|
<br />Erfecl~ of lJli on invertebrales
<br />
<br />were probed with a blunt instrument; if no movement was
<br />observed, they were counted as affected (i.e., dead) larvae.
<br />Animals that could not be found at the end of the tests were
<br />also counted as affected larvae.
<br />On May 25 and June 22, 1989 (verification teSlS IA and
<br />I B for enclosure test I), the tested mosquito genus was
<br />Culisela spp. On April 25, 1990 (verification test 2 for en-
<br />closure test 2), and May 8, 1990 (verification test 3 for en-
<br />closure test 3), the genus of mosquito Aedes spp. was tested.
<br />Culisela spp. was tested on May 19, 1990 (verification test
<br />4 for enclosure test 4).
<br />Field and laboratory static acute tests were conducted on
<br />chironomid larvae found in the enclosures to evaluate the ef-
<br />fect of Bli at field application rates under controlled condi-
<br />tions. Some tests included mosquito larvae to verify the
<br />viability of the Vectobac-G. The procedures described above
<br />were used for the static acute tests, unless otherwise noted.
<br />Static test 1 was begun July 26, 1989, using 25 ml LBP
<br />sediment per chamber and water from a control enclosure,
<br />chironomid larvae (Micropseclra [TanYlarsus] spp.) from the
<br />MP, and mosquito larvae (Psorosphera spp.) collected on the
<br />Refuge. There were three replicates each of control, RAR,
<br />2 x RAR and 5 X RAR. Concentrations of pesticide were
<br />mixed in wash bottles with I L pond water and the appro-
<br />priate amount of Vectobac-G corncob granules. Wash bot-
<br />tle contents were mixed for approximately 20 min. After the
<br />chironomid larvae had burrowed into the sediment (approx-
<br />imately 24 h), the appropriate amount of each Bti concen-
<br />tration was added to the test chambers. The final volume of
<br />water in the chambers was 200 ml. Chambers were randomly
<br />placed on a shelf in the field laboratory.
<br />Further static acute tests were conducted, beginning on
<br />May 1 (static test 2), May 8 (static test 3), May 19 (static
<br />test 4), and June 8 (static test 5), 1990. Three extra enclosures
<br />were erected, Vectobac-G was applied, and water grab sam-
<br />ples were collected to be used in the tests. Chironomid larvae
<br />were collected from the pond where the test was conducted.
<br />For static test 2, chironomid larvae (Chironomus. Dicro-
<br />tendipes, and Paratanytarsus spp.) were collected from LBP
<br />and placed in 100- x 50-mm crystallizing dishes in the field
<br />laboratory with 200 ml of treated MP water. In static test 3,
<br />chironomid larvae (Dicrotendipes spp.) from the LBP were
<br />placed in 300-ml test chambers containing treated LBP wa-
<br />ter that were put in LBP in a staked milk crate. Static test
<br />4 used the same procedures as those for static test 3,except
<br />that Paratendipes spp. larvae were tested. These chironomid
<br />larvae were collected from MP, and the test was conducted
<br />there with MP treated water. For static test 5,25 ml,sieved
<br />MP sediment and 100 ml control MP water were added to
<br />each test chamber, and the sediment was allowed to settle for
<br />24 h. Chironomid larvae (Micropsectra [Tanytarsus) and
<br />Paratendipes spp.) were then added to the test chambers and
<br />allowed to acclimate for 24 h. Next, 200 ml treated MP wa-
<br />ter was slowly added to the chambers to prevent sediment dis-
<br />turbance. Mosquito larvae (Culex spp.) were added to the test
<br />chambers, which were then placed in a staked milk crate in
<br />MP. The various toxicity tests are summarized in Table I.
<br />Statistical analysis. For all enclosure tests, only data from
<br />Amphipoda, Chironomidae, and Oligochaeta were analyzed.
<br />For this paper the term benthic organisms refers to these
<br />
<br />269
<br />
<br />Table I. Summary of toxicity teslS conducted in the field and
<br />laboratory to assess the efficacy of Vectobac-GI!J
<br />on chironomid larvae
<br />
<br />Test name
<br />
<br />Test location
<br />
<br />Verification tests lA, I B, 2, 3, 4
<br />Slatic tests I, 2
<br />Static tesls 3, 4, 5
<br />Range tests I, 2, 3, 4
<br />Temp. test
<br />Water depth tests I, 2
<br />Field depth test
<br />Macrophyte tests I, 2
<br />Water source test
<br />Food test
<br />Organic matter test
<br />I nstar test
<br />
<br />Laboratory
<br />Laboratory
<br />Field
<br />Laboratory
<br />Laboratory
<br />Laboratory
<br />Field
<br />Laboratory
<br />Laboratory
<br />Laboratory
<br />Laboratory
<br />Laboratory
<br />
<br />three groups. To normalize distributions, the distribution of
<br />means that were skewed (univariate procedure (43)) were
<br />transformed (log x + I) [36]. General linear model procedures
<br />(44) were used to conduct ANOVA Ftests to determine if the
<br />Vectobac-G treatments resulted in fewer benthic inverte-
<br />brates. Effects of treatment, time and grouping of enclosures,
<br />and interactions between classes were determined by least-
<br />squares means and tested at a = 0.05.
<br />Spatial dispersion of benthic invertebrate populations are
<br />usually contiguous (clumped), with density variance greater
<br />than the mean (36). Clumping complicates determination of
<br />population differences and may lead to an erroneous conclu-
<br />sion that no differences exist (type 2 error). Therefore, to fa-
<br />cilitate our interpretations, we used "power analysis," in
<br />addition to the normal ANOVA procedures: the more pow-
<br />erful a test, the more confident a researcher can be of detect-
<br />ing an effect if it exists [45-48). If significant differences are
<br />not detected, power analysis helps to conclude either that
<br />there were no differences or that the variance among sam-
<br />ples was so great that if differences existed, they could not
<br />be detected. The powers of the enclosure tests were deter-
<br />mined using PowerPack@ (49) and tested at {J = 0.30.
<br />For enclosure test I, data from the two pretreatment sam-
<br />ples were averaged (May 1 and May 8, 1989). Differences be-
<br />tween pretreatment and post-treatment means for all
<br />applications were compared by ANOVA. Differences be-
<br />tween the averaged pretreatment means and final (June 28,
<br />1989) post-treatment means were compared using ANOVA
<br />to determine if a series of applications of Vectobac-G over
<br />time affected benthic organisms.
<br />For enclosure test 2, differences between the pretreatment
<br />(April 23, 1990) and post-treatment (Apri126, 1990) means
<br />were tested by ANOVA.
<br />There were no pretreatment samples taken for enclosure
<br />tests 3 and 4. When data were analyzed using ANOVA, only
<br />the differences between the control and two treatment lev-
<br />els were tested.
<br />
<br />Laboratory toxicity tests
<br />
<br />General methods. Laboratory toxicity tests were con-
<br />ducted to evaluate an LC50 and factors responsible for mit-
<br />
|