9530

• UCREFRP Catalog Number: 9530
• Author: Nelson, S. M. and D. C. Andersen
• Title: Butterfly ‎(Papilionoidea and Hesperioidea)‎ Assemblages Associated with Natural, Exotic, and Restored Riarian Habitats along the Lower Colorado River, USA
• USFW Year: 1999
• USFW - Doc Type: Regulated Rivers
• Copyright Material: YES

492
100
E
E
c
0
.Q
~U
N
L
Q
a
c
0
(LS
80
60
40
20
S.M. NELSON AND D.C. ANDERSEN
------------------
~ l^^^ _ I I _^_.^^ l l~l In ^^^^^ I n ^^ fl n I I f ^^ 11 I II I1'
~~9 4. A P 6 01 0 9,~0,~~.,~q...A ti 9 M b 6 1 0 oi,~0,~'~,~'~a ~t. 9 P <a 6 1 0 N,~O,~'.,~'~(o ~L 9 b y 6 1 '0 9,~0,~~.,~~L..'~ ry A P h 61 0 9
J~ J~ X66'' J~ aP ~~ a~ v~ ~°
Month
Figure 4. Total monthly precipitation in study area. Graphed values are-the mean of total monthly precipitation for Needles and
Blythe, California and Lake Havasu City and Parker; Arizona. Dark bars represent winter months (October-March). Data from
iNational Oceanic and Atmospheric Administration
The amount of nectar produced by a `plant growing on the xerified floodplai`r is probably related to
depth to groundwater. Tamarisk flowers were. abundant- at Tam, situatednext to the river, presumably
from the presence of shallow ground water. Revegetated sites away from the river also contained tamarisk
but less nectar perhaps because of a lower water table. Negative responses of nectar to high groundwater
were observed in some cases. High groundwater at R96 may have been responsible for Bermuda grass
(Cynodon dactylon) competitively eliminating low-lying nectar sources (e.g.; Melilotus, Trianthema portu-
lacastrum, and Heliotropium curassavicum) that were readily accessible in March 1997 but effectively
concealed by the grass in November 1997.
Assemblage analysis
Correspondence analysis suggested that the butterfly assemblage at BW is unique and that assemblages
at revegetation sites are not trending toward it. CA differentiated the butterfly assemblages of BW from
those of Tam and the revegetated sites in each of the three seasonal periods examined (Figure 8a-c;
Tables II, III and IV). The hypothesis that the CA pattern explained by the data-set was the same as a
randomized data-set was rejected (p < 0.05) for the first several axes for both the March and June data
sets. The first two axes derived from these ordinations accounted for 45% of the variation in the
assemblage in March and 33% in June. Only the first axis (27% of variation) differed from the random
data-set in August.
In no case did age of revegetation site result in a closer ordination to BW assemblages. Rather, Tam
and the revegetation sites were scattered in ordination space in a manner unrelated to age. Correlation of
Axis I and Axis II scores with age of revegetation sites were all non-significant (p > 0.4). However, a
faunal basis for the ordination patterns was evident in the differences in occurrence and abundance of
butterfly taxa (Tables II, III and IV). March ordination (Figure 8a; Table II) grouped revegetated sites
from the Cibola area that contained Pyrgus communis and Hesperopsis gracielae but lacked taxa unique

Copyright ©1999 John Wiley & Sons, Ltd. Regul. Ricers: Res. Mgmt. 15: 485-504 (1999)