|
NOTE
<br />substantial: over 7,000 electrofishing samples (typical-
<br />ly 5-min stations) were collected in the main-stem
<br />Colorado River, and there were over 34,000 overnight
<br />hoop net sets in the main-stem Colorado River and
<br />LCR (Figure 2). For each fish, total length (mm) and
<br />capture location (usually to the nearest 0.1 river km)
<br />were recorded and a PIT tag was implanted in the
<br />abdominal cavity to serve as a unique fish identifier for
<br />fish larger than 150 mm total length. Each fish was
<br />then released near its capture location.
<br />Captures and recaptures were recorded in seven
<br />main-stem Colorado River reaches that were de-
<br />termined based on geomorphology, logistics, and
<br />easily identifiable areas (Figure 1; Table 1). In addition,
<br />two tributaries (LCR and Havasu Creek) were also
<br />identified as reaches along with the inflow areas of
<br />these tributaries (e.g., main-stem Colorado River near
<br />the confluence of these tributaries). Restricted distri-
<br />bution was defined as recapture and previous capture
<br />location within the same river reach or tributary. Since
<br />other studies have suggested that the humpback chub
<br />population in the main-stem Colorado River within 12
<br />km of the LCR (i.e., LCR inflow) is the same
<br />population (Douglas and Marsh 1996; Gorman and
<br />Stone 1999), we also analyzed distribution by com-
<br />bining the LCR inflow and the LCR as one river reach,
<br />thus creating six river reaches instead of seven.
<br />Logistic regression was used to determine if a size-
<br />group of fish was more likely to exhibit restricted
<br />distribution based on odds ratios (Stokes et al. 1995).
<br />Logistic regression was also used to determine if the
<br />proportion of fish that exhibited restricted distribution
<br />differed by time at liberty. In this analysis, time at
<br />liberty was separated into eight categories: 14-30 d,
<br />31-90 d, 91-120 d, 121-365 d, 1-3 years, 3-5 five
<br />years, 5-10 years, and more than 10 years. An analysis
<br />of variance was used to determine if the mean distance
<br />541
<br />between capture and recapture location (for fish that
<br />exhibited restricted distribution) differed by fish size
<br />and by river (i.e., main-stem Colorado River and the
<br />LCR) for three times between capture and recapture
<br />(14-90 d, 91-365 d, >365 d).
<br />Results
<br />A total of 14,674 recaptures of 7,127 unique fish
<br />were recorded in Grand Canyon from May 1989 to
<br />October 2002. Of the 7,127 unique fish, 3,322 (46%)
<br />were recaptured only once, whereas 219 (3%) were
<br />recaptured more than five times and one was recaptured
<br />15 times. The mean number of days between captures
<br />was 394 (range = 14-4,713). However, 75% of the
<br />recaptures occurred in less than 404 d.
<br />Of the 14,674 total recaptures, 12,868 (87.7%) were
<br />recaptured in the same tributary or area of the main-stem
<br />Colorado River (Table 1) and thus exhibited restricted
<br />distribution. For fish exhibiting a restricted distribution,
<br />the mean number of days between capture and recapture
<br />was 532 (range = 22-3,700) for the main-stem Colorado
<br />River, and 354 (range =14-4,713) for the LCR. For fish
<br />not exhibiting a restricted distribution, the mean number
<br />of days between capture and recapture was 773 (range:
<br />14-3,742) for the main-stem Colorado River and 515
<br />(range = 14-3,808) for the LCR.
<br />Of the total recaptures, 12,508 (85.2%) were
<br />captured and recaptured in the LCR, whereas an
<br />additional 241 (1.6%) were captured and recaptured in
<br />the main-stem Colorado River within 12 km of the
<br />LCR confluence (Table 1). When we considered the
<br />LCR and LCR inflow as one population, 14,526
<br />(99.0%) of recaptures occurred in the LCR or the main
<br />stem near the LCR confluence and thus exhibited
<br />a restricted distribution to this reach. Although a high
<br />percentage of fish showed restricted distribution to the
<br />LCR and surrounding areas, there was evidence a few
<br />TABLE 1.-Recaptures of humpback chub collected throughout the Colorado River in Grand Canyon, Arizona, 1989-2002.
<br />Values along the diagonal (bold italics) indicate that the capture and recapture locations were the same (i.e., site fidelity). Major
<br />tributaries are the Little Colorado River (LCR) and Havasu Creek; rkm =river kilometers. Values above the diagonal represent
<br />downstream movement, values below the diagonal upstream movement.
<br />Recapture location (don)
<br />278 Total
<br />Tag location Won) 0-117 124 (LCR) 117-136 136-199 199-201 201-227 227-244 (Havasu Creek) 244-389 recaptures
<br />0-117 26 1
<br />124 1 12,508 868 5
<br />117-136 909 241 3 1 1
<br />136-199 2 1 1 1
<br />199-201 1 8
<br />201-227 1 1
<br />227-244 2 77
<br />278 1 1
<br />244-389
<br />14,674
|