be the supreme court’s intent. Instead, in Reed, the supreme court shifted the analysis of fundamental error. Before deciding whether an error rises to the level described in one of the various definitions of this rare error, the reviewing court should first determine whether the error was harmful.6 Because the error is unpre-served, the defendant does not receive the benefit of the harmless error analysis established in DiGuilio. Rather, in the case of fundamental error on direct appeal, the defendant has the initial burden of persuasion to show that there is a reasonable probability that the error was harmful or “prejudicial.” § 924.051(l)(a), Fla. Stat. (2002); Sampson v. State, 903 So.2d 1055, 1056-57 (Fla. 2d DCA 2005). If it is, the defendant must then’show that the error is fundamental.7 This new analysis is dem*1172onstrated, for example, in State v. Weaver, 957 So.2d 586, 589 (Fla.2007) (disapproving Vega v. State, 900 So.2d 572 (Fla. 2d DCA 2004)). See also Martinez v. State, 981 So.2d 449, 455-57 (Fla.2008); Sampson, 903 So.2d at 1056-57.
Thus, returning to the dicta in Pena, the district court in that case was not actually required to perform a traditional harmless error analysis under Abreau. For an un-preserved error concerning the omission of the justifiable or excusable homicide instruction, the defendant had to demonstrate that the error was prejudicial. Because the affected lesser offense was more than one step removed, the defendant was unable to demonstrate prejudice. See Pena, 901 So.2d at 784-85.
Although this rather technical misstatement in dicta was certainly of no import in Pena, the language was picked up in both Montgomery II and Daniels. In Montgomery II, the court states:
Second-degree murder as a lesser included offense is one step removed from first-degree murder, and manslaughter as a lesser included offense is two steps removed from first-degree murder. In Pena, we concluded that “when the trial court fails to properly instruct on a crime two or more degrees removed from the crime for which the defendant is convicted, the error is not per se reversible, but instead is subject to a harmless error analysis.” Pena, 901 So,2d at 787. We explained that '
the significance of the two-steps-removed requirement is more than merely a matter of number or degree.
A jury must be given a fair opportunity to exercise its inherent “pardon” power by returning a verdict of guilty as to the next lower crime. If the jury is not properly instructed on the next lower crime, then it is impossible to determine whether, having been properly instructed, it would have found the defendant guilty of the next lesser offense.
Id. at 787. The lesser included offense of manslaughter is just one step removed from second-degree murder. Because Montgomery’s conviction for second-degree murder was only one step removed from the necessarily lesser included offense of manslaughter, under Pena, fundamental error occurred in his case which was per se reversible where the manslaughter instruction erroneously imposed upon the jury a requirement to find that Montgomery intended to kill Ellis.
39 So.3d at 259.
It is likely that the defendant in Montgomery II demonstrated that the un-preserved error was prejudicial, i.e., that the faulty instruction pertained to an element disputed at trial, but that fact is not emphasized in the opinion. The above-quoted language appears to extend the dicta of Pena to become a holding that applies not only to the omission of an entire instruction on a lesser-included offense or a critical exception to an offense, but also to an unpreserved error in the description of one element of such an offense. This discussion is repeated in Daniels, 121 So.3d at 415 & n. 5.8
*1173We do not believe that the supreme court actually intends for this court to use an analysis that declares the erroneous instruction to be a fundamental error subject to automatic reversal as a per se error merely because the instruction contained a bad element and is the instruction for the next-lesser offense. As explained above, we believe that Reed applies and requires a defendant to make a showing by a reasonable probability that the error in the lesser instruction was prejudicial to him. Because the supreme court has already determined that this particular error fulfills the second requirement for fundamental error, Mr. Berube is only required to establish this first requirement. If we have misunderstood the supreme court’s mandate, then we are in conflict with Montgomery II and Daniels.
IV. DETERMINING WHETHER THE ERROR WAS HARMFUL IN THIS CASE
In this case, the victim was strangled with a lamp cord while she was naked on a bed in a motel room. Shortly after the homicide, the police investigated the scene. In addition to the woman’s body and the lamp cord, the police found a knife, a broken lamp shade, and a pair of glasses on the bed. The glasses ultimately were determined to be Mr. Berube’s glasses.
There was blood on the wall above the bed, on the lamp (the cord of which was used to strangle the victim), and on the exterior door jamb to the bathroom. Mr. Berube’s DNA matched the blood on the wall, the lamp, and some of the blood on the door jamb. His DNA was also found under the victim’s fingernails. A witness in an adjoining motel room had heard a loud bang on the wall and believed that he had seen a man matching Mr. Berube’s description leaving the room a short time after he heard the noise.
The medical examiner testified at length about the cause of death and opined that the strangulation process would have taken at least two to three minutes. The victim had five separate ligature marks on her neck, indicating that the cord was either wrapped five times around her neck or was tightened in five separate locations at five separate times during the strangulation. The victim’s hyoid bone was fractured, and she had other bruises on her body and fingernail marks near the ligature marks.
On cross-examination, Mr. Berube’s attorney did not significantly challenge the medical examiner’s opinion as to the cause of death or the condition of the victim’s body. Instead, he relied upon this evidence to create a theory of innocence for his client. He suggested that the husband had killed his wife after Mr. Berube left the room.
As described in Berube I, Mr. Berube had a sexual encounter with the victim shortly before her death. In an unusual twist of events, the victim’s husband was hiding in the bathroom of the motel room during this sexual encounter. In his video-recorded interview, Mr. Berube claimed that the victim had performed oral sex on him while he was sitting in a chair near the door. He fled when confronted by her *1174knife-wielding husband.9 Mr. Berube’s claim was that he left his glasses in the room because he left in a hurry. He surmised that his blood was on the walls as a result of the confrontation he had with the victim’s husband. He claimed the victim was fully clothed and sitting on the bed when he fled. He theorized that it was the victim’s husband who removed her clothing and strangled her on the bed. This theory was supported to some extent by evidence that the husband and wife had a turbulent marriage affected by their mutual addiction to cocaine.
In closing argument, Mr. Berube’s attorney argued that strangulation was a long process and that his client could not have killed the victim during the short time between the struggle that created the bang on the wall and the time the witness saw the man hurrying from the motel room. His theory was that the husband had strangled the victim in an argument occurring after Mr. Berube left the room. As he succinctly explained near the close of his argument:
And I suggest to you the reasonable doubt in this case arises from the conflict in the evidence. The conflict in the evidence is obviously [the victim’s husband] or Leo Berube. So if you look at [the victim’s husband’s] story and you look at Leo Berube’s story as told in the core of the video where he says he left and he ran and he fell and he broke his leg, you will see that there is plenty of doubt when compared to the Hintons, the next door neighbors, that say there was a thud and there was a short period of time before Leo Berube walked out of that room. From that conflict of evidence, based on the reasonable doubt that arises from that, you should find that Leo Berube is not guilty.
In Berube I, the opinion discusses Mr. Berube’s theory that the victim could have died of erotic asphyxiation rather than from a conscious purpose to cause her death. 5 So.3d at 743-44. As that opinion notes, there, was no evidence to support this theory at the trial. Id. Although this argument was made to the judge for purposes of a motion for judgment of acquittal, Mr. Berube’s attorney never argued this theory to the jury or suggested that the jury could reach any outcome in this case based on an accidental death. Id. at 742-44. In short, whether Mr. Berube intended to kill the victim was not in dispute and was not pertinent or material to what the jury needed to convict Mr. Be-rube of either second-degree murder or manslaughter. See Nieves, 144 So.3d at 651 (“Where the defendant concedes the facts of the crime, arguing only that it was committed by somebody else, no [ ] dispute [regarding intent] arises.”).
In this case, the jury received a correct instruction on manslaughter by culpable negligencé. Even if Mr. Berube had abandoned his story in the recorded statement — that he was a customer of a fully-clothed prostitute performing oral sex on him while he sat in a chair — and had chosen instead to argue that he had negligently strangled this woman while performing consensual acts of erotic asphyxiation with a lamp cord around her neck while lying on a bed, this version would have involved the manslaughter by culpable negligence instruction and not the manslaughter by act instruction. Cf. Haygood, 109 So.3d at 741-42 (holding that faulty manslaughter by act instruction was fundamental error even though correct instruction for man*1175slaughter by culpable negligence was given where intent to kill was a disputed issue at trial and evidence that defendant beat the victim supported manslaughter by act but not manslaughter by culpable negligence).
Moreover, the erroneous manslaughter instruction given in this case explained that the State must prove “Leo Richard Berube intentionally caused the death of [the victim].” The irrefutable physical evidence in this case was that the person who killed the victim strangled her with a lamp cord for a period of minutes while she struggled violently to survive. If anything, the error in the instruction, by including the concept of an intentional cause of death, increased the likelihood that a jury might consider the possibility of returning a verdict of manslaughter instead of second-degree murder, which does not require an intent to kill.
If this case were remanded for a third trial, the jury would be instructed on manslaughter with the current instruction. That instruction would explain that the State must prove that “Leo Richard Be-rube intentionally committed an act or acts that caused the death of [the victim].” Under the evidence in this case, we cannot conceive of any new argument that Mr. Berube’s attorney might make based on this instruction that was not already made in the second trial. We remain convinced beyond a reasonable doubt that the error in the manslaughter instruction was harmless in this ease and that the erroneous instruction did not pertain to a disputed element of the lesser-included offense of manslaughter by act.
Affirmed.
SILBERMAN and WALLACE, JJ„ Concur.. Williams v. State, 110 So.2d 654 (Fla.1959).
. We have selected the test of "reasonable probability,” recognizing that the precise burden placed on the defendant in this context may be open to debate. In a typical harmless error analysis in a criminal appeal, the State must show there is no "reasonable possibility” of harm. See Mendoza v. State, 87 So.3d 644, 660-61 (Fla.2011) ("The test for harmless error is whether there is a reasonable possibility that the error affected the *1169verdict.”) (quoting Mendoza v. State, 700 So.2d 670, 678 (Fla.1997))). In a postconviction proceeding, the defendant must show a "reasonable probability” that the error affected the outcome. See Kilgore v. State, 55 So.3d 487, 504 (Fla.2010) ("Therefore, Kil-gore has failed to demonstrate that the proffered evidence had a reasonable probability of changing the outcome, which is a probability sufficient to undermine our confidence in the verdict.”). Likewise, "reasonable probability" is a test often used in civil cases. See Chrysler v. Dep't of Prof'l Regulation, 627 So.2d 31, 34-35 (Fla. 1st DCA 1993) ("An error is harmful where there is a reasonable probability a different result would have been reached but for the error committed.”). The "reasonable probability” test seems the appropriate test to fulfill the narrow function served by fundamental error.
. See State v. DiGuilio, 491 So.2d 1129, 1135 (Fla. 1986).
. As explained in more detail later in this opinion, in Reed, 837 So.2d at 370 n. 3, the supreme court receded from Clark to the extent that Clark held that fundamental error could be harmless error.
. The supreme court does not appear to have ever overruled or narrowed its holding in Lucas.
. The shift announced in Reed can be compared to the shift announced in Parkway Bank v. Fort Myers Armature Works, Inc., 658 So.2d 646 (Fla. 2d DCA 1995), which altered the analysis used in certiorari review. Because attorneys and judges often struggled with the type of error that constituted a departure from the essential requirements of the law, Parkway Bank treated the requirement of an irreparable harm as an initial jurisdictional test. Id. at 649. In that way, many certiorari proceedings could be resolved on this relatively easy issue without need to reach the much harder question of whether an error is a departure from the essential requirements of law. Similarly, in Reed, the court focused on the need for an error first to be harmful. This often eliminates the need to decide whether the error is fundamental under the various definitions of fundamental error described infra at footnote 7.
. With modest success, Florida courts have employed numerous definitions to describe such a "fundamental” error. It can be an error going to the "heart of the judicial process,” Martinez v. State, 933 So.2d 1155, 1159 (Fla. 3d DCA 2006), or to the "heart .of the case,” Cruz v. State, 554 So.2d 586, 587 (Fla. 3d DCA 1989). It can be an error going to the "foundation” of the case or merely to the "merits” of the case. See Abdool v. State, 53 So.3d 208, 220 (Fla.2010) (quoting Bailey v. State, 998 So.2d 545, 554 (Fla.2008)); see also Daniels, 121 So.3d at 417 ("To justify not imposing the contemporaneous objection rule, 'the error must reach down into the validity of the trial itself to the extent that a verdict of guilty could not have been obtained without the assistance of the alleged error.’ ”) (quoting State v. Delva, 575 So.2d 643, 644-45 (Fla.1991))).
From a functional perspective, such an error is one that requires correction even in the absence of an objection to protect "the public’s confidence in our system of justice.” Murphy v. Int’l Robotic Sys., Inc., 766 So.2d 1010, 1033 (Fla.2000) (Pariente, J., concurring) (quoting Hagan v. Sun Bank of Mid-Florida, 666 So.2d 580, 584 (Fla. 2d DCA 1996))). It seems obvious that the public’s confidence in our system of justice would be severely eroded if we allowed juries to return verdicts of guilt for major felonies when the jury was not given a correct description of the elements of the offense. When the evidence *1172demonstrates that a jury reasonably could have given serious consideration to the next-lesser offense, an error in that instruction may equally erode the public’s confidence in its courts.
. The court went on to explain in Daniels that "[flailing to instruct on an element of the crime over which the record reflects there was no dispute is not fundamental error.” Daniels, 121 So.3d at 417-18. It therefore held, in accordance with Reed, 837 So.2d 366, and Delva, 575 So.2d 643, that "a defective instruction in a criminal case can only constitute fundamental error if the error pertains to *1173a material element that is disputed at trial,” and that element is both pertinent and material to what the jury must consider to convict the defendant. Daniels, 121 So.3dat418. In other words, fundamental error is not based on a per se rule, but requires a showing of actual harm or prejudice to the defendant.
The court further explained that its conclusion in Montgomery was based on this rule. Daniels, 121 So.3d at 418. But it did so without reconciling the conflicting per se rule from Pena, which it quoted without qualification. See Daniels, 121 So.3d at 415 & n. 5.
. Mr. Berube's version of events evolved during his interview, but this was the version that was most helpful to him at the time of trial.