Skip to main content


Reset A A Font size: Print

Court of Appeal, Second District, Division 2, California.

The PEOPLE, Plaintiff and Respondent, v. Ryan WILLIS, Defendant and Appellant.

No. B160530.

Decided: January 28, 2004

 Verna Wefald, Pasadena, under appointment by the Court of Appeal, for Defendant and Appellant.  Bill Lockyer, Attorney General, Robert R. Anderson, Chief Assistant Attorney General, Pamela C. Hamanaka, Senior Assistant Attorney General, Jaime L. Fuster and Victoria B. Wilson, Deputy Attorneys General, for Plaintiff and Respondent.

Ryan O'Neal Willis was convicted by a jury of first degree murder, with the special circumstances of torture and lying in wait.   He was sentenced to life in prison without the possibility of parole.   On appeal, Willis challenges the trial court's admission of dog scent identifications, arguing that “this unreliable yet highly prejudicial evidence” deprived him of a fair trial.

We agree that the dog scent evidence was improperly admitted because it lacked foundation and scientific proof of reliability.   Nevertheless, the error was harmless because there is no reasonable probability that the jury would have reached a different result had the dog scent evidence been excluded.   Accordingly, we affirm the judgment.


The Victim Is Found in a Burned Car

At approximately 12:50 a.m. on January 6, 2000, a Palmdale resident came upon a car engulfed in flames in a church parking lot.   The witness promptly drove to her nearby home and called 911.   After the fire department extinguished the flames, a body charred beyond recognition was found curled on the floor behind the driver's seat.   The victim's bones were fractured from the intense heat of the fire, with brain matter leaking from the skull.   High blood levels of carbon monoxide, and soot deposited in the throat and lungs, revealed that the victim was alive and breathing while trapped in the burning car.

An arson investigator opined that the fire began toward the rear of the passenger compartment, most likely on the floor.   The fire was not caused by any mechanical failure of the electrical or fuel systems.   An open flame was used to intentionally ignite paper material;  once the seat cushions were affected, the fire would have burned rapidly and become very hot.

Some coins, papers and the driver's license of Crystal Stahl were retrieved from the ground near the burned car, strewn about as if there had been a struggle.   Through dental records, the body in the car was identified as that of Crystal Stahl.

 The Violent Relationship Between Appellant and the Victim

Crystal Stahl was a taxicab driver who cohabited with appellant Ryan Willis from 1998 until a few weeks before her death.   During his relationship with Stahl, appellant called the taxicab dispatcher as often as 20 to 30 times per week asking to speak to Stahl.   As a result, the taxicab dispatcher was readily able to recognize appellant's voice.   The dispatcher saw appellant riding in Stahl's cab from time to time.

The relationship between appellant and Crystal Stahl was marked by domestic violence.   In December 1998, Stahl's foster mother, Beverly Williams, witnessed a fight between Stahl and appellant that began in Stahl's bedroom.   When Stahl began yelling, Williams banged on the door and threatened to call the police.   Stahl came out in tears, wearing only panties, with bite marks on her breast, back and neck, and pinch marks on the other breast.   Williams struggled for 10 to 15 minutes to separate the couple.   While Stahl cried out for him to stop, appellant pulled Stahl up seven or eight times, then let her fall so that her head hit the sink or the floor.

One of appellant's friends reported seeing bite marks on Stahl's arm.   On three or four occasions, Stahl came to work with bruises on her face, arms and neck, according to the taxicab dispatcher and another coworker.   In mid-December 1999, a friend visited Stahl's apartment.   She saw appellant hit Stahl several times during the visit.   When the friend tried to intervene, she was struck in the face and arm.   During the course of the fight, appellant called Stahl a “bitch” and a “slut” and twice said to Stahl, “You're never going to leave me.”   Stahl yelled, “Get away from me,” while appellant pinned her to a wall.

The Victim Breaks Off Her Relationship with Appellant

Three weeks prior to her death, Stahl broke up with appellant and began a new romantic relationship.   Stahl called the police on December 17, 1999, after appellant attacked and bit Stahl when she refused to allow him back into her apartment.   Appellant was arrested for this attack.   The arresting officer described Stahl as “upset and scared,” while appellant “was very calm.”   Stahl moved in with her new beau.   Appellant confided to a friend that he was angry because Stahl “was screwing another guy” and because of his December 17 arrest for domestic violence.

Appellant Announces His Intent to Kill the Victim

Appellant threatened to kill Stahl shortly before her burned body was found in the taxicab.   Appellant told a group of friends that he was angry at  Stahl for his December 17 arrest.   According to one of the individuals present, appellant said, “Shit, I should kill the bitch for putting me in jail.”   Another witness testified that appellant announced, “I'm going to kill that fucking bitch.   I'm going to burn that fucking bitch to the fucking ground in her car and all, going to burn her to death.”   Appellant claimed he would kill Stahl, then go to Las Vegas, where no one would find him.   The people who heard appellant's threat did not think that he was serious.

Shortly after learning of Stahl's death, one of the individuals who heard appellant's threat called the police.   The witness told the investigating officer that appellant said he was going to burn Stahl to death in her taxicab because other cab drivers get killed, so no one would suspect him.

The Victim Agrees to Pick up Appellant Shortly Before Her Death

Stahl worked the night shift on January 5-6, 2000.   At 11:40 p.m. on January 5, appellant called the cab company and asked to be picked up by Stahl and no other cab driver.   The dispatcher recognized appellant's voice with absolute certainty, even though he gave a false name.   The dispatcher informed appellant that Stahl was busy transporting another fare, and appellant replied that he would wait at a gas station called The Pit Stop until Stahl was free to collect him.

Between 12:15 and 12:20 a.m. on January 6, Stahl called the dispatcher, who told her that appellant was waiting to be picked up at The Pit Stop. The dispatcher offered to tell appellant that Stahl was unavailable, but Stahl rejected the suggestion and told the dispatcher that she would collect appellant, who was only five blocks away from Stahl's location.   That was the last time the dispatcher ever heard from Stahl.   The Pit Stop is one mile and a half away from the church where Stahl's body was found inside the burned taxicab.

Appellant Leaves Town

On the afternoon of January 6, 2000, appellant called a friend to tell her that he had arrived in Las Vegas.   Toward the end of January 2000, one of appellant's friends saw him walking down the street in the Antelope Valley.   Appellant instructed his friend not to say that he had seen appellant.   After his arrest, appellant informed the homicide investigator that he went to Las Vegas on January 4, 2000, and remained there until February.   Appellant admitted that he had a tempestuous relationship with the victim.   He also stated that he shopped at The Pit Stop.

 The Dog Scent Identification

Investigators tried to link appellant to the scene with dog scent identification.   A civilian canine handler employed on an hourly basis by the sheriff's department used a bloodhound named Scarlet to detect a scent.   Until 1995, dogs were primarily used to track down missing hikers, children or Alzheimer's patients.   The main problem with using dogs to identify people in criminal investigations is that there is less scent to work with compared to tracking cases, where the dog is given an entire bed sheet or article of clothing to smell.   Scarlet, who was involved in some 600 criminal investigations, was trained to jump on the person being tracked when she located the individual.

The dog handler testified that a scent trail on the ground can last for at least several weeks.   A scent on an item, if properly stored, will last for months or years.   Scent is collected from objects using a vacuum device called a “scent transfer unit” (STU). The STU holds a five-by-nine-inch gauze pad that ostensibly collects and preserves the scent from virtually any object a person has touched.   The dog handler uses the scent pad from the STU instead of simply using the object itself for scent to ensure consistency and to prevent the dog from damaging the evidence.

In this instance, Scarlet and her handler arrived at the church parking lot at 6:00 a.m. on January 6, 2000.   The handler used the STU to create scent pads from certain objects found in the vicinity of the taxicab:  a black purse, knit cap, black glove, black hair extension, and a minimart matchbook.   Using the matchbook scent, Scarlet “showed interest” (i.e., was animated or excited) in a vacant lot near the church and at several apartment houses where appellant lived or spent time.   After appellant was arrested, Scarlet went right up to appellant in the police station and sat in his lap, in the presence of five deputies.

The defense attempted to debunk the dog scent evidence.   A veterinary professor testified that scent dogs are invaluable aids to criminal investigations.   Nevertheless, the dog must unambiguously “alert” or signal a discovery, for example by sitting and barking, in a way that is not open to subjective interpretation.   If a dog merely “shows interest” it means the dog is working, not that it has made a discovery.   If a dog shows repeated interest in numerous locations it could suggest that the dog was following a trail.   But the dog still needs to give an unambiguous alert.   Scarlet was not trained to give an unambiguous alert to an area, as opposed to a person:  she is trained to jump up and put her paws on people who are being tracked.

With respect to the STU, the veterinary professor observed that the basic idea was “pretty good”;  however, the STU device is made of plastic, which  “absorbs odors immensely well” and is extremely hard to clean of odors that have impregnated the plastic.   The STU might be reliable if it were made of another material.   At present, it is not a reliable method for scent collection.   In any event, the veterinarian felt that we do not know enough about the reliability of canine identification of human scent, and the bias of the handler in subconsciously giving subtle signals to the dog is a critical consideration.


1. Admissibility of Dog Scent Evidence

Appellant made a pretrial motion to exclude all dog scent and STU evidence, asserting that it lacked foundation and was experimental.   The prosecution made a counter-motion to have the dog scent evidence admitted, arguing that the Kelly-Frye rule does not apply to the collection or preservation of evidence.1  The trial court ruled that the evidence was admissible.

 The Kelly rule applies to new scientific techniques, especially in cases involving novel devices or processes.   To protect the jury from scientific techniques that convey a misleading aura of certainty, a new technique must be subjected to a three-part test of reliability.   First, there must be proof that the technique is considered reliable in the scientific community.   Second, the witness testifying about the technique must be a qualified expert on the subject.   Third, there must be proof that the person performing the test used correct scientific procedures.   The analysis is needed where the evidence is so foreign to everyday experience that it is difficult for laypersons to evaluate merely using their own common sense and good judgment.  (People v. Mitchell (2003) 110 Cal.App.4th 772, 782-783, 2 Cal.Rptr.3d 49.)

 “The scent transfer unit is a novel device used in furtherance of a new technique” and, as such, is subject to a Kelly analysis.  (People v. Mitchell, supra, 110 Cal.App.4th at p. 789, 2 Cal.Rptr.3d 49.) It is not obvious that a vacuum device can properly transfer scent to a gauze pad from an object.   Nor is it evident that the scent would not degrade or become contaminated, or that scent would not remain in the STU after it was cleaned with alcohol.  (Ibid.) As noted during trial, the plastic housing of the STU is porous, so it is difficult to remove odors.

 The evidence in this case did not meet the requirements of Kelly.   The dog handler who testified for the prosecution is not a scientist or an engineer;   therefore, he is not qualified to testify about the characteristics of the STU or the unit's acceptance in the scientific community.   There was also no proof that the dog handler used correct scientific procedures while employing the STU.

 Apart from the Kelly rule deficiency in this case, there is a foundational weakness in the dog identification evidence.   As has already been observed, “ ‘dog trailing is a lot different from dog scent recognition.’ ”  (People v. Mitchell, supra, 110 Cal.App.4th at p. 790, 2 Cal.Rptr.3d 49.)   Testimony showing that a dog pursued a fleeing suspect to the site where he was hiding is a clear-cut case of dog tracking, and-depending on the facts of each case-may be admissible.  (See People v. Malgren (1983) 139 Cal.App.3d 234, 237-238, 188 Cal.Rptr. 569, disapproved on other grounds in People v. Jones (1991) 53 Cal.3d 1115, 1144-1145, 282 Cal.Rptr. 465, 811 P.2d 757 [allowing evidence that the defendant was tracked from the victims' house to some bushes less than one mile from the house, with the dog on the trail within a half hour of the incident];  and People v. Craig (1978) 86 Cal.App.3d 905, 910-911, 915-917, 150 Cal.Rptr. 676 [allowing use of dog tracking evidence where the suspects were followed from their stolen getaway car to the point inside an apartment complex where they were detained];  but cf. People v. Gonzales (1990) 218 Cal.App.3d 403, 267 Cal.Rptr. 138 [reversing a conviction where a dog found the defendant in a vineyard less than one mile from a home that was burglarized half an hour earlier].)

A more difficult case is presented when the dog is not tracking a suspect but rather is given a scent from a gauze pad some length of time after an incident and is watched to see if the dog “shows interest” in various locales frequented by the defendant.   Showing interest in locations is a far cry from tracking a suspect and giving an unambiguous alert that the person has been located.   Furthermore, there is no proof that appellant ever touched the matchbook from which a scent was collected.   The matchbook was found in a parking lot used by many people, any one of whom could have thrown the matchbook on the ground.

 The prosecution cannot rely solely on anecdotes regarding the dog's capabilities.   Instead, a foundation must be laid from academic or scientific sources regarding (a) how long scent remains on an object or at a location;  (b) whether every person has a scent that is so unique that it provides an accurate basis for scent identification, such that it can be analogized to human DNA;  (c) whether a particular breed of dog is characterized by acute powers of scent and discrimination;  and (d) the adequacy of the certification procedures for scent identifications.  (People v. Mitchell, supra, 110 Cal.App.4th at pp. 791-792, 2 Cal.Rptr.3d 49.)   None of these foundational requirements were met in this case.

 2. Prejudice

 Even if the dog scent evidence was improperly admitted, we find no prejudice because there was abundant evidence pointing to appellant's guilt.   Appellant cohabited with Crystal Stahl, the victim in this case.   Appellant was violently abusive toward Stahl.   Coworkers and friends saw bruises and bite marks on Stahl's face, arms and neck.   Stahl's foster mother saw appellant attack Stahl.   A few weeks before Stahl was murdered, a friend tried to intervene while appellant was beating up Stahl and heard appellant warn Stahl that she would “never” leave him.   The police arrested appellant for domestic violence in mid-December 1999, after appellant attacked and bit Stahl when she refused to allow him into her apartment.

Stahl broke up with appellant in the weeks preceding her death, and started a new romantic relationship.   Appellant voiced anger because Stahl was “screwing another guy” and because she had him arrested for domestic violence.

A few days before Stahl's death, appellant told a group of friends that he was angry at Stahl.   Witnesses heard appellant declare that he was going to burn Stahl to death in her taxicab and avoid blame because taxicab drivers are regularly killed, so no one would suspect him.   Appellant also stated that he would go to Las Vegas after committing the murder.   In fact, Stahl was burned to death in her taxicab and appellant went to Las Vegas, where he arrived on the afternoon of January 6, the day of the murder.   The fire was intentionally set and was not the result of any mechanical or electrical failure.

On the night of Stahl's death, the taxicab dispatcher received a call from appellant asking that Stahl personally come and collect him at The Pit Stop. When advised that Stahl was busy, appellant indicated that he would wait until Stahl was free to come and collect him.   Though appellant used an assumed name, the dispatcher was absolutely certain that the caller was appellant, who had a habit of calling taxi dispatch 20 to 30 times per week asking to speak to Stahl.   The dispatcher readily recognized appellant's voice. Stahl set out to pick up appellant at approximately 12:20 a.m. She was never heard from again.   Her burning vehicle was seen at 12:50 a.m., only one mile and a half away from The Pit Stop.

After appellant returned to the Antelope Valley, he warned a friend not to tell anyone that he had seen appellant, indicating consciousness of guilt and a desire to evade authorities.

Given the compelling evidence of appellant's history of violence toward the victim, his announced intention to burn her to death in her taxicab, and  the taxicab dispatcher's absolute certainty that appellant was the last person that Stahl picked up shortly before her death, it is not reasonably probable that the jury would have reached a different result had the dog scent evidence been excluded.   Any error in admitting the evidence must be deemed harmless and is not a cause for reversal.   There was no miscarriage of justice.  (See People v. Heard (2003) 31 Cal.4th 946, 978, 4 Cal.Rptr.3d 131, 75 P.3d 53;  People v. Watson (1956) 46 Cal.2d 818, 835-837, 299 P.2d 243;  People v. Mitchell, supra, 110 Cal.App.4th at pp. 794-795, 2 Cal.Rptr.3d 49.)


The judgment is affirmed.


1.   People v. Kelly (1976) 17 Cal.3d 24, 130 Cal.Rptr. 144, 549 P.2d 1240, Frye v. United States (D.C.Cir.1923) 293 F. 1013.   Due to changes in federal law that supersede the Frye case, the rule is now referred to in California as the Kelly test or rule.  (People v. Bolden (2002) 29 Cal.4th 515, 545, 127 Cal.Rptr.2d 802, 58 P.3d 931.)


We concur:  DOI TODD and ASHMANN-GERST, J.

Copied to clipboard