4 Results and Discussion
Results of Phase 1 and Phase 2 assessments are summarized in Figure 4.1 with additional details provided in sections below.
##make colors for the priorities
pal <-
leaflet::colorFactor(palette = c("red", "yellow", "grey", "black"),
levels = c("High", "Moderate", "Low", "No Fix"))
pal_phase1 <-
leaflet::colorFactor(palette = c("red", "yellow", "grey", "black"),
levels = c("High", "Moderate", "Low", NA))
map <- leaflet::leaflet(height=500, width=780) |>
leaflet::addTiles() |>
# leafem::addMouseCoordinates(proj4 = 26911) |> ##can't seem to get it to render utms yet
# leaflet::addProviderTiles(providers$"Esri.DeLorme") |>
leaflet::addProviderTiles("Esri.WorldTopoMap", group = "Topo") |>
leaflet::addProviderTiles("Esri.WorldImagery", group = "ESRI Aerial") |>
leaflet::addPolygons(data = wshd_study_areas, color = "#F29A6E", weight = 1, smoothFactor = 0.5,
opacity = 1.0, fillOpacity = 0,
fillColor = "#F29A6E", label = wshd_study_areas$watershed_group_name) |>
# `wshds` is NULL in a season with no Phase 2 habitat confirmation sites
(\(m) if (is.null(wshds) || nrow(wshds) == 0) m else
m |> leaflet::addPolygons(data = wshds, color = "#0859C6", weight = 1, smoothFactor = 0.5,
opacity = 1.0, fillOpacity = 0.25,
fillColor = "#00DBFF",
label = wshds$stream_crossing_id,
popup = leafpop::popupTable(x = dplyr::select(wshds |> sf::st_set_geometry(NULL),
Site = stream_crossing_id,
elev_site:area_km),
feature.id = F,
row.numbers = F),
group = "Phase 2"))() |>
leaflet::addLegend(
position = "topright",
colors = c("red", "yellow", "grey", "black"),
labels = c("High", "Moderate", "Low", 'No fix'), opacity = 1,
title = "Fish Passage Priorities") |>
# `assess_type_*` is recorded as "yes" in some regions and "Yes" in others, so
# compare case-insensitively. The layer is skipped outright when nothing matches -
# leafpop::popupTable() indexes mat[1, 1] and fails on a zero-row table.
(\(m) if (nrow(dplyr::filter(tab_map_phase_1, tolower(assess_type_phase1) == "yes" | tolower(assess_type_reassessment) == "yes")) == 0) m else
m |>
leaflet::addCircleMarkers(data=dplyr::filter(tab_map_phase_1, tolower(assess_type_phase1) == "yes" | tolower(assess_type_reassessment) == "yes"),
label = dplyr::filter(tab_map_phase_1, tolower(assess_type_phase1) == "yes" | tolower(assess_type_reassessment) == "yes") |>
dplyr::pull(pscis_crossing_id),
# label = tab_map_phase_1$pscis_crossing_id,
labelOptions = leaflet::labelOptions(noHide = F, textOnly = TRUE),
popup = leafpop::popupTable(x = dplyr::select((tab_map_phase_1 |> sf::st_set_geometry(NULL) |> dplyr::filter(tolower(assess_type_phase1) == "yes" | tolower(assess_type_reassessment) == "yes")),
Site = pscis_crossing_id, Priority = priority_phase1, Stream = stream_name, Road = road_name, `Habitat value`= habitat_value, `Barrier Result` = barrier_result, `Culvert data` = data_link, `Culvert photos` = photo_link, `Model data` = model_link),
feature.id = F,
row.numbers = F),
radius = 9,
fillColor = ~pal_phase1(priority_phase1),
color= "#ffffff",
stroke = TRUE,
fillOpacity = 1.0,
weight = 2,
opacity = 1.0,
group = "Phase 1"))() |>
# survey tracks are absent until the season's GPS tracks are cleaned and tagged
(\(m) if (is.null(habitat_confirmation_tracks) || nrow(habitat_confirmation_tracks) == 0) m else
m |> leaflet::addPolylines(data=habitat_confirmation_tracks,
opacity=0.75, color = '#e216c4',
fillOpacity = 0.75, weight=5, group = "Phase 2"))() |>
# `photo_metadata` is NULL when no photos are tagged for publication
(\(m) if (is.null(photo_metadata) || nrow(photo_metadata) == 0) m else
m |> leaflet::addAwesomeMarkers(
lng = as.numeric(photo_metadata$gps_longitude),
lat = as.numeric(photo_metadata$gps_latitude),
popup = leafpop::popupImage(photo_metadata$url, src = "remote"),
clusterOptions = leaflet::markerClusterOptions(),
group = "Phase 2"))() |>
# `tab_map_phase_2` is NULL in a season with no Phase 2 habitat confirmation sites
(\(m) if (is.null(tab_map_phase_2) || nrow(tab_map_phase_2) == 0) m else
m |>
leaflet::addCircleMarkers(
data=tab_map_phase_2,
label = tab_map_phase_2$pscis_crossing_id,
labelOptions = leaflet::labelOptions(noHide = T, textOnly = TRUE),
popup = leafpop::popupTable(x = dplyr::select((tab_map_phase_2 |> sf::st_drop_geometry()),
Site = pscis_crossing_id,
Priority = priority,
Stream = stream_name,
Road = road_name,
`Habitat (m)`= upstream_habitat_length_m,
Comments = comments,
`Culvert data` = data_link,
`Culvert photos` = photo_link,
`Model data` = model_link),
feature.id = F,
row.numbers = F),
radius = 9,
fillColor = ~pal(priority),
color= "#ffffff",
stroke = TRUE,
fillOpacity = 1.0,
weight = 2,
opacity = 1.0,
group = "Phase 2"
)
)() |>
leaflet::addLayersControl(
baseGroups = c(
"Esri.DeLorme",
"ESRI Aerial"),
overlayGroups = c("Phase 1", "Phase 2"),
options = leaflet::layersControlOptions(collapsed = F)) |>
leaflet.extras::addFullscreenControl() |>
leaflet::addMiniMap(tiles = leaflet::providers$"Esri.NatGeoWorldMap",
zoomLevelOffset = -6, width = 100, height = 100)
mapFigure 4.1: Map of fish passage and habitat confirmation results
4.1 Site Assessment Data
A summary of fish passage assessment procedures conducted through SERNbc since 2020 — including assessments, habitat confirmations, design, remediation, monitoring, fish sampling, eDNA sampling, and drone imagery — is provided in Appendix - Assessment Data Summary.
Orthoimagery and elevation model rasters generated since 2020 are catalogued in Appendix - UAV Imagery.
4.2 Climate Departure
Four findings from the regional climate-departure analysis carry to fish-passage prioritisation.
Warming is broad and significant region-wide. All six ecoregions of the study area warmed about 2.1 °C over the recent decade relative to the pre-warming reference — annual mean (+2.1 °C), daytime maximum (+1.9 °C), and overnight minimum (+2.2 °C) all tell the same story (Mann-Kendall p < 0.001 in every ecoregion). The rate has been steady rather than accelerating — the 1981–present slope (0.026 °C/yr) is shallower than the 1951–present slope (0.035 °C/yr). What is notable here is how little the signal varies across the study area: cumulative warming since 1951 spans only 0.28 °C between the most and least affected ecoregion, so the warming applies to every watershed group alike and the thermal question falls to elevation and reach scale — see Appendix - Climate Departure Figure 5.11.
The atmosphere is drying despite flat precipitation. Vapour pressure deficit rose 0.34 hPa region-wide, significant in every ecoregion. Annual precipitation rose +2 % regionally but the long-term trend is not statistically significant in any ecoregion; soil moisture is essentially flat because the warmer atmosphere is drinking the modest precipitation increase back through evaporation and transpiration.
The snowpack signal is about timing, and about volume. Total annual snowfall is essentially unchanged (-8 %, not significant), but annual snow water equivalent is down 20 % and that decline is significant (p = 0.007). The sharpest signals are on the summer side and at the midpoint: summer SWE collapsed 60 % and the snowmelt midpoint moved 15 days earlier (p < 0.001 regionally and in every ecoregion). The counterbalancing rise is concentrated in spring, which gained 52 % and is highly significant; winter snowmelt rose a comparable 46 % but does not clear significance on its own. The freshet is arriving more than two weeks earlier and the high-elevation snowpack that historically lingered into summer no longer does.
Day-night asymmetry is present. Overnight minimums warmed 2.2 °C while daytime maximums warmed 1.9 °C — overnight outpacing daytime is the textbook fingerprint of greenhouse warming (Karl et al. 1993). Mean diurnal range has narrowed by 0.3 °C across the record.
For barrier prioritisation these signals cut both ways. Where upstream reaches are cold-limited, added growing-degree-days can raise the intrinsic potential of habitat above the barrier and increase the value of restoring passage; where upstream reaches sit closer to the upper edge of the cold-water thermal niche, the same warming reduces usable habitat and lowers the payoff. Because the departure is so even across this study area, which of those applies is an elevation and reach-scale question rather than a watershed-group one. The six-ecoregion signal is mapped onto the 5 watershed groups via a percentage crosswalk in Appendix - Climate Departure.
4.3 Planning — Habitat, Connectivity and Floodplain Modelling
4.3.1 Habitat Modelling
Habitat modelling from bcfishpass including access model, linear spawning/rearing habitat model and lateral habitat
connectivity models for watershed groups within our study area were updated for the spring of 2025 and are included
spatially in the collaborative GIS project. A snapshot of these outputs related to each modeled and PSCIS stream
crossing structure are also included within an sqlite database within this year’s project reporting/code repository here.
4.3.2 Statistical Support for Habitat Modelling
The statistical modelling underpinning these outputs — channel-width prediction, and Bayesian Spatial Stream Network modelling of stream temperature with Poisson Consulting (Hill et al. 2024; Hill et al. 2025) — is described in the Methods chapter. The 2025 effort was carried out in this study area, extending the approach to a larger network and adding Growing-Season Degree Days as the productivity-relevant thermal metric, with observations drawn from water-temp-bc (Irvine [2025] 2025b). Both studies are linked in Attachment - Water Temperature Modelling.
These predictions are the route by which thermal condition enters prioritization: reaches too cold to support production can be discounted from intrinsic habitat models, and reaches becoming warm enough to support it recognised as gaining value.
4.3.3 Floodplain Delineation
Modelled functional floodplain in the Bulkley River Watershed Group covers 48,344 ha — 6.2 % of the 7,762 km² watershed group — attached to 1,587 km of coho accessible order-3+ streams. Within that footprint sit 180 lakes and 345 mapped wetland polygons — the off-channel inventory that crossing remediations in these reaches stand to reconnect or leave severed, depending on whether design accounts for the floodplain context around the structure. The full metric table, watershed-scale map, and a detail map of the valley south of Smithers are in Appendix - Floodplain Delineation.
4.4 Fish Passage Assessments
Field assessments were conducted between September 29, 2025 and October 04, 2025 by Allan Irvine, R.P.Bio. The 2025 field season was directed at effectiveness monitoring and environmental DNA sampling rather than at new fish passage assessments. A total of 1 crossings were assessed, all of them reassessments of structures already inventoried in PSCIS. No Phase 1 assessments were completed.
Reassessments update the provincial record where a structure has been replaced or remediated, or where the existing PSCIS entry no longer reflects conditions at the crossing. They do not add new crossings to the inventory, so they do not carry the cost estimates and priority rankings that accompany Phase 1 results. The amalgamated assessment record for the study area, spanning 2020 to 2025, is presented in Appendix - Assessment Data Summary.
4.5 Habitat Confirmation Assessments
No Phase 2 habitat confirmation assessments were conducted in 2025. Habitat confirmations characterise the quantity and quality of habitat upstream and downstream of a crossing prioritised during Phase 1 assessment; with no Phase 1 work completed this season, none were scheduled.
Habitat confirmation results from previous years in the Bulkley River, Kispiox River, Kalum River, Morice River, and Zymoetz River watershed groups are carried in the amalgamated record in Appendix - Assessment Data Summary, and in the reporting for those years.
## Environmental DNA (eDNA) Sampling
New in 2025, environmental DNA (eDNA) sampling was incorporated into the program at both previously assessed and newly added sites. A total of 20 collection events were undertaken across 11 streams, comprising 19 real environmental samples, 1 field blanks, and 0 office blanks collected for quality assurance.
Per-species detection results across the 19 real environmental samples are summarized in Table 4.1. Detections use the four-droplet call threshold applied by the laboratory; results below that threshold, including two- and three-droplet results that were retested without reaching the cutoff, are reported as not detected. Laboratory methods and per-sample results are in the UNBC technical report, available here. Site-level detail — the per-site detection table, field blanks, and retests — is provided in Appendix - Environmental DNA Results, and the interactive eDNA map shows toggleable per-species layers.
edna_summary |>
fpr::fpr_kable(
caption_text = "Per-species summary of eDNA results across real environmental samples in the project area (excludes field and office blanks).",
scroll = FALSE
)| Species | Code | Sites tested | Detected (≥4 droplets) | Not detected |
|---|---|---|---|---|
| Rainbow Trout | RAIN | 13 | 9 | 4 |
| Bull Trout | BULT | 18 | 7 | 11 |
| Coho Salmon | COHO | 19 | 3 | 16 |
| Chinook Salmon | CHIN | 17 | 0 | 17 |
| Sockeye Salmon | SOCK | 9 | 0 | 9 |
4.6 Engineering Design
No new designs were commissioned in 2025 due to uncertainty related to forest harvesting activities and lack of funding for the 50% costs of replacing structures. All sites with a design for remediation of fish passage can be found in Table 5.10 of Appendix - Assessment Data Summary by filtering using Design = yes.
4.7 Monitoring
In 2025, baseline or follow up monitoring data was gathered through completion of an effectivness montoring form (sites where remediation has been completed) and/or through eDNA sampling and aquistion of aerial imagery at the following sites (Table 4.2:
Effectiveness monitoring was conducted on Peacock Creek (PSCIS crossing 197962) on the Morice FSR, where the oval culvert was replaced with a bridge in 2021. Environmental DNA sampling upstream and downstream of the crossing confirmed coho, rainbow trout and bull trout in the reach. Results are presented in the Peacock Creek - 197962 - Monitoring Appendix.
Effectiveness monitoring was conducted on Waterfall Creek (PSCIS crossing 124421) at 11th Avenue, where the culvert was removed by the Gitksan Watershed Authorities. Riprap at the site has been buried and seeded, with good regrowth of grasses established and shrub planting planned. Results are presented in the Waterfall Creek - 124421 - Monitoring Appendix.
# started but not sure if this is going to be helpful.
form_monitoring |>
sf::st_drop_geometry() |>
dplyr::select(
`PSCIS ID` = pscis_crossing_id,
Stream = stream_name,
Road = road_name,
`Crossing type` = crossing_type,
`Crossing subtype` = crossing_subtype,
`Diameter or span (m)` = diameter_or_span_meters,
`Length or width (m)` = length_or_width_meters,
`Assessment comments` = assessment_comment
) |>
fpr::fpr_kable(caption_text = 'Summary of monitoring sites.', scroll = F)| PSCIS ID | Stream | Road | Crossing type | Crossing subtype | Diameter or span (m) | Length or width (m) | Assessment comments |
|---|---|---|---|---|---|---|---|
| 197962 | Peacock Creek | Morice FSR | Open Bottom Structure | BRIDGE | 20 | 10 | Bridge installed in 2020 or 2021. Bridge span estimated from photos. |
| 124421 | Waterfall Creek | 11th Avenue | Open Bottom Structure | FORD | – | – | Culvert was removed by Gitksan Watershed Authorities. Site riprap was buried with dirt and seated with good regrowth of grasses. Shrub planting is planned for this fall. |
4.8 Collaborative GIS Environment
Fieldwork activities and background spatial layers are maintained in a collaborative QGIS and Mergin Maps environment (sern_skeena_2023), which field crews and partners use to view, edit and add data. The layers loaded at the time of writing are listed in Appendix - Collaborative GIS Layers.