|
<br />Experimental Challenges and Biases
<br />Our results were conservative in terms of benefits derived [1\ pn:ddtol e\plhLlll' \\illk illl'
<br />capability of handling stress for our study groups was idenllcaL treatl1lerll fi:--h recelvl'd helld\I\lIdl dllll
<br />physical trauma that control fish did not. Some treatment fish (both bnnylail and ld/olhdLk "lIlkl'l" I
<br />exhibited physical bruising (for example. discoloration. scale I()~',;, ,ll1d rin dill\L\:,'l'l In III! 1,llll'll 1'1;'1 i. I,"
<br />attacks, which undoubtedly placed them at a performance disadvillll:rg,' I.II~ 1-1 i l I11I >llilll-\1 I
<br />small number of fish we had to work with prevented us from lemovlllg dllll~h Illdl ~lll'\1 ,'Ll ,1~li' i'l
<br />physical injury. Only those exhibiting difficulty in swimming or had major fin damage well' rel1lO\ ,'d
<br />We suspect some of the bruising and possibly internal injuries went undetected and contributed [owald
<br />poor escape performance leading to predation losses. This might be avoided in future studies by
<br />exposing large groups, for example, 100 prey and one predator.
<br />In 2006, we used ten fish test groups where predation occurred over several days, In 20m, II
<br />became evident that after the first day, all the prey had been exposed to predatol aggles,,\ol1 al1d \\ ,'Il'
<br />exhibiting predator avoidance. We felt the benefit of pre-treatment decreased with edl'il pd~~II1~ d,111\1
<br />address this, we shortened the duration and group size in the experimenls WL' ()j1ll'd III ,'\,lllIllh'
<br />predation that occurred within a day or two. rather than Idlc" ill'\llIll"ll (1\ ,'I 'Vi ,'I ,i ,i,I'
<br />Treatment and control fish schooled together. which Wd." ,ll1o\hl'l Id,.[OI \I,' Ild,I'i,'1 ""I'l
<br />would be a problem. Studies have shown that survivorship of some hatcherY-leared rl~h 111(ll'iN' II 11\,'\
<br />are released among wild cohorts (Hvidsten and Johnsen, 1993). Berejikian and olhl'r~ I ~()()() I "li~:,'l"ll'cI
<br />control fish may acquire anti-predatory behavior more quickly when released wllh Iraillt'd 11,11 111Iilll,'!'
<br />social learning processes. Brown and Laland (200 I ) recolllmended Ih~11 I rl'd I Jl)l' 11\ .Iild '\llllll,( I,,!'
<br />should be tested independently due to this problem: this is contrarv lO our apP\'(ldl'hlh' "II(1\>!III:' ill
<br />control fish with the treatment groups Illay have Inddvencl1llv pJ.lll'li lhl'lii 111,1 Ihlt,'i :',..,," .
<br />they had been introduced to predators hy lhelll'il'!v L'''
<br />
<br />
<br />Figure 14. Bonytail exhibiting bruising, discoloration, and scale loss due to f,lilec1 :111:[1 !: ".i",, )
<br />during predator, pre-exposure treatments Many trlell Ilsl' S,dt['!"
<br />during treatment sessions,
<br />
<br />20
<br />
|