concurring in part, dissenting in part.
I concur in the court’s opinion as to part IIA. I respectfully dissent, however, from part IIB. As the court notes, the parties agree that the plastic bags and tote that were the subject of the warrantless search were located in an open field. See Oliver v. United States, 466 U.S. 170, 178-79, 104 S.Ct. 1735, 80 L.Ed.2d 214 (1984). I agree with the court that the containers searched were in plain view to law enforcement. The question on appeal is whether Bob Sam Castleman had an objectively reasonable expectation of privacy in these closed containers.6 The court concludes he did not. With respect to the tote, I believe he did.7
Law enforcement officers are required to obtain a warrant or otherwise have con*919stitutional justification to access the contents of a closed container, even if the officer is lawfully present where the closed container is found. See United States v. Banks, 514 F.3d 769, 773-74 (8th Cir.2008) (“Ordinarily, a warrant is necessary before police may open a closed container because by concealing the contents from plain view, the possessor creates a reasonable expectation of privacy.”). There are exceptions, of course. One such exception is that law enforcement “may seize, without a warrant, an item that is (1) in plain view, (2) when it is observed from a lawful vantage point, (8) where the incriminating character of the item is immediately apparent.” Id. The tote in this case, however, was not incriminating in and of itself, and there was nothing externally visible to indicate the containers held evidence of a drug manufacturing operation. See id. at 775 (finding the search of a gun case without a warrant constitutionally acceptable, as the container was readily identifiable as a single-purpose container — namely, a gun case). Thus, while the officers were lawfully in the open field when they observed the tote, there was nothing about the tote itself that suggested the owner did not have a reasonable expectation of privacy in its contents. See id.
The tote was found on Castleman’s own 262-acre parcel of property, which is an important factor in determining whether his expectation of privacy in a closed container found there was objectively reasonable. Cf. United States v. Douglas, 744 F.3d 1065, 1070 (8th Cir.2014) (holding the defendant failed to demonstrate an objectively reasonable expectation of privacy over items found on his aunt and uncle’s property); United States v. Stallings, 28 F.3d 58, 61 (8th Cir.1994) (stating the defendant did not have an objectively reasonable expectation of privacy in items he had left unattended on someone else’s property). Castleman also took efforts to restrict access to his property with barbed-wire fences, a locked gate, and “No Trespassing” signs. Cf. Douglas, 744 F.3d at 1070 (holding the defendant failed to demonstrate his expectation of privacy was reasonable, as he had not attempted to regulate access to the bag beyond generally demanding the officers leave his aunt and uncle’s property). While fences, gates, and “No Trespassing” signs do not “effectively bar the public from viewing open fields in rural areas,” and thus cannot establish a reasonable expectation of privacy in an open field, Oliver, 466 U.S. at 178-79, 104 S.Ct. 1735, those same fences, gates, and signs are relevant when determining whether an expectation of privacy in items in that open field is objectively reasonable. Cf. Stallings, 28 F.3d at 60 n. 3 (“[T]he fact that the bag was left in an open field owned by another is highly relevant, as a factual matter, in demonstrating the reasonableness of [the defendant’s] expectation of privacy.”).
Castleman also did not place the containers in a location where he expected they would be taken or accessed by a third party. Cf. California v. Greenwood, 486 U.S. 35, 40, 108 S.Ct. 1625, 100 L.Ed.2d 30 (1988) (finding the respondents did not manifest a reasonable expectation of privacy in the garbage bags left on or near the public street, as they placed the bags on the curb “for the express purpose of conveying it to a third party” — namely, trash collectors). The tote in this case was found approximately 200 yards from Cas-tleman’s house, and not near the road or driveway leading to his home where visitors might frequent. Although the tote was visible to law enforcement during a search of the open fields, it was mostly covered by a plastic tarp. Castleman had both partially protected the tote from the elements and largely shielded it from the view of potential passers-by.
*920The court relies in part on United States v. Pennington, 287 F.3d 739 (8th Cir.2002), to support the conclusion that Castleman lacked an objectively reasonable expectation of privacy in the tote. In Pennington, we found no error in the district court’s denial of Pennington’s motion to suppress items found in an underground bunker located in an open field on his property. Yet the court made it clear that “the open fields doctrine only allows a search of what is in plain view in the open field. It does not justify a warrantless search of a man-made enclosure found in an open field.” Pennington, 287 F.3d at 745. Recognizing the important distinction between a war-rantless search of a man-made enclosure (prohibited) and a warrantless search of “what is in plain view in the open field” (permissible), Pennington presented the court with “a very close issue.” Id. Only the facts that the entrance to the underground bunker was “readily visible” with an “unprotected ladder facilitating access to the tunnel” and “no lock or door impeding access” to. that entrance rendered the warrantless search permissible. Id. at 746.
The reasoning we applied with regard to a “man-made enclosure” in Pennington should, in my view, also apply to the tote. A man-made enclosure may be in plain view in an open field, but law enforcement is not permitted to search that enclosure without a warrant, even if officers are lawfully in the open field in the first instance. See id. at 745. Similarly, the tote itself was in plain view in an open field— and law enforcement lawfully observed it there — but its contents were not. The tote was opaque plastic and the lid was closed, thereby completely concealing the contents from view. Law enforcement was therefore required to obtain a warrant before searching the tote.
The fact that a closed container is located in an open field does not mean a person has abandoned any objectively reasonable expectation of privacy in its contents. Instead, such circumstances require a fact-specific inquiry. See, e.g., Douglas, 744 F.3d at 1070-72; Stallings, 28 F.3d at 61. In my view, the facts of this ease present a close call, but the evidence presented in this case was enough to establish a reasonable expectation of privacy. The contents of the tote were not visible or readily ascertainable by sight, and.the tote itself was mostly covered by a tarp. The tote, a common type of storage bin for personal items, was located on property owned by Castleman, and that property was fenced, had a locked gate, and had posted “No Trespassing” signs. Because Castleman took minimally sufficient steps to preserve the tote as private, I conclude he had an expectation of privacy in the tote that society is prepared to recognize as objectively reasonable.
I also respectfully dissent from- part IIC of the court’s opinion. I agree that evidence of threats, including death threats, may be admissible to illustrate “consciousness of guilt.” See United States v. Zierke, 618 F.3d 755, 759 (8th Cir.2010). In some cases, it may be probative of guilt of the charged offense. Id.; see also United States v. Burnett, 579 F.3d 129, 133 (1st Cir.2009) (holding that evidence of a death threat may be introduced for a purpose “such as demonstrating consciousness of guilt by showing the lengths to which a defendant will go to keep damaging testi-' mony out of his trial”). On this issue, I do not disagree with the court. Yet even probative evidence may be inadmissible if the prejudicial effect of that evidence substantially outweighs its probative value. United States v. Weir, 575 F.2d 668, 671 (8th Cir.1978). In my view, the danger of unfair prejudice from introduction of the disputed evidence — that Castleman not only threatened a government witness, but in fact carried out that threat by murder*921ing him — substantially outweighs its probative value. See Fed.R.Evid. 403.
When considering the admissibility of a threat, there are several relevant factors to consider in determining whether the danger of unfair prejudice outweighs the probative value. These factors may include (1) whether the threat was “an emotional or impulsive reaction” as opposed to a calculated or premeditated plan to carry out the threat; (2) whether the jury heard “graphic details of how the threat would be carried out”; and (3) the inflammatory nature of the threat. Burnett, 579 F.3d at 134.
First, this case did not merely involve an emotional or impulsive threat. Rather, it also involved evidence of a premeditated murder carrying out that threat. Evidence that Castleman told his son, Jerrod Castleman, that Travis Perkins “had to go” may have been an impulsive statement or threat; the record offers little more than the statement itself, so it is difficult to know. Yet even if this threat was made on impulse, the evidence presented at trial did not stop there. Instead, the government introduced evidence of a calculated murder. Specifically, Jerrod testified that his father left his cell phone behind and drove from West Memphis, where he had been at the dog track, to the small town of Pocohontas, where Perkins lived. He then went to Perkins’s apartment wearing a wig and a trench coat, walked into Perkins’s bedroom, and shot Perkins in the face and chest with a handgun before disposing of the gun, holster, two clips, wig, and trench coat in the river. Rather than hearing about a simple, impulsive threat, the jury heard evidence of a murder and attempted cover-up, thus increasing the risk of an unfair prejudicial effect. Cf. id. (holding the threat was merely an impulsive reaction, which supported the district court’s decision to allow the evidence).
Second, instead of hearing how the threat would be carried out, the jury heard how the threat was in fact carried out by Castleman. And the government’s case agent — after already testifying in the government’s case-in-chief — was recalled to the witness stand to corroborate some of the details Jerrod provided, testifying that Perkins’s “feet and legs were hanging off the side of the bed ... as if either he was about to get up or had just laid down, one or the other, when he had been killed.” The agent also confirmed Perkins was shot in “the nasal area” and that there were two shell casings found on the floor near the body. These details, amounting to far more than a mere threat showing a consciousness of guilt, provided the jury with a graphic description of an actual murder of a government witness by Castleman. Cf. id. (In weighing the probative value of death threat against danger of unfair prejudice, we consider, among other things, “whether the jury heard graphic details of how the threat would be carried out.”).
Third, the evidence in this case — evidence of a murder of a government witness, presented unexpectedly in the middle of a drug conspiracy trial — was inflammatory in nature, thereby adding to its prejudicial impact. See id. (“We accept that Burnett’s threat was inflammatory because it was aimed at [the witness’s] daughter.”); see also United States v. McManaman, 606 F.2d 919, 926 (10th Cir.1979) (“[H]earing and reading the taped conversation lead us to the conviction that the inflammatory talk of the plans of murders clearly must have predominated in impact over the discussion of drug dealings.”). Even if probative of guilt, there was a substantial risk that this evidence improperly “diverged] the jury’s attention from the. material issues in the trial.” United States v. Fawbush, 634 F.3d 420, 423 (8th Cir.1990) (concluding other-act evidence that defendant had sexually abused his daughters *922and fathered a child with one of them was so inflammatory that it diverted the jury’s attention and should have been excluded under Rule 403). The material issues at Castleman’s trial included whether he voluntarily and intentionally joined a conspiracy to manufacture methamphetamine and whether he voluntarily and intentionally allowed other individuals to use his property to manufacture methamphetamine. Then, on the third day of trial, the jury heard from Castleman’s son that Castle-man had also committed murder. Murder is a type of crime quite different in kind and quality than drug trafficking, and it is difficult to imagine that this evidence did not change the tone, tenor, and focus of the trial.8
Another relevant factor for consideration in this case is the government’s delay in filing its notice of its intent to offer evidence of the murder until the morning of the third day of trial. Trial started on Thursday; on the following Monday morning, the government said they “learned on Thursday morning of last week about this information, but [we] did not have a chance to talk directly with Mr. Jerrod Castleman about the information.”9 The district court allowed the evidence and granted a one-day continuance. The evidence was presented the following day.10 Thus, any “danger of unfair prejudice” was exacerbated because Castleman, as a practical matter, had no opportunity to investigate or prepare a meaningful defense to this new, highly damaging consciousness of guilt evidence. And, as a result, the jury was left with only the government’s version of events. Perhaps Castleman would not have been successful, even with additional time, in presenting an alternative explanation or lodging a credible defense to the murder allegation. As it happened, however, he was not given an opportunity.
Federal Rule of Evidence 403 “does not offer protection against evidence that is merely prejudicial in the sense of being detrimental to a party’s case. The rule protects against evidence that is unfairly prejudicial, that is, if it tends to suggest decision on an improper basis.” United States v. Myers, 503 F.3d 676, 681-82 (8th Cir.2007) (quotation omitted); see also Fed.R.Evid. 403 advisory committee’s note (explaining that unfair prejudice “means an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one”). In this case, the danger of unfair prejudice was simply too strong. Without notice prior to trial, the government introduced evidence not only that Castleman had made a threat against a government witness, but that he had in fact carried through on that threat and violently murdered that witness prior to trial. See Burnett, 579 F.3d at 133 (While evidence that defendant threatened to kill a witness pre-trial may be admissible to show consciousness of guilt, “[s]uch evidence can be highly prejudicial ... and *923it should not be admitted if its probative value is substantially outweighed by the danger of unfair prejudice.” (quotation omitted)). It is unrealistic, and unfair, to expect a jury to compartmentalize this accusation into the category of “consciousness of guilt” and “not to consider this evidence for any other purpose.” See Instruction No. 17. The risk that the jury would render its decision “on an improper basis” under these circumstances is unacceptably high.
I recognize that the district court gave a limiting instruction in this case, both at the time of the testimony and again in the final instructions read to the jury. I also recognize the deference properly given to district courts who must make difficult evi-dentiary decisions during the course of a trial. I believe, however, the probative value of the evidence that Castleman murdered a co-conspirator prior to trial was substantially outweighed by the danger of unfair prejudice and should have been excluded. See Weir, 575 F.2d at 670-71. On this issue, I respectfully dissent.
Because the error in admitting the testimony about Castleman’s alleged murder of Perkins was not harmless, I would reverse the conviction on that grounds. I therefore would not reach the sentencing issues addressed in part IID.
. The district court and this court on appeal have assumed Castleman had a subjective expectation of privacy in the bags and tote searched. I have done the same.
. The containers seized and searched included full plastic trash bags and a hard-sided plastic tote. I recognize that plastic trash bags are sometimes used for storage of items a person wants to save, not discard. Thus, in some situations, a person may be able to show a subjective expectation of privacy in a closed plastic trash bag, and that expectation may be objectively reasonable. In this case, however, there is no indication that the plastic bags were anything but full of trash. I therefore concur in the court’s conclusion that Castleman failed to meet his burden of proving that any expectation of privacy he may have had in discarded trash was objectively reasonable.
.While not a dispositive factor, courts have considered "the Government’s need for the evidence" to prove its case at trial. See Burnett, 579 F.3d at 134; United States v. Check, 582 F.2d 668, 685 (2d Cir.1978) (noting that, due to the "severe prejudice [that] can result from the use of death threat testimony ... we have carefully limited that use to situations where there was a clear need for the prosecution to use such evidence”). The government did not explain why the consciousness of guilt evidence in this case was necessary.
. The government also stated that "Jerrod Castleman’s safety” "played into the timing” of the filing of the notice of intent to offer additional evidence, although it did not elaborate on this concern.
. Keep in mind, too, that jury selection would not have filtered any individual venire members who might have expressed a real concern about sitting on a jury at a trial where evidence of a murder would be introduced and who, as a result, could not be fair and impartial.