Angler's Atlas

When MyCatch first launched in 2018, the first goal was to see if the app (and website) could collect reliable creel data from anglers. To keep it as simple as possible, we focused on collecting data on just four core variables:

Where did they fish?
When did they fish?
How long did they fish?
What did they catch (including zeroes)?

After we launched the project and began collecting the data, the next challenge was assessing the quality of the data. Were we getting data that was consistent with conventional creel surveys? Fortunately we had two conventional surveys taking place at the same time (spring/summer of 2018) on the Bow River and Oldman/Livingston Rivers in southern Alberta. This provided two good surveys that we could benchmark MyCatch data against.

Fig. 1. The Fisheries Management Zones in Alberta, Canada (panel a), and the Bow River and Oldman River system in southwestern Alberta where creel surveys were conducted in 2018 (panel b). Magenta points indicate the upstream (Bearspaw Dam) and downstream (Carseland Weir) limits of the creel survey area on the Bow River.

SPECIES COMPOSITION

As a starting point for the analysis, we compared species composition as a proportion of catch between the two methods. In both systems MyCatch data tracked closely with the data obtained through the creel surveys. Some differences were noted between cutthroat trout and rainbow trout on the Oldman / Livingston rivers, however we expect that had to do with hybridization between these two species. At the time of the study the option to select hybrids through the app was not available so anglers had to select one or the other.

Fig. 6. The relative composition of commonly caught species in catches from the MyCatch app or the 2018 creel surveys for (a) the Bow River (Creel N = 6105 fish, MyCatch N = 161 fish) and (b) the Oldman River system (Creel N = 2472 fish, MyCatch N = 240 fish). * Asterisks above the bars indicate significant differences at alpha = 0.05, and numbers above the bars are sample size. N = the total number of fish caught. BNTR = brown trout, Salmo trutta; BLTR = bull trout, Salvelinus confluentus; CTTR = cutthroat trout, Oncorhynchus clarki; MNWH = mountain whitefish, Prosopium williamsoni; RNTR = rainbow trout, Oncorhynchus mykiss. Cutthroat trout × rainbow trout hybrids (N = 73 for the creel and 0 for the app) in the Oldman River system were included in the CTTR category.

CATCH RATE

As part of the evaluation, a series of models were evaluated to find the most parsimonious model where catch rate data was compared between MyCatch and the two creel surveys. Through this model development, we were able to generate nearly identical catch rate predictions using MyCatch data for both systems. The red and blue lines in Figure 7 reflect the predicted catch rates for the creel surveys (red) and MyCatch (blue).

Fig. 7. Data and predictions for the catch per trip of (a) rainbow trout (Oncorhynchus mykiss) from the Bow River (Creel N = 2740 reports, MyCatch N = 91 reports) and (b) cutthroat trout (Oncorhynchus clarki) from the Oldman River system (Creel N = 790 reports, MyCatch N = 43 reports), in relation to the time spent fishing. The panels illustrate the most parsimonious model (Model 8: Catch ∼ OT | 1, in both cases) when MyCatch catches were compared to creel catches pooled over all angler segments. Dots represent the raw data and lines are the model predictions. The dot size indicates the proportion of trips in a group. Model details can be found in Table 2. [Colour online.]

DISCUSSION

Overall, the creel comparisons for both the Bow River and the Oldman/Livingston Rivers show the MyCatch data can yield similar species compositions from anglers, as well as similar catch rates. Much work is still needed on this front, however as a first step it shows the potential for using app data to supplement ongoing creel surveys.

The real advantage to MyCatch comes from the scale of data that can be collected compared to creel surveys. Where a creel survey is very expensive and covers a very limited geographic area, MyCatch is not constrained in this way, and the volume of data is exponentially greater covering hundreds of waterbodies across the landscape. Figure 2 below shows the data collected across the province of Alberta in 2018, when the two creel surveys were conducted.

Figure 2. (a) Distribution of MyCatch-registered anglers (N = 369 anglers) from Alberta by residence, and the distribution of the 379 waterbodies that anglers reported fishing using the MyCatch app or website by (b) number of trips (2217 trips), and (c) and hours fished (N = 9767 hours). Locations for larger streams and rivers are represented by a single position rather than fishing locations. Black lines represent the borders of Alberta’s three Fisheries Management Units, green lines represent the borders of national parks, and gray lines represent the borders of Alberta’s ten Fisheries Management Watershed Units (approximate because this spatial layer is not publicly available). PP = Parkland Prairie, ES = Eastern Slopes, NB = Northern Boreal.

These results were published in the Canadian Journal of Fisheries and Aquatic Sciences in 2021. You can access to the full paper here.

Creel Comparisons with MyCatch

Comparing MyCatch with Fall Index Netting — 2018

MyCatch Benchmarking with DFO Survey Data

Creel Comparisons with MyCatch

Read Related Articles

Comparing MyCatch with Fall Index Netting — 2018

MyCatch Benchmarking with DFO Survey Data