|
<br />1371
<br />
<br />Desolation-Gray Canyon
<br />
<br />ST A TUS OF GREEN RIVER COLORADO PIKEMINNOW
<br />
<br />lower Green River
<br />
<br />200]
<br />
<br />'"1
<br />]00
<br />~ 80
<br />~ 60
<br />1 ~ .......i.......
<br />
<br />
<br />
<br /><&> <&> ,,~ <&> ,<&> <&> <Jo~ <&> <&> '"
<br />"" ~' ~'" <5'''! &" <5'~ "" ~~ .s;y" ~'l"
<br />-..; "0 '1) ~ "3 ~ ^i q,'J'
<br />
<br />2002
<br />
<br />:~J
<br />g 80
<br />~ 60
<br />J: 40 d
<br />2~ ......... r .... r. . . . .
<br />
<br />\':f~ ,~ ,.,'to, ,.,,~o, ;)'-0, /;)~o, tJ'o, '\~ $' ?J~
<br />,<S'O'~,,~,,<S'~~#~~'b~ ~
<br />
<br />:~~ I
<br />~ 80
<br />g 60
<br />
<br />1 ~~ ., ........ ~.........I.I. . . . , , ,
<br />f!',,~ <&> '~Y'~ '~<Jo~ <&> <&>#
<br />'" ,~' o,cS?' ,,~"! ~ ~~~ # ~~~ 'b~" ~
<br />
<br />2003
<br />
<br />Colorado pikeminnow TL (mm)
<br />
<br />'ID ili
<br />g 120
<br />
<br />g 80
<br />[
<br />"' 40
<br />___I
<br />o " , " "U... ..
<br />
<br />200]
<br />
<br />. . I .
<br />
<br />!>~ <&> ~ ~ o)r~ ~,~ J" ,,<&> ~ ~<S'
<br />'" ,~' oeftS'" ,,~"! ,~ ~~ 'c~ ~~ 'b~" ~
<br />
<br />2002
<br />
<br />~::L.
<br />g 80
<br />if
<br />~ 40
<br />o ... . r......,.. . . , , . ,
<br />~ <&> <&> <&>_~ ~ J" ~ # #
<br />or ,~' o,~'" "",or"! ~ ~~ # ~~ 'b~ ~
<br />
<br />2003
<br />
<br />'IDj
<br />g 120
<br />g 80
<br />
<br />1 4~ ..... ,.~..I,I"l. ~.... . . . , , .
<br />!>~ <&> <&> ,~ o)r~ ,~ J" ,,<&> <&> ~<S'
<br />'" ~' ~'" <5'''! ~ <5'~ <5' ~ <5''' ~
<br />"; "Y '1) ~ ') B ^i q;
<br />
<br />FIGURE 9,-Continued.
<br />
<br />Colorado pikeminnow TL (mm)
<br />
<br />within a sampling season (reflected in the values of 0/),
<br />which suggest that movement into or out of the
<br />sampling area was unlikely. The relatively short period
<br />encompassed by the spring sampling passes also
<br />ensured that mortality and recruitment were minimized.
<br />The effect of the heterogeneity of capture probabil-
<br />ities of individuals due to fish size was minimized by
<br />adding the covariate TL to analyses where appropriate,
<br />which adjusted capture probabilities over the range of
<br />fish lengths encountered. This was important because
<br />sampling crews noted that large fish were sometimes
<br />more difficult to capture. We also investigated the
<br />likelihood of a population composed of groups of
<br />individuals with inherently different capture probabil-
<br />ities (e.g., behavioral or size differences). We found no
<br />support for differing capture probabilities based on
<br />preliminary analyses using mixture models (Pledger
<br />2000); such models typically require five or more
<br />sampling passes to detect differences (G.C.W., person-
<br />al observation). We also demonstrated that initial
<br />capture probabilities were equal to recapture probabil-
<br />ities among the short-term sampling passes (i.e., Pk =
<br />ck) by including an additive parameter to the top model
<br />(see model 4 in Table 1). This additive model
<br />suggested that recapture probabilities were nearly
<br />
<br />identical to initial capture probabilities. Because
<br />minimal differences in the initial capture rates and
<br />recapture rates suggested that confounding factors such
<br />as fish avoidance of sampling boats did not introduce
<br />heterogeneity into capture probabilities, the p's and c's
<br />were held equal in the Huggins models. Thus, to the
<br />extent possible, we tested for and found no effects of
<br />heterogeneity other than that attributable to fish body
<br />size.
<br />Another relevant assumption in this study was that
<br />
<br />fJ
<br />'2
<br />u
<br />Q.l
<br />...
<br />6'2.
<br />
<br />80.0
<br />
<br />60.0
<br />
<br />40,0
<br />
<br />20.0
<br />
<br />0,0
<br />1990 1992 1994 1996 1998 2000 2002 2004
<br />
<br />
<br />Year
<br />
<br />FIGURE ro.-Percentage of recruit-size (400-449 mm TL)
<br />Colorado pikeminnow relative to the number of adults (2'450
<br />mm) in ISMP samples collected from the Green River basin,
<br />1991-2003,
<br />
|