by Adela Morris and Rita Martinez (CSST members)

This research project was initiated in January of 1997 as a response to a frequently asked question by agencies that use our services. How, we are asked, do the dogs define death? At what point during the decomposition process of a human body will the dogs demonstrate that they recognize the scent as post-mortem?

We set up this research project to begin to understand when and how our dogs can discriminate live scent from post-mortem scent. All of the dogs used in this project have been "cross-trained", that is trained in both the discipline of finding and indicating on live human scent and also on post-mortem (cadaver) scent. *

The general medical definition of death, "The irreversible cessation of respiratory and heart activity" (Spitz and Fisher, 1993), describes the exact moment a human being becomes a dead body. Physiologic changes begin immediately, and within the first two hours after death onset of lividity, chemical changes in the blood, relaxation of muscle tissue, and other changes have been documented. In fact, in some cases, lividity can be seen as soon as 15 minutes post-mortem. Between 2-4 hours after death, body temperature begins to drop (Haglund and Sorg, 1997), chemical changes in muscles begin and autolysis advances.

We want to begin to learn, understand, and document the minimum post mortem time interval for which our dogs can perceive the difference between live and dead scent. Our window into the vivid realm of canine olfaction is, obviously, the dog's body language and more precisely, the alert sequence. The dog uses the alert to intentionally communicate to us "I have found a smell that you want". We want to know the minimum interval at which our dogs can classify a scent as dead, as opposed to live, and when that classification is distinct enough that the dog is moved to perform the trained alert.


*All post-mortem scent samples consist of sterile gauze pads, (sealed until just prior to use) placed on abdominal area of decedent for exactly 20 minutes. Gauze pads are then placed in unused plastic bags and double sealed with packing tape. All post-mortem scent samples are handled with latex gloves, and at no time do these samples come in skin contact with live human scent.
*All Live Human scent samples are placed on the abdominal area of a living person for exactly 20 minutes, then placed in unused plastic bags and double sealed in the same manner as the post-mortem samples.

All trials are conducted using a lineup method:
Each consists of a line of three gauze pads - one exposed to post-mortem scent, one exposed to live human scent, and one sterile unused pad opened with gloved hands just prior to each trial.
All trials are done on a hard surface, either asphalt, cement, or hard packed dry dirt ground.
Each dog is given the "cadaver" command by its handler and walked through the lineup. The dog is then asked to choose and indicate to the handler which gauze contains post-mortem scent.
Every trial was "blind" to each handler working it, that is the handler had no prior knowledge of lineup sequence.
Dogs are given only a single choice per trial.
Choices are recorded as CORRECT or INCORRECT only.
Any dog that appeared unwilling to commit to a single sample was removed from that trial and the choice was recorded as Incorrect.
All samples are properly disposed of after each trial.

TRIALS BEGUN: January 1997
NUMBER OF DOGS USED: Five different dogs
POST-MORTEM INTERVAL RANGE: From 70 minutes to 3 days
NUMBER OF TRIALS COMPLETED: As of July 1997, total of 52 trials completed
PRELIMINARY RESULTS: The shortest post-mortem interval for which we received a correct response was one hour and 25 minutes. However, the post-mortem interval for which we received a consistently correct response from all dogs involved is 2.5 - 3 hours.

We emphasize our inability to control all variables during our project. It has been impossible for us to obtain scent samples and carry out the trials under strictly controlled conditions. We had no control over the environmental conditions under which our samples were gathered, manner and cause of death, nor could we exercise tight regulation over the quantity of samples of any one given time interval. Since there are periods in which our samples are few and far between, this project has no set deadline and will be ongoing for a time. Our intention is not to offer an absolute answer to the questions posed, but to initiate an ongoing dialogue and offer what we hope is a point of departure for other exciting research projects. This will shed light on the under-navigated and sometimes mysterious arena of canine olfaction applied to body recovery.

It is important to note what the dogs are not being asked to do, as well as what they are being asked to do. They are not choosing from a lineup of actual human tissue, but only from gauze pads for which great care has been taken to expose them only to dry skin regions, not to any blood or perceivable body fluids. Also note that the exposure time is exactly 20 minutes.

In addition, there are many obvious variables not addressed by this project. It will be exciting to undertake future studies that deal with such factors as difficulty of detection according to age, sex, and race of decedent and manner and cause of death.

Comments, suggestions and questions regarding this project are welcome.