Emily McAuley

PhD Candidate

Senior Supervisor: Ronald C. Ydenberg


  • BSc (Honours), Carleton University, Ottawa, ON
  • MSc in Biology, Carleton University, Ottawa, ON
  • Director, WildResearch, Vancouver, BC


Ronald C. Ydenberg

Thesis Title

Do introduced fish lower the quality of harlequin duck breeding streams?

Areas of Research

Effects of fish on harlequin duck breeding behaviour
My Ph.D. project investigates the effect of introduced fish on harlequin duck, Histrionicus histrionicus, breeding behaviour and success, and the resulting effects on their population. Harlequin ducks, along with other sea ducks species, have been experiencing a slow, but steady decline for decades. In eastern North America, the number of harlequin ducks was estimated to be about 1500 in 1990, and the species was subsequently listed as endangered by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC).  This listing was downgraded to a Species of Special Concern in 2001, due to population stability and a marginal increase in numbers. The western population is Yellow-listed in B.C. and Alberta, where there is concern for its stability due to long-term declines in numbers, and is considered a Sensitive Species in most of the north-western U.S.
The widespread introduction of fish to historically fishless streams may be contributing to the declines noted in harlequin ducks, as well as other sea duck species. There are many streams which fish could not have naturally colonized due to the presence of natural barriers, such as waterfalls. However, they have since been introduced to these waters, sometimes accidentally, but often through government-sanctioned stocking programs. The presence of fish induces anti-predator behaviours in aquatic invertebrates, such as decreased drifting behaviour, increased hiding, and a switch from diurnal or aperiodic activity to nocturnal activity. All of these behavioural changes reduce the availability of this key resource to harlequin ducks, which are primarily diurnal, visual predators. Decreased nutrient availability can lead to increased female mortality during the incubation period and/or decreased breeding success and juvenile survival. Ducks may also compensate for lower nutrient availabilities by breeding at decreased densities. If high quality habitat is limited, this could lead to fewer ducks breeding overall, meaning lower overall recruitment within the population.

Why use more than one sexual signal?
My M.Sc. project investigated the reason for multiple sexual signals, using the field cricket as a model organism. Multiple sexual signals are hypothesized to indicate (1) separate aspects of condition (multiple messages) or (2) independently signal overall condition (redundant signals).  Multiple messages should only respond to specific changes in condition (those for which they signal). As such, during selection, individuals should prioritize the signals that are indicative of the aspects of condition in which they are most interested. Redundant signals, on the other hand, should all respond similarly to any change in condition; thus, individuals should weigh all signals equally.  
In field crickets, males use acoustic signals (calls) to attract females, who select mates based on different components of these calls. I tested whether these call components are used as multiple messages or redundant signals in Texas, Gryllus texensis, and Jamaican, G. assimilis, field crickets. I manipulated juvenile and adult condition independently in male crickets via dietary limitation and measured the effects on calling behaviour. In both species, call components appeared to be acting as multiple messages; most were unaffected by changes in either juvenile or adult condition and those that were affected responded only to changes in one or the other, but not both. I then tested whether female condition affected which call components were most important to them. Gryllus texensis females prioritized different call components depending on their own body condition. Gryllus assimilis females also exhibited priorities, but these were unaffected by condition. Overall, the multiple messages hypothesis was supported.