Smart system for detecting owls: Automated technology identifies calls of Barred Owls, Boreal Owls, and Great Horned Owls

OLYMPUS DIGITAL CAMERA

An autonomous recording unit deployed in the field to record owl calls.

PhD student Julia Shonfield in Erin Bayne’s lab at the University of Alberta has developed an automated system for detecting owl calls, eliminating the need for researchers to spend nights in the field to survey owls. This new approach combines using audio recorders with a software program that can detect owl calls, called recognizers, and is similarly accurate and far more efficient than other, more traditional methods.

For more details of this work, see the university press release here:
https://www.ualberta.ca/science/science-news/2018/april/automated-system-detects-owl-calls

To hear Julia Shonfield talk about owls and her research, check out a clip of a radio interview on the local CBC Edmonton station: http://www.cbc.ca/listen/shows/radio-active/segment/15537010

“Utility of automated species recognition for acoustic monitoring of owls” by Julia Shonfield, Sarah Heemskerk, and Erin M. Bayne was published in the March 2018 issue of the Journal of Raptor Research, available at https://doi.org/10.3356/JRR-17-52.1.

Field Fun Friday

With the spring breeding season right around the corner, here is a photo of three Ferruginous Hawk chicks during a routine nest check last summer. Nick Parayko is studying these amazing raptors, in southern Alberta, for his Master’s project under Dr. Erin Bayne. Looks like fun!

Photo by: Nick Parayko

 

 

 

Male red squirrels kill other squirrels’ offspring to increase their own chances of having kids

I was walking in the Yukon one evening when I heard a commotion in the forest. At the time, I was working on my PhD under the supervision of Stan Boutin at the University of Alberta as part of the Kluane Red Squirrel Project. This meant that I spent a lot of time hiking on my study area monitoring the red squirrels that I was working with. It is common to hear squirrels calling because they are pretty chatty, but this commotion was different: this squirrel was very upset.

I walked quickly towards the calls, arriving to find a female squirrel yelling at her male next-door neighbor, who had intruded onto her territory. This is unusual, as squirrels live solitary lives and are usually respectful of each other’s boundaries. However, as I watched the male’s intentions soon became clear because within seconds of my arrival the male killed one of the female’s pups. I realized that I had just witnessed infanticide.

SONY DSC

Infanticide is when an adult kills the young of their own species. I later found a second dead pup from the same litter, whose wounds were consistent with the infanticide that I had watched. The whole litter eventually died and their mother, the female I had seen yelling at the male, later had a second litter that summer. I was able to show using genetics that the male who killed the pup was not the father of any of the pups in the first litter, but that he was the father of all of the pups in her second litter.

This is an example of sexually-selected infanticide: this is when a male kills another male’s offspring in order to increase the chances that he’ll be able to father kids of his own when the female breeds again. I was fascinated by this behaviour, so I decided to explore it in more detail in our paper that was published today in the journal Ecology. I found evidence that this behaviour is linked to fluctuations in white spruce cones, the main food that red squirrels eat.

My colleagues previously showed that red squirrels can predict the future cone availability. The cones mature in autumn and at that time they can be harvested by the squirrels to be stored in a cache on their territory called a midden. Squirrel pups strike out on their own in the fall and during a bumper crop, called a mast year, they have a really good chance of surviving the winter because there are lots of cones for them to harvest and cache. Very few pups are able to store enough food in non-mast years so few of them will survive. Having access to cones is really key for pups to survive.

SONY DSC

Squirrels are strategic and they can predict the future cone availability. During a mast year, the females will have more than one litter because they know that their pups will have access to cones once the fall arrives, and thus the pups will have a good chance of surviving. The females respond to masts in this way even though they breed in the spring but the cones aren’t available until autumn. In contrast, during non-mast years they will typically have only one litter.

I showed that litters die more frequently during mast years, suggesting that infanticide is more common during mast years. When litters die, their mother is more likely to have a second litter and will breed again sooner than if her litter had survived. So male red squirrels commit infanticide in mast years because the females will have that second litter, giving the males a second chance at fatherhood.

SONY DSC

 

Post and photos by: Jess Haines

Field Fun Friday

 

 

Boutin lab members, Sean Konkolics and Mike Peers, lead Edmonton families for a wildlife snow-tracking workshop hosted by Nature Alberta’s: Nature Kids program. The group spent a few hours in Strathcona Science Park looking at different wildlife tracks including coyote, fox, red squirrel, and snowshoe hare. Even though it was a chilly -16C, this was a great opportunity for kids from the city to learn about animal tracks, while getting outside in the winter! 

Mike shows the families a spot where a coyote killed a snowshoe hare.The City of Edmonton in the distance.

Special thanks to Nature Kids coordinators Alyssa Bohart and Emily Dong.

Photo credit: Alyssa Bohart

 

 

 

 

 

How does industrial noise affect owls in Northeastern Alberta?

Recently published research by Julia Shonfield, a PhD student in Erin Bayne’s lab:

Shonfield, J & Bayne, EM. 2017. The effect of industrial noise on owl occupancy in the boreal forest at multiple spatial scales. Avian Conservation and Ecology, 12 (2): 13. doi: 10.5751/ACE-01042-120213.

Photo 1 – Barred Owl, one of three species of owls I’m studying in this project.

Energy development creates several types of disturbance that can impact wildlife, including the physical footprint of the infrastructure and the chronic noise from facilities. Chronic noise sources can pose problems for animals that communicate vocally because the noise can mask important signals.

Owls use vocal communication to attract mates and defend territories, and hunt by listening for acoustic cues made by prey (e.g. mice scurrying along the forest floor). Chronic noise has been shown to negatively affect owl hunting success and ability to detect prey. So this suggests that noise can have negative effects on owls, but we don’t know whether this affects where owls are distributed on the landscape.

For part of my PhD work in Erin Bayne’s lab, I wanted to know if owls avoid chronic industrial noise sources. To get at this question, I used autonomous recordings units (ARUs) to survey for owls.

Photo 2 – An autonomous recording unit (ARU) all set up and ready to record some owls calling.

OLYMPUS DIGITAL CAMERA

These units were programmed to record every hour for 10 minutes throughout the night for up to 2 weeks at each site. I surveyed at three types of sites in northern Alberta:

  1. Chronic noise sites – with a compressor station or oil processing facility at the center of the site
  2. Intermittent noise sites – with a road bisecting the site
  3. No noise sites – with no traffic noise or industrial noise sources

I processed the recordings using recognizers I developed in a program called Song Scope to detect the calls 3 owl species found in northern Alberta: Barred owls, Great Horned owls and Boreal owls.

Photo 3 – An example of the call of the Great Horned owl, shown as a spectrogram in the program Song Scope.

Once I obtained the owl detections from the recordings, I analyzed the data using occupancy models. I analyzed the data at two spatial scales, at a larger scale that is roughly equivalent to an owl’s home range size, and a smaller scale representing an area within a home range.

Photo 4 – The two spatial scales that I analyzed to determine owl occupancy. The small white circles are ARU stations.

I found that the occupancy of all three owl species was not different between sites in the three noise categories at the larger scale. I also found that at the smaller scale, owl occupancy was not affected by industrial noise levels.

Photo 5 – Results of the owl occupancy models at the larger scale, showing occupancy estimates for each of the three noise categories of sites.

Take-home message:

At the spatial extent I assessed there was no evidence of noise effects on owls, suggesting that noise is not likely to have negatively affected owl populations at current noise levels.

  • Photos and blog post written by Julia Shonfield