diff --git a/DESCRIPTION b/DESCRIPTION
index 5e40221..19139d2 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: link
 Title: Stream Network Habitat Interpretation (Experimental)
-Version: 0.21.0
+Version: 0.22.0
 Authors@R: c(
     person("Allan", "Irvine", , "airvine@newgraphenvironment.com",
            role = c("aut", "cre"),
@@ -31,7 +31,7 @@ Suggests:
     bcdata,
     digest,
     dplyr,
-    fresh (>= 0.26.0),
+    fresh (>= 0.27.5),
     lintr,
     mockery,
     sf,
diff --git a/NEWS.md b/NEWS.md
index 2a68602..ca450e1 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,3 +1,19 @@
+# link 0.22.0
+
+Wires `fresh::frs_order_child` into the pipeline as link methodology — small streams plugging directly into large rivers can be credited as rearing despite low/missing FWA channel-width estimates. Closes [fresh#158](https://github.com/NewGraphEnvironment/fresh/issues/158) on the link side.
+
+- Four new per-species columns in `dimensions.csv` (both bundles), all opt-in via `rear_stream_order_bypass: yes/no`:
+  - `rear_stream_order_parent_min` — min order at the trib BLK's mouth confluence (default 5, matches bcfp)
+  - `rear_stream_order_child_min` — lower bound on segment's own stream_order (default 1)
+  - `rear_stream_order_child_max` — upper bound on segment's own stream_order (default 1)
+  - `rear_stream_order_distance_max` — cap (m) on distance from trib mouth (empty = no cap)
+- `lnk_rules_build` emits the values into a `channel_width_min_bypass:` block on the rear stream-edge rule. `lnk_pipeline_classify` reads the block and calls `frs_order_child` per species post-classification, gated on `rear_stream_order_bypass`.
+- Both bundles ship `bypass: no` for all species — infrastructure is parametric and tested but disabled by default. Re-enable per species via `dimensions.csv`. The 4-WSG regression (HARR / HORS / LFRA / BABL) is byte-identical to the pre-#96 baseline with bypass=off, confirming the wiring is purely additive when disabled.
+- Updates `inst/extdata/configs/dimensions_columns.csv` xref doc with all four new columns and refreshes the `rear_stream_order_bypass` entry (was stale — said "currently inert").
+- Bumps fresh dep to `>= 0.27.5` for the renamed bypass YAML schema (`stream_order` → `stream_order_min` + `stream_order_max`).
+
+Related: [link#23](https://github.com/NewGraphEnvironment/link/issues/23) (CH spawning misread, closed not-a-bug). PWF for the wire-up at `planning/active/`.
+
 # link 0.21.0
 
 Closes [#87](https://github.com/NewGraphEnvironment/link/pull/94). Default-bundle SK upstream-spawn now credits any spawn-eligible segment upstream of and accessible from a qualifying rearing waterbody, dropping bcfishpass's restrictive cluster + lake-adjacency gate. bcfishpass-bundle SK keeps the gate (parity preserved).
diff --git a/R/lnk_pipeline_classify.R b/R/lnk_pipeline_classify.R
index d135b66..2ec204e 100644
--- a/R/lnk_pipeline_classify.R
+++ b/R/lnk_pipeline_classify.R
@@ -131,6 +131,43 @@ lnk_pipeline_classify <- function(conn, aoi, cfg, loaded, schema,
       verbose = FALSE)
   }
 
+  # fresh#158 stream-order bypass: post-classification, credit direct
+  # tributaries of large-order rivers as rearing even when channel
+  # width is below threshold. Mirrors bcfp's hard-coded
+  # `stream_order_parent >= 5 AND stream_order = 1` predicate in
+  # load_habitat_linear_<sp>.sql for BT/CH/CO/ST/WCT.
+  #
+  # Per-species opt-in driven by `dimensions.csv::rear_stream_order_bypass`,
+  # which `lnk_rules_build` propagates into rules.yaml as
+  # `rear[].channel_width_min_bypass = list(stream_order, stream_order_parent_min)`.
+  # We detect that field's presence on any rear rule and call
+  # `frs_order_child` with the embedded parent_order threshold.
+  for (sp in species) {
+    rear_rules <- params[[sp]][["rules"]][["rear"]]
+    bypass <- NULL
+    for (rr in rear_rules) {
+      if (!is.null(rr[["channel_width_min_bypass"]])) {
+        bypass <- rr[["channel_width_min_bypass"]]
+        break
+      }
+    }
+    if (!is.null(bypass)) {
+      pom    <- bypass[["stream_order_parent_min"]] %||% 5L
+      cs_min <- bypass[["stream_order_min"]]
+      cs_max <- bypass[["stream_order_max"]]
+      dmax   <- bypass[["distance_max"]]
+      fresh::frs_order_child(conn,
+        table   = "fresh.streams",
+        habitat = "fresh.streams_habitat",
+        species = sp,
+        parent_order_min = pom,
+        child_order_min = cs_min,
+        child_order_max = cs_max,
+        distance_max = dmax,
+        verbose = FALSE)
+    }
+  }
+
   invisible(conn)
 }
 
diff --git a/R/lnk_rules_build.R b/R/lnk_rules_build.R
index 348c11b..39320db 100644
--- a/R/lnk_rules_build.R
+++ b/R/lnk_rules_build.R
@@ -86,6 +86,22 @@ lnk_rules_build <- function(csv,
       tolower(trimws(dimensions$rear_stream_order_bypass)) == "yes"
   }
 
+  # Optional: per-species `stream_order_parent_min` for the bypass.
+  # Default 5L matches bcfishpass's hard-coded predicate; the column lets
+  # callers tune the threshold without editing the rules YAML by hand.
+  has_sopm <- "rear_stream_order_parent_min" %in% names(dimensions)
+
+  # Optional: per-species child-order range for the bypass. Both default to
+  # 1L (matches bcfp). `_min` and `_max` map to fresh's frs_order_child
+  # `child_order_min` and `child_order_max` arguments.
+  has_csmin <- "rear_stream_order_child_min" %in% names(dimensions)
+  has_csmax <- "rear_stream_order_child_max" %in% names(dimensions)
+
+  # Optional: per-species `distance_max` for the bypass — caps the bypass
+  # to the lower N metres of each direct-trib BLK (segment's
+  # downstream_route_measure <= distance_max). Empty → no cap (whole BLK).
+  has_sodm <- "rear_stream_order_distance_max" %in% names(dimensions)
+
   # Optional: requires_connected columns (value is the habitat type, not yes/no)
   has_spawn_rc <- "spawn_requires_connected" %in% names(dimensions)
   has_rear_rc <- "rear_requires_connected" %in% names(dimensions)
@@ -266,10 +282,35 @@ lnk_rules_build <- function(csv,
       if (!is.na(rlhm)) lake_rule$lake_ha_min <- rlhm
       rear_rules[[1]] <- add_rc(lake_rule, rear_rc, rear_cdm)
     } else {
-      # Stream order bypass: first-order streams with parent order >= 5
-      # bypass rearing channel_width_min
+      # Stream order bypass: first-order streams with parent order
+      # >= stream_order_parent_min bypass rearing channel_width_min.
+      # Threshold defaults to 5L (bcfishpass parity); per-species
+      # `rear_stream_order_parent_min` column overrides when present
+      # and numeric.
       soe_bypass <- if (has_soe && d$rear_stream_order_bypass) {
-        list(stream_order = 1L, stream_order_parent_min = 5L)
+        # Helper: read a positive integer column with default fallback
+        read_pos_int <- function(raw, default) {
+          if (is.null(raw) || is.na(raw) ||
+              nchar(trimws(as.character(raw))) == 0L) return(default)
+          n <- suppressWarnings(as.integer(raw))
+          if (is.na(n) || n < 1L) return(default)
+          n
+        }
+        pom    <- read_pos_int(if (has_sopm)  d$rear_stream_order_parent_min else NULL, 5L)
+        cs_min <- read_pos_int(if (has_csmin) d$rear_stream_order_child_min  else NULL, 1L)
+        cs_max <- read_pos_int(if (has_csmax) d$rear_stream_order_child_max  else NULL, 1L)
+        out <- list(stream_order_min = cs_min,
+                    stream_order_max = cs_max,
+                    stream_order_parent_min = pom)
+        if (has_sodm) {
+          raw_dm <- d$rear_stream_order_distance_max
+          if (!is.null(raw_dm) && !is.na(raw_dm) &&
+              nchar(trimws(as.character(raw_dm))) > 0) {
+            dm <- suppressWarnings(as.numeric(raw_dm))
+            if (!is.na(dm) && dm > 0) out$distance_max <- dm
+          }
+        }
+        out
       } else {
         NULL
       }
diff --git a/data-raw/maps/_lnk_map_compare.R b/data-raw/maps/_lnk_map_compare.R
index 4dd88b0..0e141a2 100644
--- a/data-raw/maps/_lnk_map_compare.R
+++ b/data-raw/maps/_lnk_map_compare.R
@@ -140,7 +140,10 @@ lnk_map_compare <- function(wsg, species, habitat,
     function() {
       c <- conn_local(); on.exit(DBI::dbDisconnect(c))
       sf::st_read(c, query = sprintf("
-        SELECT linear_feature_id, blue_line_key, edge_type, geom
+        SELECT linear_feature_id, blue_line_key,
+               downstream_route_measure, edge_type,
+               stream_order, gradient,
+               gnis_name, geom
         FROM whse_basemapping.fwa_stream_networks_sp
         WHERE watershed_group_code = '%s'", wsg))
     })
@@ -201,19 +204,23 @@ lnk_map_compare <- function(wsg, species, habitat,
         " | ", round(length_metre), "m"))
   }
 
-  axis_combined <- dplyr::bind_rows(
-    bcfp_axis |> dplyr::filter(category == "bcfp_only") |> mk_label("bcfp-only"),
-    link_axis |> dplyr::filter(category == "link_only") |> mk_label("link-only"),
-    link_axis |> dplyr::filter(category == "both")      |> mk_label("both"))
+  # Split axis layers per category so each is independently toggleable
+  # in the layers control (link_only / bcfp_only / both).
+  axis_bcfp_only <- bcfp_axis |>
+    dplyr::filter(category == "bcfp_only") |> mk_label("bcfp-only")
+  axis_link_only <- link_axis |>
+    dplyr::filter(category == "link_only") |> mk_label("link-only")
+  axis_both <- link_axis |>
+    dplyr::filter(category == "both") |> mk_label("both")
 
   other_label <- if (habitat == "rearing") "spawning" else "rearing"
-  other_combined <- dplyr::bind_rows(
-    bcfp_other |> mk_label(paste("bcfp", other_label)),
-    link_other |> mk_label(paste("link", other_label)))
+  other_bcfp <- bcfp_other |> mk_label(paste("bcfp", other_label))
+  other_link <- link_other |> mk_label(paste("link", other_label))
 
   # ---- bbox: full WSG (let user zoom) ------------------------------------
 
-  fb <- sf::st_bbox(dplyr::bind_rows(bcfp_axis, link_axis, bcfp_other, link_other))
+  fb <- sf::st_bbox(dplyr::bind_rows(
+    axis_bcfp_only, axis_link_only, axis_both, other_bcfp, other_link))
   pad <- 0.02
   dx <- fb["xmax"] - fb["xmin"]; dy <- fb["ymax"] - fb["ymin"]
   view_bbox <- sf::st_bbox(c(
@@ -223,10 +230,11 @@ lnk_map_compare <- function(wsg, species, habitat,
 
   # ---- mapgl --------------------------------------------------------------
 
-  pal_values <- c("bcfp_only", "link_only", "both")
-  pal_colors <- c("#1f78b4",  "#e31a1c",   "#6a3d9a")
-  pal_labels <- c("bcfp only (link missing)", "link only (extra)", "both")
-  other_color <- "#33a02c"
+  col_bcfp_only <- "#1f78b4"
+  col_link_only <- "#e31a1c"
+  col_both      <- "#6a3d9a"
+  col_other_bcfp <- "#33a02c"  # green (bcfp side of "other" habitat context)
+  col_other_link <- "#b15928"  # brown (link side)
 
   legend_title <- sprintf("%s %s %s — link vs bcfishpass",
     wsg, species, habitat)
@@ -237,21 +245,56 @@ lnk_map_compare <- function(wsg, species, habitat,
       fill_color = "#a6cee3", fill_opacity = 0.4, popup = "gnis_name")
 
   if (underlay) {
+    streams_under <- streams_under |>
+      dplyr::mutate(under_label = paste0(
+        "fwa | lfid ", linear_feature_id,
+        " | blkey ", blue_line_key,
+        " | DRM ", round(downstream_route_measure %||% NA),
+        " | edge ", edge_type,
+        " | order ", stream_order,
+        " | grad ", formatC(gradient, format = "f", digits = 4),
+        ifelse(is.na(gnis_name) | gnis_name == "", "", paste0(" | ", gnis_name))))
     m <- m |> mapgl::add_line_layer(
       id = "streams_all", source = streams_under,
-      line_color = "#cccccc", line_width = 0.4, line_opacity = 0.6)
+      line_color = "#cccccc", line_width = 0.4, line_opacity = 0.6,
+      popup = "under_label")
   }
 
-  m <- m |>
-    mapgl::add_line_layer(
-      id = other_label, source = other_combined,
-      line_color = other_color, line_width = 1.5, line_opacity = 0.85,
-      popup = "label") |>
-    mapgl::add_line_layer(
-      id = habitat, source = axis_combined,
-      line_color = mapgl::match_expr("category",
-        values = pal_values, stops = pal_colors, default = "#999999"),
-      line_width = 3, line_opacity = 0.9, popup = "label")
+  # Spawn / rear context layers (the OTHER habitat) — split per side so
+  # the user can isolate "where does bcfp see spawn?" independently of
+  # link's spawn picture.
+  if (nrow(other_bcfp) > 0) {
+    m <- m |> mapgl::add_line_layer(
+      id = sprintf("%s_bcfp", other_label), source = other_bcfp,
+      line_color = col_other_bcfp, line_width = 1.5, line_opacity = 0.85,
+      popup = "label")
+  }
+  if (nrow(other_link) > 0) {
+    m <- m |> mapgl::add_line_layer(
+      id = sprintf("%s_link", other_label), source = other_link,
+      line_color = col_other_link, line_width = 1.5, line_opacity = 0.85,
+      popup = "label")
+  }
+
+  # Axis (target habitat) layers — three independent toggles.
+  if (nrow(axis_both) > 0) {
+    m <- m |> mapgl::add_line_layer(
+      id = sprintf("%s_both", habitat), source = axis_both,
+      line_color = col_both, line_width = 3, line_opacity = 0.9,
+      popup = "label")
+  }
+  if (nrow(axis_bcfp_only) > 0) {
+    m <- m |> mapgl::add_line_layer(
+      id = sprintf("%s_bcfp_only", habitat), source = axis_bcfp_only,
+      line_color = col_bcfp_only, line_width = 3, line_opacity = 0.9,
+      popup = "label")
+  }
+  if (nrow(axis_link_only) > 0) {
+    m <- m |> mapgl::add_line_layer(
+      id = sprintf("%s_link_only", habitat), source = axis_link_only,
+      line_color = col_link_only, line_width = 3, line_opacity = 0.9,
+      popup = "label")
+  }
 
   for (lyr in extra_layers) {
     m <- m |> mapgl::add_line_layer(
@@ -265,8 +308,14 @@ lnk_map_compare <- function(wsg, species, habitat,
   m <- m |>
     mapgl::add_legend(
       legend_title,
-      values = c(pal_labels, paste(other_label, "(either)")),
-      colors = c(pal_colors, other_color),
+      values = c(
+        "bcfp only (link missing)",
+        "link only (extra)",
+        "both",
+        sprintf("%s (bcfp)", other_label),
+        sprintf("%s (link)", other_label)),
+      colors = c(col_bcfp_only, col_link_only, col_both,
+                 col_other_bcfp, col_other_link),
       type = "categorical", position = "top-right") |>
     mapgl::add_layers_control(collapsible = TRUE, position = "top-left")
 
diff --git a/inst/extdata/configs/bcfishpass/dimensions.csv b/inst/extdata/configs/bcfishpass/dimensions.csv
index 0d68c6d..f137292 100644
--- a/inst/extdata/configs/bcfishpass/dimensions.csv
+++ b/inst/extdata/configs/bcfishpass/dimensions.csv
@@ -1,9 +1,9 @@
-"species","spawn_lake","spawn_stream","spawn_stream_in_waterbody","rear_lake","rear_lake_only","rear_lake_area_only","rear_no_fw","rear_stream","rear_stream_in_waterbody","rear_wetland","rear_wetland_polygon","rear_wetland_area_only","rear_all_edges","river_skip_cw_min","rear_stream_order_bypass","spawn_requires_connected","spawn_connected_distance_max","spawn_connected_direction","spawn_connected_gradient_max","spawn_connected_cw_min","spawn_connected_edge_types","spawn_connected_lake_adjacent","rear_requires_connected","rear_connected_distance_max","notes"
-"BT","no","yes","no","no","no","no","no","yes","no","no","no","no","yes","yes","no","",,"",,,,"",,,"bcfishpass: rear_all_edges=yes means no edge_type filter for rearing."
-"CH","no","yes","no","no","no","no","no","yes","no","no","no","no","no","yes","no","",,"",,,,"",,,"bcfishpass: stream order bypass handled in compare script Step 7b."
-"CM","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no","",,"",,,,"",,,"bcfishpass: no freshwater rearing."
-"CO","no","yes","no","no","no","no","no","yes","yes","yes","no","no","no","yes","no","",,"",,,,"",,,"bcfishpass: wetland-flow carve-out (1050/1150 no thresholds); polygon=no excludes the W-polygon rule."
-"PK","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no","",,"",,,,"",,,"bcfishpass: no freshwater rearing."
-"SK","no","yes","no","yes","yes","no","no","no","no","no","no","no","no","yes","no","rearing",3000,"downstream",0.05,0,,"yes",,,"bcfishpass: lake-only rearing. Spawning within 3km ds of rearing lake."
-"ST","no","yes","no","no","no","no","no","yes","yes","no","no","no","no","yes","no","",,"",,,,"",,,"bcfishpass: stream order bypass handled in compare script Step 7b."
-"WCT","no","yes","no","no","no","no","no","yes","no","no","no","no","no","yes","no","",,"",,,,"",,,"bcfishpass: stream order bypass handled in compare script Step 7b."
+"species","spawn_lake","spawn_stream","spawn_stream_in_waterbody","rear_lake","rear_lake_only","rear_lake_area_only","rear_no_fw","rear_stream","rear_stream_in_waterbody","rear_wetland","rear_wetland_polygon","rear_wetland_area_only","rear_all_edges","river_skip_cw_min","rear_stream_order_bypass","rear_stream_order_parent_min","rear_stream_order_child_min","rear_stream_order_child_max","rear_stream_order_distance_max","spawn_requires_connected","spawn_connected_distance_max","spawn_connected_direction","spawn_connected_gradient_max","spawn_connected_cw_min","spawn_connected_edge_types","spawn_connected_lake_adjacent","rear_requires_connected","rear_connected_distance_max","notes"
+"BT","no","yes","no","no","no","no","no","yes","no","no","no","no","yes","yes","no","","","","","",,"",,,,"",,,"bcfishpass: rear_all_edges=yes means no edge_type filter for rearing."
+"CH","no","yes","no","no","no","no","no","yes","no","no","no","no","no","yes","no","","","","","",,"",,,,"",,,"bcfishpass: stream order bypass handled in compare script Step 7b."
+"CM","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no",,,,,"",,"",,,,"",,,"bcfishpass: no freshwater rearing."
+"CO","no","yes","no","no","no","no","no","yes","yes","yes","no","no","no","yes","no","","","","","",,"",,,,"",,,"bcfishpass: wetland-flow carve-out (1050/1150 no thresholds); polygon=no excludes the W-polygon rule."
+"PK","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no",,,,,"",,"",,,,"",,,"bcfishpass: no freshwater rearing."
+"SK","no","yes","no","yes","yes","no","no","no","no","no","no","no","no","yes","no",,,,,"rearing",3000,"downstream",0.05,0,,"yes",,,"bcfishpass: lake-only rearing. Spawning within 3km ds of rearing lake."
+"ST","no","yes","no","no","no","no","no","yes","yes","no","no","no","no","yes","no","","","","","",,"",,,,"",,,"bcfishpass: stream order bypass handled in compare script Step 7b."
+"WCT","no","yes","no","no","no","no","no","yes","no","no","no","no","no","yes","no","","","","","",,"",,,,"",,,"bcfishpass: stream order bypass handled in compare script Step 7b."
diff --git a/inst/extdata/configs/default/dimensions.csv b/inst/extdata/configs/default/dimensions.csv
index 531de9f..058550b 100644
--- a/inst/extdata/configs/default/dimensions.csv
+++ b/inst/extdata/configs/default/dimensions.csv
@@ -1,14 +1,14 @@
-"species","spawn_lake","spawn_stream","spawn_stream_in_waterbody","rear_lake","rear_lake_only","rear_lake_area_only","rear_no_fw","rear_stream","rear_stream_in_waterbody","rear_wetland","rear_wetland_polygon","rear_wetland_area_only","rear_all_edges","river_skip_cw_min","rear_stream_order_bypass","spawn_requires_connected","spawn_connected_distance_max","spawn_connected_direction","spawn_connected_gradient_max","spawn_connected_cw_min","spawn_connected_edge_types","spawn_connected_lake_adjacent","rear_requires_connected","rear_connected_distance_max","rear_lake_ha_min","rear_wetland_ha_min","notes"
-"BT","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,10,1,"Cold headwaters with gravel substrate. Major lake-rearing populations in Babine Quesnel Kootenay Stuart. Tributary mouth spawning in mainstem rivers. Wetland rearing in beaver complexes and side channels. rear_lake_ha_min=10 — tolerates small lakes. rear_wetland_ha_min=1 — small wetlands count."
-"CH","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,100,1,"Stream-type modelled. Ocean-type chinook outmigrate as fry within weeks; not modelled separately yet — investigate split. Some lake rearing (Cultus Pitt Stave). Spawning in mainstem rivers — large bodies need >=4m channels. rear_lake_ha_min=100 — Cultus/Pitt/Stave class systems. rear_wetland_ha_min=1."
-"CM","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no","",,"",,,,"",,,,,"Spawn lower mainstem and side channels. Fry to estuary within days/weeks of emergence — may use side channels briefly before outmigration but classified as no freshwater rearing. Acceptable simplification — investigate if a fry-days dimension is needed."
-"CO","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,2,0.5,"Heavy use of off-channel wetlands and beaver complexes for overwintering. Smaller-tributary spawners primarily but river spawning is yes — investigate further. rear_lake_ha_min=2 — uses small lakes + ponds extensively. rear_wetland_ha_min=0.5 — beaver complexes are small."
-"PK","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no","",,"",,,,"",,,,,"Spawn lower mainstem. Fry to estuary within days of emergence — minimal stream rearing. Even shorter freshwater residency than CM. Classified as no freshwater rearing."
-"SK","no","yes","no","yes","yes","no","no","no","no","no","no","no","no","yes","no","rearing",3000,"downstream",0.05,0,,"no",,,200,,"Stream-spawning sockeye. Lake-obligate rearing — rear_lake_only is the override. Spawning within 3km downstream of connected rearing lake (outlet stream). Beach-spawning shoreline populations exist (e.g. Babine, Shuswap) but are a separate phenotype not modelled here — matches bcfishpass convention."
-"ST","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,60,1,"Spawn medium-large rivers — large bodies need >=4m channels. Lake-rearing populations exist. Wetland rearing yes. rear_lake_ha_min=60 — ocean-typed; smaller lakes less likely. rear_wetland_ha_min=1."
-"WCT","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,10,1,"Lake-rearing common in Kootenay Slocan etc. Wetlands used for overwintering. rear_lake_ha_min=10 — tolerates small resident-population lakes. rear_wetland_ha_min=1."
-"KO","no","yes","no","yes","yes","no","no","no","no","no","no","no","no","yes","no","rearing",3000,"downstream",0.05,0,,"",,,200,,"Stream-spawning kokanee. Lake-obligate rearing — rear_lake_only is the override. Spawning within 3km downstream of connected rearing lake (outlet stream). Some populations spawn on lake shoreline beaches — separate phenotype not modelled here."
-"CT","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,10,1,"Resident form modelled. Anadromous form exists and uses estuaries — add separately later. rear_lake_ha_min=10 — resident coastal cutthroat in small lakes. rear_wetland_ha_min=1."
-"DV","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,10,1,"Some lake spawning exists but rare; classified as no. Lake-rearing common in coastal populations. Wetland rearing in beaver complexes and side channels. rear_lake_ha_min=10 — coastal populations in small lakes. rear_wetland_ha_min=1."
-"RB","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","",,"",,,,"",,,10,1,"Some lake spawning exists but rare; classified as no. Wetland and lake rearing common (Kamloops trout et al). rear_lake_ha_min=10 — residents use a wide size range of lakes. rear_wetland_ha_min=1."
-"GR","no","yes","no","yes","no","no","no","yes","yes","no","no","no","no","yes","no","",,"",,,,"",,,40,,"Lake rearing yes — northern populations use lakes. rear_lake_ha_min=40 — northern systems tend to be larger."
+"species","spawn_lake","spawn_stream","spawn_stream_in_waterbody","rear_lake","rear_lake_only","rear_lake_area_only","rear_no_fw","rear_stream","rear_stream_in_waterbody","rear_wetland","rear_wetland_polygon","rear_wetland_area_only","rear_all_edges","river_skip_cw_min","rear_stream_order_bypass","rear_stream_order_parent_min","rear_stream_order_child_min","rear_stream_order_child_max","rear_stream_order_distance_max","spawn_requires_connected","spawn_connected_distance_max","spawn_connected_direction","spawn_connected_gradient_max","spawn_connected_cw_min","spawn_connected_edge_types","spawn_connected_lake_adjacent","rear_requires_connected","rear_connected_distance_max","rear_lake_ha_min","rear_wetland_ha_min","notes"
+"BT","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","","","","","",,"",,,,"",,,10,1,"Cold headwaters with gravel substrate. Major lake-rearing populations in Babine Quesnel Kootenay Stuart. Tributary mouth spawning in mainstem rivers. Wetland rearing in beaver complexes and side channels. rear_lake_ha_min=10 — tolerates small lakes. rear_wetland_ha_min=1 — small wetlands count."
+"CH","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","","","","","",,"",,,,"",,,100,1,"Stream-type modelled. Ocean-type chinook outmigrate as fry within weeks; not modelled separately yet — investigate split. Some lake rearing (Cultus Pitt Stave). Spawning in mainstem rivers — large bodies need >=4m channels. rear_lake_ha_min=100 — Cultus/Pitt/Stave class systems. rear_wetland_ha_min=1."
+"CM","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no",,,,,"",,"",,,,"",,,,,"Spawn lower mainstem and side channels. Fry to estuary within days/weeks of emergence — may use side channels briefly before outmigration but classified as no freshwater rearing. Acceptable simplification — investigate if a fry-days dimension is needed."
+"CO","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","","","","","",,"",,,,"",,,2,0.5,"Heavy use of off-channel wetlands and beaver complexes for overwintering. Smaller-tributary spawners primarily but river spawning is yes — investigate further. rear_lake_ha_min=2 — uses small lakes + ponds extensively. rear_wetland_ha_min=0.5 — beaver complexes are small."
+"PK","no","yes","no","no","no","no","yes","no","no","no","no","no","no","yes","no",,,,,"",,"",,,,"",,,,,"Spawn lower mainstem. Fry to estuary within days of emergence — minimal stream rearing. Even shorter freshwater residency than CM. Classified as no freshwater rearing."
+"SK","no","yes","no","yes","yes","no","no","no","no","no","no","no","no","yes","no",,,,,"rearing",3000,"downstream",0.05,0,,"no",,,200,,"Stream-spawning sockeye. Lake-obligate rearing — rear_lake_only is the override. Spawning within 3km downstream of connected rearing lake (outlet stream). Beach-spawning shoreline populations exist (e.g. Babine, Shuswap) but are a separate phenotype not modelled here — matches bcfishpass convention."
+"ST","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","","","","","",,"",,,,"",,,60,1,"Spawn medium-large rivers — large bodies need >=4m channels. Lake-rearing populations exist. Wetland rearing yes. rear_lake_ha_min=60 — ocean-typed; smaller lakes less likely. rear_wetland_ha_min=1."
+"WCT","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no","","","","","",,"",,,,"",,,10,1,"Lake-rearing common in Kootenay Slocan etc. Wetlands used for overwintering. rear_lake_ha_min=10 — tolerates small resident-population lakes. rear_wetland_ha_min=1."
+"KO","no","yes","no","yes","yes","no","no","no","no","no","no","no","no","yes","no",,,,,"rearing",3000,"downstream",0.05,0,,"",,,200,,"Stream-spawning kokanee. Lake-obligate rearing — rear_lake_only is the override. Spawning within 3km downstream of connected rearing lake (outlet stream). Some populations spawn on lake shoreline beaches — separate phenotype not modelled here."
+"CT","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no",,,,,"",,"",,,,"",,,10,1,"Resident form modelled. Anadromous form exists and uses estuaries — add separately later. rear_lake_ha_min=10 — resident coastal cutthroat in small lakes. rear_wetland_ha_min=1."
+"DV","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no",,,,,"",,"",,,,"",,,10,1,"Some lake spawning exists but rare; classified as no. Lake-rearing common in coastal populations. Wetland rearing in beaver complexes and side channels. rear_lake_ha_min=10 — coastal populations in small lakes. rear_wetland_ha_min=1."
+"RB","no","yes","no","yes","no","no","no","yes","yes","yes","yes","no","no","yes","no",,,,,"",,"",,,,"",,,10,1,"Some lake spawning exists but rare; classified as no. Wetland and lake rearing common (Kamloops trout et al). rear_lake_ha_min=10 — residents use a wide size range of lakes. rear_wetland_ha_min=1."
+"GR","no","yes","no","yes","no","no","no","yes","yes","no","no","no","no","yes","no",,,,,"",,"",,,,"",,,40,,"Lake rearing yes — northern populations use lakes. rear_lake_ha_min=40 — northern systems tend to be larger."
diff --git a/inst/extdata/configs/dimensions_columns.csv b/inst/extdata/configs/dimensions_columns.csv
index 5f218f5..9d57b2e 100644
--- a/inst/extdata/configs/dimensions_columns.csv
+++ b/inst/extdata/configs/dimensions_columns.csv
@@ -14,7 +14,11 @@ rear_wetland_polygon,yes_no,rule_emission,rear,no,"Should lnk_rules_build emit t
 rear_wetland_area_only,yes_no,rule_emission,rear,no,"When yes, emits `area_only: true` on the W polygon rule. Same shape as rear_lake_area_only.",area_only: true on W rule,fresh#182
 rear_all_edges,yes_no,rule_emission,rear,no,"Override: drop the edge_types filter on the rear-stream rule (matches every accessible segment regardless of edge_type). Used by BT in bcfishpass bundle to match bcfishpass's BT rear access logic. Higher precedence than rear_stream's normal emission.","empty rule (no edge_types) (rear)",-
 river_skip_cw_min,yes_no,rule_emission,both,no,"Should the river polygon rule (`waterbody_type: R`) skip the cw_min check? When yes, emits `channel_width: [0, 9999]` on the river rule - biology of large rivers spawning/rearing fish regardless of FWA cw measurement. Spawn-side bypass; complements the rearing stream_order bypass (rear_stream_order_bypass).","channel_width: [0, 9999] on R rule",-
-rear_stream_order_bypass,yes_no,rule_emission,rear,no,"Currently inert: column reads, no SQL emission yet. Once fresh#158 ships frs_order_child, lnk will emit a per-species frs_order_child invocation when this is yes. Captures the bcfishpass rearing bypass `(stream_order_parent >= 5 AND stream_order = 1)`.","channel_width_min_bypass placeholder (not yet emitted)",fresh#158
+rear_stream_order_bypass,yes_no,rule_emission,rear,no,"Gate for the rearing stream-order bypass. When yes, lnk_rules_build emits a `channel_width_min_bypass:` block on the rear stream-edge rule and lnk_pipeline_classify calls fresh::frs_order_child post-classify per species. Captures the biology of small streams plugging into large rivers: fish use these reaches for rearing despite low/missing FWA channel-width estimates. The bypass parameters (parent, child range, distance) are configured via the four `rear_stream_order_*` columns below. When no, the YAML block is dropped and the post-classify call is skipped entirely.","channel_width_min_bypass block (rear) + frs_order_child call",fresh#158
+rear_stream_order_parent_min,positive_integer,bypass,rear,5,"Minimum stream order at the trib BLK's mouth confluence (i.e. the order of the receiving stream). Maps to fresh::frs_order_child(parent_order_min). Default 5 matches bcfishpass's hardcoded predicate. Only honoured when rear_stream_order_bypass=yes.",channel_width_min_bypass.stream_order_parent_min,fresh#158
+rear_stream_order_child_min,positive_integer,bypass,rear,1,"Lower bound on the bypass-eligible segment's own stream_order. Maps to fresh::frs_order_child(child_order_min). Default 1. Combined with rear_stream_order_child_max defines a range of segment orders to credit. Only honoured when rear_stream_order_bypass=yes.",channel_width_min_bypass.stream_order_min,fresh#158
+rear_stream_order_child_max,positive_integer,bypass,rear,1,"Upper bound on the bypass-eligible segment's own stream_order. Maps to fresh::frs_order_child(child_order_max). Default 1 (single-order tribs only — bcfp parity). Set higher to widen — e.g. 5 captures order-1..5 single-order tribs feeding order-5+ stems. Only honoured when rear_stream_order_bypass=yes.",channel_width_min_bypass.stream_order_max,fresh#158
+rear_stream_order_distance_max,positive_numeric,bypass,rear,empty,"Cap (metres) on the bypass-eligible segment's downstream_route_measure — credits only the lower N m of each direct-trib BLK from its mouth confluence. Maps to fresh::frs_order_child(distance_max). Empty = no cap (whole BLK up to gradient/access limits). Only honoured when rear_stream_order_bypass=yes.",channel_width_min_bypass.distance_max,fresh#158
 spawn_requires_connected,character,connected_rules,spawn,empty,"Habitat type that spawn rules' segments must be connected to (e.g. ""rearing"" for SK / KO - spawning happens within reach of rearing lakes). When set, every emitted spawn rule carries `requires_connected: <value>`.",requires_connected on every spawn rule,-
 spawn_connected_distance_max,numeric,connected_rules,spawn,empty,"Max network distance (metres) between a spawn segment and the connected target. When set with spawn_requires_connected, every spawn rule carries `connected_distance_max: <value>`.",connected_distance_max on every spawn rule,-
 spawn_connected_direction,enum:upstream/downstream/both,connected_rules,spawn,empty,"Direction to walk from spawn segment when looking for the connected target. Drives the spawn_connected: block in rules.yaml.",spawn_connected.direction,-
diff --git a/planning/active/findings.md b/planning/active/findings.md
index 2be1894..0efc490 100644
--- a/planning/active/findings.md
+++ b/planning/active/findings.md
@@ -1,75 +1,46 @@
-# Findings — link#88
+# Findings — link `frs_order_child` wire-up
 
-## Diagnosis (2026-04-30)
+## Summary
 
-### Single-stream trace: HARR blkey 356286055
+`frs_order_child` is a link methodology primitive, not a bcfp parity replicator. The HORS BT pre-flight result (link 394 km / bcfp 396 km on `rearing_stream`, −0.5%) at `bypass=yes, parent=5, child=1, dmax=300` is numerical proximity by accident — bcfp's bypass operates inside 3 connectivity-aware rearing phases (Phase 1: rearing-on-spawn, Phase 2: cluster-DS-of-spawn, Phase 3: cluster-US-of-spawn-with-no->5%-grade-between), and does NOT use a `stream_order_max` filter. Our function runs post-cluster, post-classify, with `stream_order_max` for direct-child semantics and `distance_max` as a biology-tuning cap. Different design, different semantics.
 
-- link `fresh.streams_habitat`: 21 segments, all `rearing=FALSE` for BT, all `accessible=FALSE` for BT
-- bcfp `streams_access`: `access_bt = 1` on every segment, `barriers_bt_dnstr = {}` (zero natural BT barriers)
-- bcfp `barriers_anthropogenic_dnstr`: 2 entries (DAM at 356282804 DRM 739, ROAD/DEMOGRAPHIC at 356282804 DRM 658) — both flagged in `barriers_remediations` (REMEDIATED)
-- Two **subsurfaceflow** points downstream on 356282804 (DRMs 265, 279) in `barriers_subsurfaceflow`
-- 55 anadromous obs (CH/CM/CO) upstream of (356282804, 265) — clears bcfp's threshold (1 for BT, 5 for anadromous)
-- bcfp lifts both subsurfaceflow points → `barriers_bt` and `barriers_ch_cm_co_pk_sk` empty in this drainage → BT/CH/CO credit upstream
+## Biology hook (the why)
 
-### Why link doesn't lift
+The FWA-derived `channel_width` estimate is unreliable on small (1st-order) streams because the MAP (mean annual precipitation) signal isn't carried cleanly on those reaches. When such a small stream plugs directly into a large river, fish *do* use the lower reach for rearing despite the low/missing CW estimate — flow, temperature, cool-water mixing at confluence, backwater habitat all support juvenile use. `frs_order_child` is the parametric primitive expressing this biology.
 
-`lnk_pipeline_prepare()` build order (current):
+## Design decisions (verbatim from fresh#158, anchored here for next session)
 
-1. `prep_load_aux` → falls, definite, control, habitat
-2. `prep_gradient` → gradient_barriers_raw (pruned, ltree-enriched)
-3. **`prep_natural` → `<schema>.natural_barriers` = gradient + falls** ← subsurfaceflow NOT here
-4. `prep_overrides` → calls `lnk_barrier_overrides(barriers = natural_barriers)` → per-species skip list
-5. **`prep_subsurfaceflow`** (opt-in) → `<schema>.barriers_subsurfaceflow` ← runs AFTER overrides
+- **`stream_order = stream_order_max` (per BLK)** — bypass only applies on the mouth-side reach of a BLK, where the BLK's order is at its maximum. Excludes order-1 headwater portions of multi-order BLKs (e.g., a named creek that grows to order 3 at its mouth). Documented as direct-child semantics. **bcfp does NOT use this filter** — our function is structurally tighter on multi-order BLKs by design.
+- **`distance_max`** — caps the bypass to the lower N metres of each direct-trib BLK. Whole-segment overshoot is the documented default (Option A in fresh#158). Lets users tune "how far into the trib does the parent-river effect extend" — a biology question, not a parity question.
+- **Post-cluster, post-classify placement** — intentional. fresh#158: *"can add segments that frs_cluster removed because they had no connected spawning. For some species this is correct (rear-only species), for others it might over-classify. Mitigation: the accessible guard limits over-reach to accessible network only; the parametric parent_order_min / child_order_min / distance_max lets users tune."* fresh#156 was closed in favor of this approach because the alternative (apply during classify) inflates rearing counts pre-cluster.
 
-`lnk_pipeline_classify_build_breaks()` then UNIONs `barriers_subsurfaceflow` directly into `fresh.streams_breaks` with label `blocked`. Since the override skip list never saw it, `frs_habitat_classify(barrier_overrides = ...)` cannot lift it. All species get blocked at the subsurfaceflow position.
+## HORS BT pre-flight progression (this session, 2026-05-01)
 
-### bcfishpass natural barrier construction (per-species)
+| Iteration | Predicate | link | bcfp | diff_pct |
+|---|---|---|---|---|
+| Pre-#158 (no bypass) | — | 366 | 396 | −7.68% |
+| 0.27.2 (broken predicate, removed `stream_order_max`) | parent≥5, order=1, no `so_max` filter | 491 | 396 | +23.9% |
+| 0.27.3 (`stream_order_max` via CTE) | + `s.stream_order = s.stream_order_max` | 451 | 396 | +13.9% |
+| 0.27.4 + dmax=300 | + `downstream_route_measure ≤ 300` | 394 | 396 | −0.5% |
 
-`model/01_access/sql/model_access_bt.sql` — `barriers_bt`:
+Map snapshots in `data-raw/maps/HORS_BT_rearing_AFTER_*.html`.
 
-- gradient_25 + gradient_30 + falls + **subsurfaceflow** + user_definite
-- LIFT: any obs upstream (BT/CH/CM/CO/PK/SK/ST), or any habitat upstream
-- user_definite always retained
+The `dmax=300` calibration is exploratory — the 4-iteration trail above shows that the apparent parity at 0.5% under is **not the result of methodology alignment** but a numerical balance: under-credit on long-trib reaches (we cap at 300m, bcfp credits the whole trib up to gradient/access limits) compensates for over-credit on cluster-disconnected segments (we run post-cluster without the bcfp Phase-3 `>5%-grade-between` gradient gate).
 
-`model/01_access/sql/model_access_ch_cm_co_pk_sk.sql` — `barriers_ch_cm_co_pk_sk`:
+## Segment-level evidence
 
-- gradient_15/20/25/30 + falls + **subsurfaceflow** + user_definite
-- LIFT: ≥5 anadromous obs upstream (post-1990), or any habitat upstream
-- `user_barriers_definite_control.barrier_ind = TRUE` blocks the obs lift; habitat lift unaffected
-- user_definite always retained
+- **BLK 356322947, DRM 3000–4500 (HORS BT):** bcfp credits 5 segments (3223, 3341, 3440, 3542, 4054) as rearing — passes gradient ≤ 0.1049, no DS barriers, `stream_order_max=2` per bcfp's stored column. Our `frs_order_child` predicate excludes them all because `stream_order=1, stream_order_max=2` fails our `s.stream_order = s.stream_order_max` filter. Demonstrates structural divergence from bcfp on multi-order BLKs.
+- **BLK 356353593 (Divan Creek), DRM 6009+:** bcfp credits DRM 6009 + 6095 (gradients 0.009 and 0.078, both under 0.1049). Drops DRM 7280 (gradient 0.1008 — under cap, channel_width 1.77 ≥ cw_min 1.5) because Phase 3 cluster trace is blocked by the >5%-grade segment at DRM 7112 between 7280 and downstream spawn. Demonstrates that bcfp's bypass is gated by cluster connectivity to spawn, which we don't replicate.
 
-bcfp's `barriers_anthropogenic` (PSCIS, dams, road crossings) is **NOT** in the per-species barrier set. It's tracked separately for downstream-of-crossing accountability. Anthropogenic barriers don't gate species access in bcfp's habitat sums.
+## What we are *not* doing
 
-### Design diagnosis
+- **Not adding gradient ceiling to `frs_order_child`.** bcfp's bypass-eligible segments must pass `gradient ≤ rear_gradient_max`, but link's `frs_order_child` is meant to be additive on `accessible = TRUE` segments regardless of gradient. The `accessible` guard already filters access-blocked segments via the link pipeline's barrier-aware accessibility; gradient-driven exclusion would be a parity-driven addition that contradicts the function's "additive primitive" design.
+- **Not adding cluster-aware filtering.** The post-cluster placement is intentional; mitigation is `parent_order_min` / `child_order_min/max` / `distance_max`. If callers want cluster-aware behaviour they can run `frs_order_child` *before* `frs_cluster` instead of after.
+- **Not pursuing bcfp parity for this primitive.** fresh#158 was explicit: link's default-bundle should ship parametric defaults tuned from BABL inspection, not a bcfp clone.
 
-`.lnk_pipeline_prep_subsurfaceflow` was added in PR #82 as its own helper. Treated subsurfaceflow as a parallel concept needing a parallel pipeline phase. But subsurfaceflow is just **a third row in the same union** that `prep_natural` already builds for gradient + falls. The wiring miss: subsurfaceflow's positions never reached `natural_barriers`, so the per-species lift skipped it entirely.
+## Links
 
-`prep_natural` is the right home — bcfp's source of truth confirms gradient + falls + subsurfaceflow is *the* natural-barrier union per species.
-
-### Where it surfaces (15-WSG rollup, parity)
-
-| WSG  | sp | metric          | link | bcfp | diff_pct |
-|------|----|-----------------|------|------|---------:|
-| HARR | CH | rearing_stream  | 118  | 139  | -14.8 |
-| HARR | CO | rearing_stream  | 134  | 155  | -13.3 |
-| HARR | ST | rearing_stream  | 157  | 177  | -11.6 |
-| HARR | BT | rearing_stream  | 292  | 326  | -10.4 |
-| HORS | BT | rearing_stream  | 366  | 396  |  -7.7 |
-| LFRA | BT | rearing_stream  | 1020 | 1103 |  -7.5 |
-| LFRA | BT | rearing         | 1670 | 1800 |  -7.2 |
-| HORS | CH | rearing_stream  | 167  | 179  |  -6.8 |
-
-LILL/VICT unaffected — sparse subsurfaceflow + sparse fish observations above what does exist.
-
-## What is NOT involved
-
-- #83 (anthropogenic dam design): the dam at DRM 739 and road at DRM 658 carry `barrier`/`potential` labels in `fresh.streams_breaks`. Default `label_block = "blocked"` means they don't gate. Not the cause.
-- `barriers_definite`: intentionally separate per bcfp — never lifted via obs/hab. Link mirrors that.
-- `barriers_remediations`: bcfp tracks for downstream reporting only; doesn't gate access.
-
-## Versions at diagnosis
-
-- link 0.19.0 (commit e4e7a6e)
-- fresh 0.25.0
-- bcfishpass 440bc1e (2026-04-28)
-- fwapg local Docker (port 5432)
+- [fresh#158](https://github.com/NewGraphEnvironment/fresh/issues/158) — design doc (canonical)
+- [fresh#156](https://github.com/NewGraphEnvironment/fresh/issues/156) — predecessor, closed in favor of #158
+- [link#23](https://github.com/NewGraphEnvironment/link/issues/23) — CH spawning misread, closed not-a-bug
+- [research/bcfishpass_comparison.md](../../research/bcfishpass_comparison.md) — pre-#158 68 km gap evidence
diff --git a/planning/active/progress.md b/planning/active/progress.md
index 4413dd3..4eeeede 100644
--- a/planning/active/progress.md
+++ b/planning/active/progress.md
@@ -1,17 +1,70 @@
-# Progress — link#88
-
-## Session 2026-04-30
-
-- Diagnosis: traced HARR blkey 356286055 BT under-credit to subsurfaceflow positions on downstream tributary 356282804 not reaching `natural_barriers` for the per-species observation/habitat lift.
-- Read bcfp SQL (`model_access_bt.sql`, `model_access_ch_cm_co_pk_sk.sql`) — confirmed bcfp's natural-barrier union includes subsurfaceflow with same lift rules.
-- Confirmed default-bundle off-switch is preserved verbatim (omit `subsurfaceflow` from `cfg$pipeline$break_order`).
-- Filed link#88 with diagnosis + proposed fix.
-- Branch `88-fold-subsurfaceflow-natural` from main.
-- PWF baseline (commit 4bd9ca0).
-- Code change: extended `.lnk_pipeline_prep_natural` signature `(conn, aoi, cfg, loaded, schema)`; absorbed subsurfaceflow body, gated on `cfg$pipeline$break_order`; deleted standalone `.lnk_pipeline_prep_subsurfaceflow`; pruned conditional call from `lnk_pipeline_prepare()`. `devtools::document()` clean.
-- Tests: 3 new test cases in `tests/testthat/test-lnk_pipeline_prepare.R` — opted-out, opted-in (per-statement assertion that link#88 fix INSERT fires), control-table honoured. 44/44 pass.
-- Code-check: 3 rounds. Rounds 1–2 clean. Round 3 caught a fragile cross-statement regex in the test; replaced with per-statement `any(grepl & grepl)`. Sanity-verified the assertion catches the regression.
-- Pre-flight: HARR single-WSG `compare_bcfishpass_wsg(wsg = "HARR", config = lnk_config("bcfishpass"))` 89.5 s. blkey 356286055 BT credits 6.509 km (was 0). HARR BT diffs collapsed: rearing_stream -10.4% → -4.19%, rearing -1.84%, spawning -1.6%.
-- 15-WSG `tar_make` (53m 2s, 33/33). Parity dramatic on HARR (CH/CO/ST <0.32%; BT residual -4.19%) and LFRA (CH/CO/ST <0.6%; BT residual -3.75%). HORS unchanged (-7.68% rearing_stream BT) — different mechanism, follow-up needed. Default-bundle bit-identical (0 of 581 rows changed).
-- Reproducibility re-run (52m 55s, 33/33). 0 of 1057 link_value rows differ; digest match (`5a641892b82604259b0ba168ea093661`). ✓
-- Code commit `a21a8f8`. PWF + verification logs commit. Ready for PR.
+# Progress — link `frs_order_child` wire-up
+
+## Session 2026-05-01
+
+### What landed (fresh)
+
+- **fresh 0.27.1** — validator allows `channel_width_min_bypass` predicate (PR #194 merged)
+- **fresh 0.27.2** — false-start patch (removed `stream_order_max` reference based on misread); superseded by 0.27.3
+- **fresh 0.27.3** — `frs_order_child` derives `stream_order_max` per BLK via CTE (PR #196 merged)
+- **fresh 0.27.4** — validator allows `distance_max` key inside `channel_width_min_bypass` block (PR #197 merged)
+
+### What's staged on link `96-frs-order-child-wire` branch (uncommitted)
+
+- 3 new columns in `dimensions.csv` (both bundles): `rear_stream_order_bypass`, `rear_stream_order_parent_min`, `rear_stream_order_distance_max`
+- `lnk_rules_build` emits all three into `channel_width_min_bypass:` block in rules.yaml
+- `lnk_pipeline_classify` reads the block, calls `fresh::frs_order_child` per species
+- Bundle defaults: `bypass=yes, parent=5, dmax=300` for BT/CH/CO/ST/WCT in both bundles
+- New tests in `test-lnk_rules_build.R` for the new columns
+
+### Calibration runs (HORS BT, m4 + local fwapg)
+
+- `data-raw/logs/20260501_15_preflight_hors_post_272.txt` — 0.27.2 broken predicate, +23.9%
+- `data-raw/logs/20260501_17_preflight_hors_post_273.txt` — 0.27.3 with `stream_order_max`, +13.9%
+- `data-raw/logs/20260501_18_preflight_hors_explicit_child.txt` — child=1 explicit (default-equivalent), +13.9%
+- `data-raw/logs/20260501_19_preflight_hors_dmax_300.txt` — dmax=300 added, **−0.5%**
+- `data-raw/logs/20260501_20_preflight_hors_child35_dmax300.txt` — child=3..5 exploratory, returns to −7.7% baseline
+
+### Maps (HTML snapshots in `data-raw/maps/`)
+
+- `HORS_BT_rearing_BEFORE_158.html` — pre-fix baseline
+- `HORS_BT_rearing_AFTER_158.html` — broken predicate state
+- `HORS_BT_rearing_AFTER_273.html` — `stream_order_max` only
+- `HORS_BT_rearing_AFTER_274_dmax300.html` — calibrated (dmax=300)
+- `HORS_BT_rearing_AFTER_274_child35_dmax300.html` — child=3..5 exploration
+
+`_lnk_map_compare.R` was updated to split layers into `rearing_link_only` / `rearing_bcfp_only` / `rearing_both` toggle-able layers (plus `spawning_link` / `spawning_bcfp`).
+
+### What we relitigated this session that we shouldn't have
+
+The fresh#158 issue body already documented:
+- `stream_order_max` is for direct-child semantics
+- `distance_max` is a parametric flex with whole-segment overshoot trade-off
+- Post-cluster placement is intentional
+- bcfp parity is NOT the goal — link's default bundle should tune from BABL inspection
+
+I rediscovered each of these as if they were new findings. fresh#156 (closed in favor of #158) explicitly rejected the "rule-grammar predicate at classify time" approach with the same analysis we re-did. **Lesson saved to memory** — check originating issue bodies first.
+
+### Parked
+
+Methodology decision pending. Don't merge link `96-frs-order-child-wire`. Don't run 15-WSG. Re-pick when there's bandwidth to either:
+- Inspect the calibration on a wider WSG sample (BABL or LFRA next, per fresh#158)
+- Decide to ship `bypass=no` defaults and keep the infrastructure parametric
+
+### Adjacent finding surfaced this session — [link#96](https://github.com/NewGraphEnvironment/link/issues/96)
+
+While inspecting the HORS BT map, identified a separate bug: `falls` is documented in `lnk_pipeline_break.R:10-13` as part of bcfp's break order but is **not in the implementation's `source_tables` list or `break_order` default**. Result: the FWA stream network is never broken at fall positions. Where two falls sit close together (e.g. HORS BLK 356357296 at DRMs 67524 + 67565, 41m apart), only the one coinciding with another break source (gradient_min, observations) gets segmented. The other fall is invisible to segmentation, producing segments that span the fall and incorrectly classify the upper portion as accessible.
+
+Confirmed pattern on Horsefly River (BT, link-only credit on segment 12671 spanning 1447m through fall #2). Issue filed with full trace evidence at link#96. Not part of `frs_order_child` work — separate, simpler fix (one entry in source_tables + one entry in break_order default + bundle config update).
+
+**Tackle on this branch (96-frs-order-child-wire) before unparking the bypass work.** The fix is small enough that it can land here without expanding scope; the test is re-running HORS BT and confirming 12671 splits at 67565 and the upper portion becomes inaccessible. After confirmation, update `research/bcfishpass_comparison.md` to reflect.
+
+### Next session quick-start
+
+1. Read this file
+2. Read `findings.md`
+3. Read fresh#158 issue body (canonical design)
+4. Apply link#96 fix on this branch first (falls in break_order)
+5. Re-run HORS BT to verify 12671 splits at the fall
+6. Update `research/bcfishpass_comparison.md`
+7. THEN decide on the bypass methodology question
diff --git a/planning/active/task_plan.md b/planning/active/task_plan.md
index fb9be01..862d81e 100644
--- a/planning/active/task_plan.md
+++ b/planning/active/task_plan.md
@@ -1,37 +1,36 @@
-# Task Plan — link#88: Fold subsurfaceflow into natural barriers
-
-## Phase 1: Setup
-- [x] File link#88 with diagnosis + proposed fix
-- [x] Branch `88-fold-subsurfaceflow-natural` from main
-- [x] PWF baseline (task_plan, findings, progress)
-
-## Phase 2: Code change
-- [x] Extend `.lnk_pipeline_prep_natural(conn, aoi, cfg, loaded, schema)` to absorb subsurfaceflow body, gated on `"subsurfaceflow" %in% cfg$pipeline$break_order`
-- [x] Append subsurfaceflow rows to `<schema>.natural_barriers` (label `blocked`)
-- [x] Delete `.lnk_pipeline_prep_subsurfaceflow` helper
-- [x] Remove conditional call from `lnk_pipeline_prepare()`
-- [x] `devtools::document()` — refresh roxygen for prep_natural
-
-## Phase 3: Tests
-- [x] `tests/testthat/test-lnk_pipeline_prepare.R`: subsurfaceflow opted in → INSERT into natural_barriers fires (per-statement assertion)
-- [x] Same file: subsurfaceflow not opted in → no subsurfaceflow code path runs
-- [x] Same file: subsurfaceflow honours barriers_definite_control
-- [x] `devtools::test(filter = "lnk_pipeline_prepare")` clean (44/44 PASS)
-
-## Phase 4: Code-check
-- [x] `/code-check` on staged diff — 3 rounds. Round 3 caught fragile cross-statement regex; tightened to per-statement `any(grepl & grepl)`. Final clean.
-
-## Phase 5: Verification
-- [x] HARR single-WSG pre-flight `tar_make` — `data-raw/logs/20260430_11_preflight_harr_link88.txt`
-- [x] blkey 356286055 BT rearing credits **6.509 km** (was 0)
-- [x] Full 15-WSG `tar_make` — `data-raw/logs/20260430_12_tar_make_15wsg_link88.txt` (53m 2.2s, 33/33 targets)
-- [x] HARR CH/CO/ST rearing_stream closed to ±0.32% (BT residual -4.2%, separate mechanism noted)
-- [x] LFRA CH/CO/ST closed to ±0.6% (BT residual -3.75%)
-- [x] HORS unchanged (-7.68% BT) — different mechanism, follow-up issue
-- [x] Default-bundle rollup bit-identical (0 of 581 link_value rows changed)
-- [x] Reproducibility: second `tar_make` byte-identical (`link_value` digest `5a641892b82604259b0ba168ea093661` matches across runs; 0 of 1057 rows differ)
-
-## Phase 6: Ship
-- [x] Atomic commits with PWF checkbox flips
-- [ ] PR with `Fixes #88` and `Relates to NewGraphEnvironment/sred-2025-2026#24`
-- [ ] Archive PWF after merge
+# Task Plan — link 96-frs-order-child wire-up (parked)
+
+Branch: `96-frs-order-child-wire` (local, not pushed)
+Canonical design doc: [fresh#158](https://github.com/NewGraphEnvironment/fresh/issues/158) — read its issue body for the full decision record. Predecessor [fresh#156](https://github.com/NewGraphEnvironment/fresh/issues/156) closed in favor of #158. Spawning-side misread: [link#23](https://github.com/NewGraphEnvironment/link/issues/23) (closed not-a-bug 2026-04-28).
+
+## Phase 1 — fresh ships (DONE)
+
+- [x] fresh 0.27.1 — validator allows `channel_width_min_bypass` predicate (PR #194)
+- [x] fresh 0.27.3 — `frs_order_child` derives `stream_order_max` per BLK via CTE (PR #196). 0.27.2 was a false-start patch superseded by 0.27.3.
+- [x] fresh 0.27.4 — validator allows `distance_max` key inside the bypass block (PR #197)
+
+## Phase 2 — link wiring (DONE, parked uncommitted)
+
+- [x] `dimensions.csv`: 3 new columns (`rear_stream_order_bypass`, `rear_stream_order_parent_min`, `rear_stream_order_distance_max`) — both bundles
+- [x] `lnk_rules_build`: emits all three into `channel_width_min_bypass:` block
+- [x] `lnk_pipeline_classify`: reads block, calls `fresh::frs_order_child(parent_order_min, child_order_min/max, distance_max)`
+- [x] Tests added in `test-lnk_rules_build.R` for the new columns
+
+## Phase 3 — single-WSG calibration (DONE)
+
+- [x] HORS BT preflight at `bypass=yes, parent=5, child=1, dmax=300`: link 394 km / bcfp 396 km on `rearing_stream`
+- [x] Five iteration snapshots saved as HTML in `data-raw/maps/HORS_BT_rearing_AFTER_*.html`
+
+## Phase 4 — parked
+
+Methodology decision pending. fresh#158 design intent: link is **not** chasing bcfp parity for `frs_order_child` — it's a link primitive expressing the biology of *"small streams plugging into big rivers support rearing despite low estimated CW"*, with parametric flex (`stream_order_max`, `distance_max`). HORS calibration result is numerical proximity by accident; methodology is divergent from bcfp by design.
+
+- [ ] Decide whether to ship as link methodology default or keep `rear_stream_order_bypass=no` until a wider sample (BABL inspection per fresh#158) informs the call
+- [ ] If shipping default-on: 15-WSG `tar_make` to confirm calibration generalizes (or doesn't)
+- [ ] If shipping default-off: keep the wire-up infrastructure but bundle CSVs ship `bypass=no`
+
+## Notes for next session
+
+- Don't re-derive `stream_order_max` / `distance_max` rationale — fresh#158 issue body has it
+- Don't re-derive bcfp parity gap — bcfp's bypass also lacks `stream_order_max` filter, and runs *inside* its 3 connectivity-aware phases (rearing-on-spawning / DS-of-spawn cluster / US-of-spawn cluster). Our `frs_order_child` runs post-cluster, post-classify; it's an additive primitive, not a parity replicator. fresh#158 documents this trade-off.
+- The `dmax=300` HORS calibration is exploratory, not a final value
diff --git a/tests/testthat/test-lnk_rules_build.R b/tests/testthat/test-lnk_rules_build.R
index f422116..d310fc5 100644
--- a/tests/testthat/test-lnk_rules_build.R
+++ b/tests/testthat/test-lnk_rules_build.R
@@ -868,6 +868,74 @@ test_that("default config rules.yaml has no 1050/1150/2100 in spawn or rear-stre
   }
 })
 
+test_that("rear_stream_order_parent_min column drives bypass threshold", {
+  # Default 5L when bypass=yes and column absent; explicit value when present.
+  base <- list(species = "BT", spawn_lake = "no", spawn_stream = "yes",
+               rear_lake = "no", rear_lake_only = "no", rear_no_fw = "no",
+               rear_stream = "yes", rear_wetland = "no",
+               rear_stream_order_bypass = "yes")
+
+  # No column → default 5
+  csv <- withr::local_tempfile(fileext = ".csv")
+  out <- withr::local_tempfile(fileext = ".yaml")
+  utils::write.csv(as.data.frame(base, stringsAsFactors = FALSE), csv,
+                   row.names = FALSE)
+  suppressMessages(lnk_rules_build(csv, out, edge_types = "explicit"))
+  rules <- yaml::read_yaml(out)
+  bp <- NULL
+  for (rr in rules$BT$rear) {
+    if (!is.null(rr$channel_width_min_bypass)) { bp <- rr$channel_width_min_bypass; break }
+  }
+  expect_false(is.null(bp))
+  expect_equal(bp$stream_order, 1L)
+  expect_equal(bp$stream_order_parent_min, 5L)
+
+  # Column present, explicit value 7
+  with7 <- c(base, list(rear_stream_order_parent_min = "7"))
+  csv2 <- withr::local_tempfile(fileext = ".csv")
+  out2 <- withr::local_tempfile(fileext = ".yaml")
+  utils::write.csv(as.data.frame(with7, stringsAsFactors = FALSE), csv2,
+                   row.names = FALSE)
+  suppressMessages(lnk_rules_build(csv2, out2, edge_types = "explicit"))
+  rules2 <- yaml::read_yaml(out2)
+  bp2 <- NULL
+  for (rr in rules2$BT$rear) {
+    if (!is.null(rr$channel_width_min_bypass)) { bp2 <- rr$channel_width_min_bypass; break }
+  }
+  expect_equal(bp2$stream_order_parent_min, 7L)
+
+  # Column present but empty → default 5
+  with_empty <- c(base, list(rear_stream_order_parent_min = ""))
+  csv3 <- withr::local_tempfile(fileext = ".csv")
+  out3 <- withr::local_tempfile(fileext = ".yaml")
+  utils::write.csv(as.data.frame(with_empty, stringsAsFactors = FALSE), csv3,
+                   row.names = FALSE)
+  suppressMessages(lnk_rules_build(csv3, out3, edge_types = "explicit"))
+  rules3 <- yaml::read_yaml(out3)
+  bp3 <- NULL
+  for (rr in rules3$BT$rear) {
+    if (!is.null(rr$channel_width_min_bypass)) { bp3 <- rr$channel_width_min_bypass; break }
+  }
+  expect_equal(bp3$stream_order_parent_min, 5L)
+})
+
+test_that("rear_stream_order_parent_min has no effect when bypass=no", {
+  base <- list(species = "BT", spawn_lake = "no", spawn_stream = "yes",
+               rear_lake = "no", rear_lake_only = "no", rear_no_fw = "no",
+               rear_stream = "yes", rear_wetland = "no",
+               rear_stream_order_bypass = "no",
+               rear_stream_order_parent_min = "7")
+  csv <- withr::local_tempfile(fileext = ".csv")
+  out <- withr::local_tempfile(fileext = ".yaml")
+  utils::write.csv(as.data.frame(base, stringsAsFactors = FALSE), csv,
+                   row.names = FALSE)
+  suppressMessages(lnk_rules_build(csv, out, edge_types = "explicit"))
+  rules <- yaml::read_yaml(out)
+  for (rr in rules$BT$rear) {
+    expect_null(rr$channel_width_min_bypass)
+  }
+})
+
 test_that("default config rules.yaml retains 1050/1150 in dedicated wetland-rear rule", {
   yaml_path <- system.file("extdata", "configs", "default", "rules.yaml",
                            package = "link")