> '` V!bjbjLULU ?&.?.?B %:::::::N8N
(2 "T T T /
/
/
$ph/E:/
/
/
/
/
/::T T t
/
:T :T
/
::
T Ա%
0
nn
n:
d/
/
/
/
/
/
/
///
/
/
/
/
/
/
NNN2dNNN2NNN::::::Supplemental text S1
Reparameterized von Bertalanffy growth function methods
Recent work has raised questions about the interpretability of Linf and k, and a number of authors have proposed a re-parameterized version of the VBGF (e.g. Francis 1988). The re-parameterized version of the VBGF includes three parameters: L(), L(), L(), which estimate body size at arbitrary ages , , and . The ages and are chosen to represent the early rapid growth phase and the later slower growth phase respectively, and is the mean of and . The re-parameterized version of the VBGF is:
EMBED Equation.3 , where Lt is length at time t, and EMBED Equation.3 subject to the constraints that L() < L() < L() and [L()- L()] `" [L() - L()].
Because a great deal of previous work examining age and growth in fishes used the traditional form of the VBGF but more recent work has shifted to the re-parameterized version of the VBGF (rVBGF), we report parameter estimates for the traditional VBGF in the body of the manuscript, and include parameter estimates for the rVBGF here to allow for comparisons across studies.
Reparameterized von Bertalanffy growth function results
For the rVBGF, we set (age during early fast growth phase) to 1 year for all species, and set (adult age, when growth is slowing) to 6 years for P. arcatus, C. margaritifer, and P. dickii. We set to 12 years for L. bohar since it reaches adult size much later on Palmyra,MN
8
:
<
>
nuv4@Rhjٿٷ٩ٔٷنyrjjjrh3h'6h'hdnjVhzh'EHUjoN
hHMOh'UVhgh'H*jhzh'EHUjN
hHMOh'UVjh'Uhfh'6h%fh'6H*h%fh'6h'hYwh'5 h5 h'5hq)h'5hq)hq)5)N
mn
@!
!!!!!!!0d7$8$H$^`0gd0d7$8$H$^`0gd'
d1$`gd' d1$gd'
!T!4L\^26^v@ !!!
!!!!!!!!"!$!0!2!4!8!:!F!H!Ҽҵh@
h@0Jjh@0JUh7jh7Uhqh'hqhh@h'\hDrhXh'6]hXh'hXh'5 h'6hh'hdnUh'h3h'67 and set to 5 years for A. nigricans because, despite its relatively long lifespan, it reaches asymptotic length relatively quickly, characteristic of many acanthurid species (Choat and Robertson 2002). For L. bohar and C. margaritifer, L() and L() were significantly greater at Palmyra, while for A. nigricans, L() and L() were significantly greater at Kiritimati, based on bootstrapped 95% confidence intervals. L(), size at age 1, did not differ for any species between islands (Online Appendix Table 1).
Appendix Literature Cited
Choat, J. H., and D. R. Robertson. 2002. Age-based studies. Pages 57-80 in P. F. Sale, editor. Coral Reef Fishes: Dynamics and Diversity in a Complex Ecosystem. Academic Press, San Diego.
Francis, R. I. C. C. 1988. Are growth-parameters estimated from tagging and age length data comparable? Canadian Journal of Fisheries and Aquatic Sciences 45:936-942.
PAGE
PAGE 2
! !"!4!6!8!N!P!R!T!V!&`#$gd'
H!J!L!N!R!T!V!h'h7h@
h@0Jjh@0JUh@0JmHnHu=0P&P1h:p'/ =!"#$%VDd
t4b
c$A??3"`?2( /$|D6
`!t( /$ͼ 0#pBxڥ?hQ.M$"HŶ[A :
"B!VKcM8"Ypr)"d:ԢB~.D?lgcd!cb!Y'~o襼}aJ1_ >6-1Gk0[k]nՅ2O-Bgw ϱEi>i݄im?M+m-˚~*FzwF.ޥF&\yӮϫ|o(]R>tZӮ:(^eo$3TL}`P[é/&wbNq8S\5|UOy..#eW#^s29;9(-yAsOp G9~<
`2tp7($9ȲX],Wd:ŇvɊ/xG ;p 'Ruttenberg et alTitle
FMicrosoft Office Word Document
MSWordDocWord.Document.89q@@@'NormalCJ_HaJmH sH tH DA@DDefault Paragraph FontRiRTable Normal4
l4a(k(No List4 @4'Footer
!.)@.'Page Number.(@.'Line Numberh
&NMNnoA B D E G H J K M N W X Y d e f i 0000000000000000@000@000@000@000@0@0@0@0@0@000
A i @0@0 $$$'H!V! !V!
T!Nbdh :: '!!8@0(
B
S ?%:?o~[eotB B D D E E G H J K M N f i @Bce;B B D D E E G H J K M N f i 333333qqvw#% " % * . 3 ? @ B B D D E E G H J K M N f i qqvw#% " % * . 3 ? @ A B B D D E F G I J L M N V Y c i @Dr@6g";KVXrm0q)7B'8@2 2 ;2 2 @h @
@UnknownGz Times New Roman5Symbol3&z Arial"1h,lfpf
aa!4> > 2qHX ?'2Ruttenberg et alBenjamin RuttenbergBenjamin Ruttenberg