GILLETTE, J.

The decision of the Court of Appeals is reversed. The judgment of the circuit court is affirmed.

GILLETTE, J.

This criminal case presents an interesting (and recurring) question: Does Article I, section 9, of the Oregon Constitution, require police officers to obtain a warrant before using a trained drug-detecting dog to sniff the exterior of a locked storage unit for odors that are imperceptible to the human nose? A majority of the en banc Court of Appeals answered that question in the affirmative, holding that any evidence obtained as a consequence of a warrantless dog sniff must be suppressed. State v. Smith, 148 Or. App. 235, 246, 939 P.2d 157 (1997). The state petitioned for review, arguing that, even without prior judicial authorization, use of a dog to sniff in the described manner is lawful, either because it is not a search within the meaning of the Oregon Constitution or because it is not subject to the usual "warrant based on probable cause" standard. We allowed the state's petition to consider those and other related issues. We now conclude that use of a dog to sniff property in this manner is not a search for constitutional purposes and, consequently, does not implicate Article I, section 9. We also conclude that defendant's alternative theory justifying suppression — that evidence obtained from the storage unit must be suppressed, because the police unlawfully seized that evidence when they "secured" the unit in anticipation of obtaining a search warrant — is incorrect. We therefore reverse the decision of the Court of Appeals.

In 1993, the Brookings Police Department received information from a confidential informant that defendant, a Klamath Falls resident, had a marijuana growing operation in the Brookings area and that he periodically used a unit in a specified storage facility to store harvested marijuana. Shortly thereafter, the Klamath Falls police arrested and jailed defendant, based on information supplied to them by the same informant and on evidence obtained from a search of defendant's residence. Not long after defendant's arrest, the Klamath Falls police contacted the Brookings police and told them that, according to the informant, defendant still had the same storage unit in Brookings. Officer Plaster of the Brookings police contacted the manager of the storage facility and confirmed that defendant still was renting a storage unit. The manager stated that he would not permit the police to search defendant's unit without a warrant.

Plaster and another police officer, Gardiner, went to the storage facility to "secure" defendant's unit. They took along Nitro, a trained drug-sniffing dog. While Plaster spoke to the facility manager, Gardiner took Nitro to the public area near defendant's unit and told him to sniff for drugs. Nitro "alerted" at defendant's unit, suggesting the presence of illegal drugs. The officers asked the manager to place a lock on defendant's unit while they applied for a search warrant. The manager complied. Several hours later, Plaster returned with a warrant, which he had obtained on the basis of his own affidavit describing his contacts with the informant and the Klamath Falls police, and the results of the dog sniff. During the ensuing search, Plaster found and seized marijuana and implements used in marijuana cultivation.

Defendant was charged with manufacture, possession, and delivery of a controlled substance, ORS 475.992. Before trial, he moved to suppress the evidence obtained from the storage unit, arguing that it was obtained unlawfully in two respects: (1) the search warrant was not supported by a sufficient showing of probable cause, because the supporting affidavit relied, in large part, on an unlawful search, viz., a warrantless dog sniff of the unit's exterior; and (2) padlocking of the unit was a warrantless seizure that was not justified by exigent circumstances. The trial court denied the motion, concluding that the police did not need a warrant to conduct the dog sniff because it was not a search and that, even if padlocking the unit were an unlawful seizure, that illegality was unrelated to the later search and seizure conducted pursuant to the warrant. Defendant was tried and convicted. He appealed, assigning error to the denial of his motion to suppress.

As noted, a majority of the Court of Appeals, sitting en banc, concluded that the challenged ruling was reversible error, because the warrantless dog sniff violated Article I, section 9. In reaching that conclusion, the majority considered only the narrow question of whether a "dog-sniff" search is subject to the constitutional warrant and probable cause requirements, ultimately answering that question in the affirmative. For the more fundamental proposition that use of a trained drug detection dog is a search within the meaning of Article I, section 9, the majority relied on that court's earlier decision in State v. Juarez-Godinez, 135 Or. App. 591, 900 P.2d 1044 (1995), aff'd 326 Or. 1, 942 P.2d 772 (1995), a decision that had held that dog sniffs were constitutionally significant searches. Having decided in defendant's favor on the search issue, the Court of Appeals never reached the alternative seizure argument.

In its petition for review to this court, the state challenges both the Court of Appeals' Juarez-Godinez holding and its conclusion that dog sniffs require a full probable cause showing. In our view, the first issue is the key. We begin, therefore, with the Court of Appeals' holding in Juarez-Godinez that use of a trained dog to sniff the exterior of a private enclosed space is a search within the meaning of Article I, section 9, of the Oregon Constitution.

The Court of Appeals analysis in Juarez-Godinez turned on the fact that, like the odor of marijuana detected by Nitro in this case, the odor of drugs at issue there was not perceptible to human beings. Relying on State v. Campbell, 306 Or. 157, 759 P.2d 1040 (1988), the Court of Appeals opined that the real touchstone of acceptable governmental scrutiny under Article I, section 9, is whether the conduct involved could have been done by any private individual. Thus, the court concluded, even when the police make their observations from a vantage point where they have every right to be, a search will occur if the police observe things that an ordinary individual could not have observed from that vantage point. From that standpoint, the court concluded, it is clear that dog sniffs are searches: Dogs discern odors that would not, and could not, be detected by any ordinary private citizen, and are used by the police for that precise reason. Juarez-Godinez, 135 Or App at 602-04.

The state contends that that analysis expands the notion of a protected privacy interest under Article I, section 9, to an unwarranted and unworkable degree. In the state's view, the odor of molecules that have escaped from a contained, private space into a place to which police officers legally have access is fair game for any kind of observation — aided or unaided — because those molecules and their odor-producing properties are in the public domain. The state argues, in other words, that Article I, section 9, is concerned with where the police are directing their attention and not with how they perform their observations.

1. Before we address the parties' arguments, we need to address certain misconceptions about Oregon search and seizure law that appear in the Court of Appeals' Juarez-Godinez opinion. First, we do not agree with the Court of Appeals' construction of Campbell or its conclusion that the applicability of Article I, section 9, turns on whether the evidence can be perceived directly by unenhanced human senses. Although Campbell did conclude that use of a particular enhancement — a tracking device — was a constitutionally significant "search," it never suggested that use of any device or enhancement — no matter where that device or enhancement was used — would qualify as such. In fact, in a more recent case, this court explicitly rejected the suggestion that Campbell contains such a per se rule. See State v. Wacker, 317 Or. 419, 426 n 12, 856 P.2d 1029 (1993) (use of light-enhancing device ("starscope") to aid in seeing activity in a car parked in a parking lot).

2. We also disagree with the Court of Appeals' suggestion that the fact that a dog sniff involves no invasion of protected space is inconsequential to the constitutional analysis. Our cases suggest that some form of invasion of a private space is a common, although not essential, element of the search construct under the Oregon Constitution. Compare, e.g., State v. Dixson/Digby, 307 Or. 195, 211-12, 766 P.2d 1015 (1988) (police entry into undeveloped land with "No Trespassing" signs would be search); Campbell, 306 Or at 172 (trespassory act of attaching tracking device to car was a search); State v. Bridewell, 306 Or. 231, 759 P.2d 1054 (1988) (entry into house and workshop without statutory authority was search); State v. Kosta, 304 Or. 549, 554, 748 P.2d 72 (1987) (opening of car trunk was a search); State v. Kock, 302 Or. 29, 725 P.2d 1285 (1986) (opening car door and reaching in to seize package was a search); with State v. Ainsworth, 310 Or. 613, 801 P.2d 749 (1990) (observation from lawfully positioned helicopter not a search); State v. Jackson, 296 Or. 430, 677 P.2d 21 (1984) (walking around and looking in windows of car is not search); State v. Louis, 296 Or. 57, 672 P.2d 708 (1983) (telephoto photograph through living room window from across street not a search; camera captured image visible from the public sidewalk). True, all those cases involved the use of (and, on occasion, the modest enhancement of) ordinary human powers of perception and observation. A dog is not a human; this case thus is different. But it is not different with respect to the pivotal question, which is whether the police invaded a protected privacy interest.

3. Commonly, a search involves some form of invasion into private space. In so noting however, we are not attempting to dispose of or supplant our traditional construction of Article I, section 9, as protecting privacy interests, i.e., the individual's interest in freedom from certain forms of governmental scrutiny. See Dixson/Digby, 307 Or at 206; Campbell, 306 Or at 170. We continue to adhere to that construction. At the same time, however, we acknowledge that private space and privacy interests often are inextricably intertwined. That is so because privacy interests generally are not self-announcing and, with a few possible exceptions, can be recognized only by their association with a private place, i.e., by the fact that an object is kept or conduct occurs in a place that legitimately can be deemed private. See, e.g., Dixson/Digby, 307 Or at 211-12 (defendant had protectable privacy interest in activity on undeveloped privately owned land only to the extent that he manifested an intent to exclude others by erecting fences or "no trespassing" signs). Ultimately, then, the privacy interests that are protected by Article I, section 9, commonly are circumscribed by the space in which they exist and, more particularly, by the barriers to public entry (physical and sensory) that define that private space.

4, 5. That is not to say that, to qualify as a search, the invasion always must be of the type that the law traditionally has labeled as a "trespass" — an actual physical intrusion. As anyone who lives in the modern world should know, private space can be invaded by technological as well as physical means. Clearly, if Article I, section 9, is to have any meaning, it must be read in light of the ever-expanding capacity of individuals and the government to gather information by technological means. It must, in other words, speak to every possible form of invasion — physical, electronic, technological, and the like.

6. That being said, we turn to the present case, involving the use of a dog to detect the presence of a particular odor caused by the presence of odor molecules in the air outside a clearly defined, private space. Obviously, that sort of examination does not involve a physical invasion into private space. The question, then, is whether it involves a constitutionally significant invasion of a different sort.

The argument might be made that, because it suggests something about the contents of a private space, a dog sniff is at least a figurative invasion into that space. We do not ascribe to that view. The police regularly and lawfully make observations from the exteriors of enclosed spaces and draw reasonable inferences about the contents of those spaces that no one considers to be a violation of Article I, section 9. See, e.g., State v. Owens, 302 Or. 196, 206, 729 P.2d 524 (1986) (seizure of transparent vial containing white powder did not violate Article I, section 9; police had probable cause, based on vial's exterior, to believe that it contained a controlled substance).

On the other hand, some police techniques clearly do involve technological or figurative invasions of privacy interests. High-powered telescopes, parabolic sound gathering devices, and infrared cameras all are able to gather ordinarily imperceptible information (in the form of light rays, sound waves, or heat waves) from the interior of an enclosed space as it emanates from that space. Although such devices can operate from an external and public place, their use is invasive nevertheless, because they provide information from an enclosed space about the enclosure's contents that a police officer, standing at a lawful vantage point, cannot detect with ordinary human powers of perception. That is, they constructively move the observing police officer into the private space.

Superficially, one might argue that dog sniffs fall into that category — that trained police dogs perceive odor molecules as they emanate from an enclosed space and then indicate the contents of the enclosure by their reactions. But dog sniffs do not provide the kind of direct access to the interiors of enclosed spaces that the aforementioned devices provide. Dog sniffs operate on a sample of air whose origins are unknown. One may infer from the fact that a dog detects the presence of a substance in the public airspace surrounding an enclosure that the enclosure contains that substance. However, that would only be an inference. For the purposes of both the dog and its human interpreter, the substance that is detected remains an ambient substance in the public airspace. As such, dog sniffs do not extract information from the interior of a private space, either literally or figuratively. They are not invasive in any sense of the word.

The clear import of the foregoing is that, at least when they are conducted in a public place, dog sniffs are not searches. And, because they are not, the protections of Article I, section 9 — including the warrant requirement — do not apply. The Court of Appeals erred in concluding otherwise in Juarez-Godinez, and its reliance on that interpretation in the present case likewise was erroneous.

7. At the Court of Appeals, defendant also argued that the dog sniff was unlawful under the United States Constitution. His argument is brief and primarily relies on one federal circuit court case, United States v. Thomas, 757 F.2d 1359 (2d Cir 1985), to the effect that use of a trained dog to sniff the hallway outside of a suspect's apartment is a search requiring a warrant.

We are not persuaded. First, Thomas itself would appear to be questionable law, in light of a number of Supreme Court and federal circuit court decisions holding that dog sniffs are not searches. See, e.g., United States v. Place, 462 U.S. 696, 103 S Ct 2637, 77 L Ed 2d 110 (1983) (dog sniff of luggage in transit through an airport is not a search); U.S. v. Colyer, 878 F.2d 469 (D C Cir 1989) (dog sniff in public corridor on sleeper car of train was not a search); U.S. v. Lingenfelter, 997 F.2d 632 (9th Cir 1993) (dog sniff of public alleyway between warehouses not a search); U.S. v. Ludwig, 10 F.3d 1523 (10th Cir 1993) (dog sniff of car in motel parking lot not a search). Second, even if Thomas were good law, it is irrelevant to the present facts: By its own terms, Thomas appears to be limited to dog sniffs of dwellings. The sniff at issue here, of a storage unit, more closely resembles the circumstances in Lingenfelter, i.e., a sniff of the exterior of a warehouse, which was held not to be a search.

In short, we find no support for defendant's suggestion that the dog sniff at issue was a search under the Fourth Amendment. The trial court thus was correct in denying defendant's motion to suppress on those grounds. Contrary to defendant's view, the police acted lawfully when they conducted the dog sniff at issue without first obtaining a judicial warrant.

8. Having disposed of defendant's dog sniff argument, we proceed to his alternative argument — that the motion to suppress should have been granted, because the evidence at issue was derived from an unlawful seizure. The seizure to which defendant refers is the padlocking of defendant's storage unit. That seizure was unlawful, in defendant's view, because it was not authorized by a judicial warrant and was not justified by exigent circumstances or any of the other recognized exceptions to the warrant requirement.

9. The state recognizes that the act of padlocking was a seizure — and rightly so. As we previously have noted, property is "seized" for the purposes of Article I, section 9, when the police significantly interfere with a person's ownership and possessory interests in the property. Owens, 302 Or at 207. Padlocking the unit represented significant interference with respect to the unit. At least in theory, it deprived defendant of the use of the unit and access to its contents.

Although the state concedes that padlocking the unit was a seizure, it does not concede that that act was unlawful for lack of a warrant. In that regard, the state maintains that padlocking the unit involved such a minimal intrusion on defendant's rights that no warrant was required. In so arguing, the state relies on various treatises and cases from other jurisdictions to the effect that the police may "secure" a residence from the outside without first obtaining a warrant. The reasoning offered for that distinction — a distinction that the state finds relevant in the present case — is that securing premises from the outside invades only possessory rights, which, relative to privacy rights, are of secondary importance in the constitutional scheme. See, e.g., Segura v. United States, 468 U.S. 796, 806, 104 S Ct 3380, 3386, 82 L Ed 2d 599 (1984) (suggesting that invasion of possessory rights are less significant than invasions of privacy rights); State v. Solberg, 66 Wn. App. 66, 831 P.2d 754 (1992), rev'd 122 Wn.2d 688, 861 P.2d 460 (1993) (no warrant required to secure premises from outside when police have probable cause). See also, Wayne R. LaFave, 3 Search and Seizure § 6.5(c) (3rd ed 1996) (discussing "impoundment" of premises to prevent loss of evidence).

10, 11. We do not agree with the state's suggestion that possessory rights deserve less protection than privacy rights. Whatever other jurisdictions may have said about the subject, it is clear that Article I, section 9, speaks to both types of interests and that, with a few well-recognized exceptions, a warrant is required even when only possessory rights are implicated. See, e.g., Juarez-Godinez, 326 Or at 8-9 (warrantless seizure of car offended Article I, section 9). Padlocking the storage unit in this case was no less of a seizure because it occurred from the outside. In the absence of a warrant, it was unlawful.

In its final argument, the state contends that, even if the seizure was unlawful, the unlawfulness was not causally related to the seizure of the evidence at issue and, therefore, provides no grounds for suppression. In that regard, the state points out that the police obtained the evidence at issue in a search of the storage unit pursuant to a warrant and that, logically, the locking of the unit contributed nothing to the state's uncovering of that evidence. Under those circumstances, the state contends, the evidence was not obtained in violation of defendant's right under Article I, section 9, and should not be suppressed.

Defendant agrees that unlawful police conduct requires suppression only when there is a causal connection between the conduct and the evidence at issue. He argues, however, that the requisite causal connection exists in this case — that the police seized the evidence at issue when they padlocked the storage unit, and that the subsequent search and seizure that occurred pursuant to the later-acquired warrant could not unring that bell. In support of that view, defendant relies on State v. Hansen, 295 Or. 78, 664 P.2d 1095 (1983).

In Hansen, police officers suspected that the defendant had marijuana in his home and seized the residence unlawfully by entering it and "securing" it until a search warrant arrived. Although the police performed a cursory search of the residence when they first entered it, they did not use any information gleaned from that search in their application for a search warrant. Hours later, the search warrant arrived and the police obtained marijuana evidence in the ensuing search. This court nonetheless concluded that, because the evidence at issue was primary evidence, i.e., the very evidence the officers were seeking when they committed the illegality, it must be suppressed. In particular, the court concluded that, when the residence was seized, any later-discovered primary evidence in the residence also was seized. That was so because, in view of the fact that the purpose of securing the residence was to secure any evidence of marijuana possession, the act of securing the residence "as effectively reduced the marijuana to the control of the trespassing police as if they had actually discovered and taken physical possession of it." 295 Or at 97.

Defendant contends that the seizure at issue in the present case is no different — that because the evidence obtained was of the very sort that the police were anticipating when they unlawfully seized the storage unit, that evidence also was unlawfully seized at the time of the padlocking. Under Hansen, defendant argues, that evidence is subject to suppression.

Defendant does acknowledge that, in a more recent case, State v. Sargent, 323 Or. 455, 918 P.2d 819 (1996), involving facts very similar to those in Hansen, this court reversed a suppression order on the ground that there was no logical connection between the unlawful securing of the residence and the discovery of what was, indisputably, primary evidence. He argues, however, that Sargent does not represent a change in the court's position — that the two cases are distinguishable factually and that they ultimately deal with different aspects of Article I, section 9. In particular, he contends that Sargent is concerned only with police interference in privacy rights, i.e., searches, while Hansen is concerned with police interference with property rights, i.e., seizures.

We do not agree that Sargent and Hansen are that easily distinguished. It is clear that the factual distinction to which we alluded in Sargent — between an unlawful entry into a private residence and the "seizure of a residence and its contents [that] continued beyond a reasonable time" — did not dictate the ultimate outcome in that case: Ultimately, we assumed that an unlawful seizure had occurred, but concluded that that seizure had no effect on the subsequent valid search. 323 Or at 462-63. Moreover, it also is clear that Sargent was argued by the parties in terms of Hansen — in terms, that is, of the validity of an asserted seizure which, according to the defendant, was accomplished when the police secured the residence. The fact that we chose to ignore that rationale and, instead, couched our analysis in terms of the validity of the search is an indication that the seizure rationale was unpersuasive — that, regardless of what we said in Hansen, no seizure of the evidence at issue occurred until the police seized it directly upon searching the residence. Ultimately, Sargent must be read as tacitly rejecting Hansen, at least to the extent that Hansen holds that unlawful seizure of property necessarily triggers suppression of any evidence contained therein, whether or not the police subsequently obtain lawful authority to search that property.

12. That we no longer adhere to the view in Hansen should not be surprising. Hansen was decided in an era when this court made little effort to evaluate Oregon's constitutional guarantees as separate from those in the United States Constitution. Hansen thus appears to have been an attempt to vindicate the "police deterrence" rationale of the Fourth Amendment — that is, to prevent the police from deriving any benefit from the unlawful practice of seizing a residence in mere anticipation of obtaining a warrant to search for evidence of suspected crimes. This court since clearly has rejected that deterrence rationale as foreign to the Oregon search and seizure provision, holding, instead, that the Oregon exclusionary rule exists to vindicate a personal right to be free from unlawful searches and seizures. See Kosta, 304 Or at 553 (so stating); Sargent, 323 Or at 462 n 4 (same). To support that purpose, it is sufficient to suppress only evidence that is actually obtained out of an illegal search or seizure. The police may gain some degree of control over the contents of private property when they "secure" that property, but they do not obtain evidence out of such a seizure.

We do not deny that, in some circumstances, the act of securing certain property may permit the police to obtain evidence that otherwise would not be available to them. However, in the present case, it clearly did not. No one attempted to gain access to the unit to remove the evidence before the search warrant was executed. As such, the padlock, although unlawful, was irrelevant. The evidence would have been obtained even in the absence of the unlawful police conduct. The mere fact that evidence was discovered after that unlawful conduct does not require suppression. See also State v. Rodriguez, 317 Or. 27, 39, 854 P.2d 399 (1993) (unlawful arrest did not automatically prevent later-obtained consent to search from being valid).

The decision of the Court of Appeals is reversed. The judgment of the circuit court is affirmed.