9530 - Laserfiche WebLink

Home Browse Search

488 S.M. NELSON AND D.C. ANDERSEN the entire lower Colorado River; the No Name Lake Revegetation Site (within the Colorado River Indian Tribe Reservation); and atamarisk-dominated site and two revegetation sites on the Cibola National Wildlife Refuge. A Bureau of Reclamation dredge spoil revegetation site near the Cibola refuge and two revegetation sites on the Havasu National Wildlife Refuge were examined beginning in 1996, 1996, and 1997, respectively. Tamarisk site. The 14 ha tamarisk site (Tam) was next to the Colorado River in the Cibola National Wildlife Refuge. The site was 70-80% tamarisk, with small amounts of quailbrush (Atriplex lentiformis), Emory baccharis (Baccharis emoryi), arrowweed (Tessaria sericea), mesquite, and a few Gooddings willow (Salix gooddingii) and cottonwood plants. The site shifted from mostly tamarisk to arrowweed on the north and to mesquite on the east. Some coyote willow was present a}ong a dike separating this site from the river. Bill Williams- River site. The Bill Williams River site (BW), near the river's mouth at Lake Havasu, contained severa] hundred hectares of naturally reproducing riparian vegetation. Woody vegetation included awell-developed canopy of Fremont cottonwood and Gooddings willow along with large stands of tamarisk. Seepwillow [Baccharis salicifolia (= B. glutinosa)] and arrowweed were understory plants at ` this site. Coyote willow (Salix exigua) was found in some areas. Various perennial and annual herbaceous plants were also present. Higher floodplain areas contained honey and screwbean mesquite (Prosopis juliflora and P. pubescens), quailbrush, four-wing saltbush (Atriplex canescens), and creosote bush. Stream flow was intermittent with surface water often absent during ,the summer. Revegetation sites. Characteristics-of revegetated sites are presented in Table I. Most revegetated sites were within xerified floodplain, over a relatively deep water table. The exceptions were the Havasu Refuge sites, which were in areas of high groundwater. Revegetated sites were typically planted in evenly spaced rows and then supplied with surface water via drip or flood irrigation until pole plantings or seedlings became phreatophytic. No natural establishment of either Fremont cottonwood or Gooddings willow occurred at any revegetated site. Thus, stands were all single-aged. Following tamarisk removal, the study sites can be considered to lie along a chronosequence from young revegetation, through older revegeta- tion, to mature community. Butterfly monitoring Individual butterflies were counted during timed searches, which provided data on both species presence and an index of relative abundance. Sweep nets were used where required for verification or identification of taxa. Two person hours per day on 2 different days were spent sampling each site during each approximately 1 week long session. We conducted two-four sampling sessions each year for the years 1994-1997. Sampling took place during all 4 years in March and June, 3 years in August, and 2 years in November. Sites were visited twice during each session to mitigate against confounding effects of time of day and weather conditions. Assemblage analysis Trends in species richness (from combined March and June observations) relative to site age (the number of years since a site had been planted) were examined using Spearman rank correlation analysis. The tamarisk site was assigned the age of zero. We hypothesized that butterfly species richness at the revegetated sites would become similar to that at the natural riparian area as plantings matured and, therefore, be positively correlated with site age. Correspondence analysis (CA) was used to ordinate sites in species space and describe assemblage patterns across sites. Because of seasonal shifts in assemblage make-up, separate analyses were performed for March, June, and August observations. There were too few data to analyze November surveys. We hypothesized that as sites increased in age, the types and numbers of butterflies seen there would become more similar and ordinate nearer to the Bill Williams assemblage. First axis CA scores are produced by iterations between site scores based on weighted species scores and species scores based on weighted site scores (Gauch, 1982). Multiple axes may be extracted using the same Copyright ©1999 John Wiley & Sons, Ltd. Regul. Rivers: Res. Mgmt. 15: 48S-SO4 (1999)