Bumpy Ride

Lockdown easing continues, despite the ominous rise in Delta-variant case numbers and we are crossing our fingers that the vaccination programme will be enough to hold the line.

With a few concessions to social distancing and mask wearing, the SARDA training weekends have been getting back to normal, and July’s offering brought the familiar larger groups and two days of training back onto the menu.

Venn and I have been working on our 3-person line-ups, the final phase of the scent-discrimination part of our training. As a reminder, this is where Venn has to select the correct target from a line of 3 people spaced about 10m apart, on the basis of a scent item.

Things seemed to be going reasonably well. We attended the June training with our laurel wreathes packed, ready to be carried at shoulder-height around SARDA base having been signed off as completing this section of training.

Alas, no.

We needed to get3 out of 3 line-ups correct, and in the event, we got 2 out of 4. Puzzled and a little downtrodden (pride, fall etc.) we went home to ponder our next course of action. More practice seemed the obvious answer. However, the more we did, the worse it got. We moved from false-indications (on the wrong target), to “I’m confused” (silent sitting) to no indication at all.

In the absence of any better ideas, we stopped scent-work completely for the 10 days leading up to the July training weekend, in the hope of some kind of reset.

At training, we started out with a formal exam-conditions 3 person line-up with the target unknown to me. Venn indicated on the wrong target. Bums!

More for my sanity than anything else, I set up a human-swab search on a similar bit of ground. She was keen to search, and found the tiny scrap of scent without issue.

After some head scratching, we tried her on a single person line-up: literally a person sat in the middle of an area of meadow. This was done off the lead to minimise any unconscious influence I was having on her behaviour. From a visual perspective, this looks crazy; how could you not see the person? However, when she is working, Venn is using her nose not her eyes and so where you and I might walk directly to the target, Venn is working the scent. She was still tending to indicate a good way down-wind of the target, but at least it was something. We then encouraged her in to the target who had her toys and food rewards. It made me laugh every time to see her do a massive double-take when she realised there was a person there. She was completely oblivious.

We repeated this approach a good number of times over the weekend and the outcome was similar each time. She indicated a little further down-wind than we would like, and when we took her in to the target, she was genuinely surprised that there was a person there.

When hunting, dogs naturally follow scent to reach the strongest concentration where they will hopefully find their prey. At face value, it is therefore hard to understand why Venn is not following the scent gradient to find the person at the end, but instead indicating at distance. Some dogs, such as Bryn, make the transition from searching for small amounts of scent to larger amounts without issue, but not Venn.

So this is my theory: the game we have taught her is to find trace amounts of scent, and that is what she is continuing to do. When she is looking for a swab, she finds it no problem, but when confronted with a huge concentration of scent in the shape of a person, she is working around the periphery of the scent cone to find those traces, or she is confused or both. If she indicates on trace scent (which could be over a large part of the working area, including near other targets) and then gets told “no”, she would rightly get confused about what was being asked of her.

To try and break this pattern in favour of following the gradient to its source, I devised a new game. Using a 30m x 30m area, I hid a large scent item (person or big pile of clothing) and set Venn to find it. Next, I hid the same target in a different position within the same, now contaminated, search area. The only way Venn could find the target would be to go to the strongest source. This process was repeated up to 5 times in quick succession, and with each go, the background contaminating scent was getting higher. She seemed to manage this no problem.

We’ve done this exercise for about 5 sessions now. The first run of each set still results in a slightly further-away indication, but the gap with the target closes with each run. In one session, we introduced a decoy on run 3, and two decoys on run 5. Again, good work and correct indications, though she was flagging towards the end of this one. Too hot.

We shall see.

We’ve not thrown those laurel garlands away just yet.