Aedes taeniorhynchus (Diptera: Culicidae) Oviposition Patterns in a Florida Mangrove Forest SCOTT A. RITCHIE1 AND ERIC S. JOHNSON2
J. Med. Entomol. 28(4): 496-500 (1991)
ABSTRACT The association of Aedes taeniorhynchus eggs and several variables was studied in a Florida mangrove forest. Eggs were limited to stands of red mangrove (Rhizophora mangle L.) that were embedded within a black mangrove (Avicennia germinans L.) forest. The occurrence of eggs was related significantly to elevation and the amount of detritus. Field and laboratory studies indicated that grazing on black mangrove detritus by the snail Melampus coffeus L. may have limited detritus accumulation and soil organic content, thus restricting mosquito oviposition to the red mangrove stands. Eggshells were concentrated in the same habitat as eggs, suggesting that eggshells may be used to identify oviposition patterns. KEY WORDS
Insecta, Aedes taeniorhynchus, egg distribution, mangrove forests
1 Current address: Collier Mosquito Control District, P.O. Box 7069, Naples, Fla. 33941 2 Current address: Emory University, P.O. Box 22587, Atlanta, Ga. 30322.
of eggs were representative of most oviposition events. Materials and Methods Study Sites. The study site was a mangrove basin forest located on Marco Island in Collier County, Fla. At its northern end, the basin featured a monospecific black mangrove stand with a firm sandy soil and a thin detritus covering; this area will be referred to as the black mangrove zone. At its southern end, red mangrove (Rhizophora mangle L.) was mixed with black mangrove, creating a substrate of thick red mangrove detritus overlaying a layer of peat; this area will be referred to as the red/black mangrove zone. The geomorphology of the basin created a flooding regime (Ritchie 1990) that frequently produced salt marsh mosquitoes. The basin was topographically simple with a gentle sloping contour and scattered depressions of low relief (maximum depth 6.25 m) in red mangrove stands. The low relief allowed the basin to dry quickly, frequently exposing sites suitable for mosquito oviposition. The basin was usually dry from November to May except for occasional tidal inundations. Heavy rains and higher tides frequently flooded the basin from June to October. An isolated ovipositional event was used to identify the relationship of several environmental variables to Ae. taeniorhynchus oviposition. The study site had dried on 25 July 1987 following a prolonged submergence from 26 June to 24 July 1987. Because of the high hatching percentage of summer eggs («95% [Ritchie 1988]), most of the eggs present were laid between 25 and 28 July in a relatively discrete ovipositional event. By sampling on 28 July, the distribution of eggs could be matched with that of selected environmental variables to identify potential oviposition cues and indices. Parameters of interest were soil moisture, the abundance of the pulmonate snail, Melampus coffeus L. (a mangrove detritivore [Mook 1985] observed
0022-2585/91/0496-0500$02.00/0 © 1991 Entomological Society of America
Downloaded from http://jme.oxfordjournals.org/ by guest on June 6, 2016
FLOODWATER MOSQUITOES EXHIBIT some common ovipositional behaviors. Laboratory studies indicate that several species preferentially oviposit in response to substrate moisture (Knight & Baker 1962, Meek & Williams 1986). Field collected eggs frequently are distributed in horizontal bands that represent the preferred soil moisture zone (Lefkovitch & Brust 1968, Horsfall et al. 1973, Olson & Meek 1977, Novak 1981, Fallis & Snow 1983, Ritchie & Johnson 1991). Thick detritus (perhaps by limiting soil desiccation) also has been associated with high egg densities (Horsfall et al. 1973, Russo 1978, Ritchie & Johnson 1991). Specific plant-mosquito egg associations have been identified and used to indicate potential oviposition sites (Horsfall 1963, Scotton & Axtell 1979, Dale et al, 1986). Black mangrove (Avicennia germinans L.) basin forests are considered to be the typical oviposition site for Aedes taeniorhynchus Wiedemann in south Florida (Provost 1977). However, mangrove basin forests may feature other mangrove species (red mangrove, Rhizophora mangle L. and white mangrove, (Laguncularia racemosa Gaertn.) that can greatly affect the substrate (Twilley et al. 1986). The relationship of Ae. taeniorhynchus oviposition to these variables is poorly understood. Detritalrich areas of red mangrove basin forests may contain high densities of Ae. taeniorhynchus eggs (Ritchie & Johnson 1991) and eggshells (Ritchie & Johnson 1989). Egg and eggshell distribution were used to test if detrital-rich soil of red mangrove stands might serve as ovipositional foci within a black mangrove forest. Egg distribution was used to show ovipositional preferences for a discrete oviposition event. Because eggshells represent historic oviposition (Lopp 1957, S.A.R., unpublished data), their distribution was used to test if the distribution
July 1991
RITCHIE
&
JOHNSON:
Ae. taeniorhynchus
497
openings of 0.185 and 0.170 mm (Ritchie & Addison 1991). Data from each zone was pooled over the different sampling dates and the number of eggshells per cm3 of soil compared with a MannWhitney U test (Schlotzhauer & Littell 1987). Similarly collected 15-cm3 soil cores (n = 10 for each zone; collected on 5 May 1990) were analyzed for percent organic carbon using the ignition loss procedure (Dean 1974); data were transformed by arcsine (Zar 1974) and compared with an unpaired t test. Detritus Decomposition Rates. The decomposition rates of red and black mangrove leaves were compared in the field to see if differences could account for the differential accumulation of red and black mangrove litter. Similarly sized, succulent red and black mangrove leaves were collected from the forest floor and secured individually sideby-side to the soil with a tooth-pick. In trial 1, four leaf pairs were set at five sites from 22 to 25 July 1987. In trial 2, six leaf pairs each were placed at four sites from 28 July to 2 August 1987. A control leaf, placed in a fiber glass screen pouch (0.15-cm openings) to prevent direct access by snails, was placed at each test site. Because changes in water content affected leaf weight, leaf damage was arbitrarily scaled. Field observations indicated that skeletonization was initiated on the leaf surface (surface skeletonization), spread over the leaf and then deepened (complete skeletonization) until only the thickest leaf veins remained. This feeding pattern has been documented for M. coffeus on red mangrove leaves (Lopez et al. 1977). Leaf damage was scaled as follows: 0 (no damage), 1 (50% intact in each sample. Water table depth (cm) was measured for each core hole to provide an estimate of the relative elevation. The mean number of hatched eggs (log [x + 1] transformed) for each zone were compared by an unpaired t test. Stepwise linear regression (Schlotzhauer & Littell 1987) was used to find the variable^) that accounted for significant variability in the transformed number of eggs (transformed by log [x + 1]), snails and red and black mangrove leaves per sample. Because egg distribution was very clumped, egg occurrence also was used as the dependent variable. Dummy variables of 0 and 1 were used respectively to represent the absence and presence of eggs (Kleinbaum & Kupper 1978) and were regressed against the independent variables used in the stepwise regression. The program Surfer (Golden Software, Inc., Golden, Colo.) was used to construct surface contour plots, so the distribution of the variables in the grid could be compared visually. Kriging (inverse distance weighting power = 2) was used to estimate grid points from the 10 nearest data points. Cubic splining (X and Y expansion factor = 2) was used to smooth the surface contour. Eggshell density and soil organic carbon were sampled randomly along transects bisecting each zone. Soil samples were collected with modified plastic syringes: 3- and 15-cm3 soil cores were taken with 6- and 60-cm3 respectively. The black mangrove zone was sampled during August 1988 (n = 24) and December 1989 (n = 20) while the red/ black mangrove zone was sampled on May 1988 (n = 24), May 1989 (n = 24) and December 1989 (n = 20). Samples were processed for eggshells by selective sieving using nested brass sieves with mesh
OVIPOSITION
498
JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 28, no. 4
Downloaded from http://jme.oxfordjournals.org/ by guest on June 6, 2016
Fig. 1. Surface contour plots for (A) the number of Ae. taeniorhynchus eggs and (B) red mangrove leaves >50% intact from 100 soil cores collected in a Florida mangrove forest on 28 July 1987 in a 9 by 60 m sampling grid. The boundary of the red/black mangrove zone and black mangrove zone is approximated by the presence of red mangrove detritus (Fig. IB).
mangrove zone and 4.2 and 40.2% in the black mangrove zone, respectively. Eggshell density (x ± SE) was significantly (t = 5.641, df = 106, P < 0.001) higher in the red/black mangrove zone (0.89 ± 0.10 shells/cm3) than the black mangrove zone (0.04 ± 0.02 shells/cm3). Stepwise regression analysis indicated that only red mangrove detritus contributed significantly (R2 = 0.16; df = 1, 98; t = 4.04; P < 0.001) to the variability of transformed number of eggs per sample; inclusion of the other variables (relative elevation and the number of black mangrove leaves and M. coffeus per sample) increased the R2 slightly (from 0.16 to 0.17). Egg occurrence also was significantly related to red mangrove leaf density (R2 = 0.287; df = 1,99; F = 39.926; P < 0.001); addition of the other variables increased the R2 slightly (R2 = 0.314; df = 4, 96; F = 11.003; P < 0.001). Additional detritus may have accounted for the significantly higher (t = 4.896, df = 18, P < 0.001) percentage of organic carbon (x ± SE) in the red/ black mangrove soil (24.4 ± 1.8%) than in the soil from the black mangrove zone (11.9 ± 2.1%). There was a positive relationship between relative elevation and the number of mosquito eggs. This analysis only involved samples positive for red mangrove litter (n = 60) because mosquito eggs were limited almost strictly to this habitat. The regression of transformed number of eggs per sam-
Fig. 2. Surface contour plots for (A) the number of black mangrove leaves >50% intact, (B) relative elevation in cm and (C) the number of M. coffeus /sample in the sample grid.
pie against relative elevation (EL) was not significant (P = 0.91). However, the quadratic model (using EL and EL2) was significant (R2 = 0.155; df = 2, 57; F = 5.22; P = 0.008), suggesting that oviposition was concentrated in a preferred soil moisture zone located at intermediate elevations. Similar results were found using egg occurrence as the dependent variable (R2 = 0.370; df = 2, 58; F = 17.03; P < 0.001).
July 1991
RITCHIE
&
JOHNSON:
Ae. taeniorhynchus
Discussion Aedes taeniorhynchus showed a clear ovipositional preference for the red/black mangrove zone, which eggshell distribution indicated was representative of previous oviposition events at our site. These data refine Provost's (1977) concept that black mangrove basin forests represent the primary domain of salt marsh mosquito production in southern Florida. In some black mangrove forests, patches of red mangrove create areas attractive to ovipositing Ae. taeniorhynchus. The larger amount of detritus within the red/black mangrove zone provided greater cover, which may protect eggs from desiccation. Surface detritus also may be more attractive to resting mosquitoes, with high oviposition rates simply an artifact of this behavior. Greater surface detritus and higher soil organic content also provided for a larger larval food supply. Several factors may have interacted to make the black mangrove substrate unattractive to ovipositing Ae. taeniorhynchus. Perhaps oviposition is associated directly or indirectly with the type of mangrove detritus. However, oviposition has been
499
observed in forests dominated by white or black mangrove (S.A.R., unpublished data), indicating that oviposition is independent of mangrove type. Perhaps the lack of extensive surface litter and the low soil organic content made the black mangrove zone unattractive for oviposition. If so, the mechanisms accounting for differential detritus and soil formation under red versus black mangroves could be the key to oviposition patterns at our study site. Several mechanisms reduce detritus in monospecific black mangrove forests. Monospecific black mangrove forests have lower litter fall and litter standing crop than do mixed red and black mangrove forests (Twilley et al. 1986). Black mangrove leaves are thinner and decompose faster than red mangrove leaves (Twilley et al. 1986). The pulmonate snail, M. coffeus, may be a significant factor in the rapid decomposition of black mangrove litter. Heald (1971) reported that these snails can "quickly reduce fallen black mangrove leaves to a bare skeleton". Other potential detrivores, such as amphipods (Heald 1971), also could be significant mangrove detritivores. Elevation within a mangrove forest may affect the distribution of detritivores and thus detritus accumulation. Tidal flushing significantly reduced detritus in mangrove basin forests (Twilley et al. 1986). Perhaps the impact of tidal flushing on litter standing crop is greatest in areas with low detrital cover (e.g., the black mangrove zone). Our results indicated the limitations of using egg and eggshell sampling to elucidate ovipositional preferences. Suggars et al. (1986) also found that the contagious distribution of floodwater mosquito eggs made the identification of the environmental parameters significantly associated with oviposition difficult. We overcame this statistical problem by using egg occurrence rather than egg density as an index of oviposition. Egg-shell density also can be used to delineate patterns of aedine oviposition, although data must be interpreted with caution. Eggshell distribution represents historical oviposition and may not reflect current patterns. Nonetheless, the economy of this method (Ritchie & Addison 1991) is allowing us to examine the relationship between Ae. taeniorhynchus oviposition and site characteristics at a large number of mangrove basin forests and test the validity of these results beyond the range of a single study site.
Acknowledgment The authors wish to acknowledge Brandt G. Watson of the Collier Mosquito Control District for his kind support. They also acknowledge J. H. Frank, James Heaney, and Jon Allen of the University of Florida and Daniel Kline and Richard Brenner of the United States Department of Agriculture—Medical and Veterinary Entomology Research Laboratory for their helpful criticism of the manuscript and Sarah Linney and Jessica Miller of the Conservancy, Inc. for helping to collect and process samples.
Downloaded from http://jme.oxfordjournals.org/ by guest on June 6, 2016
Because thin detritus covering in the monospecific black mangrove forest may limit oviposition, regression analysis was used to examine the relationship of black mangrove litter, relative elevation, and M. coffeus populations. The number of black mangrove leaves per sample (F = 5.06; df = 4, 96; P =
J. Med. Entomol. 28(4): 496-500 (1991)
ABSTRACT The association of Aedes taeniorhynchus eggs and several variables was studied in a Florida mangrove forest. Eggs were limited to stands of red mangrove (Rhizophora mangle L.) that were embedded within a black mangrove (Avicennia germinans L.) forest. The occurrence of eggs was related significantly to elevation and the amount of detritus. Field and laboratory studies indicated that grazing on black mangrove detritus by the snail Melampus coffeus L. may have limited detritus accumulation and soil organic content, thus restricting mosquito oviposition to the red mangrove stands. Eggshells were concentrated in the same habitat as eggs, suggesting that eggshells may be used to identify oviposition patterns. KEY WORDS
Insecta, Aedes taeniorhynchus, egg distribution, mangrove forests
1 Current address: Collier Mosquito Control District, P.O. Box 7069, Naples, Fla. 33941 2 Current address: Emory University, P.O. Box 22587, Atlanta, Ga. 30322.
of eggs were representative of most oviposition events. Materials and Methods Study Sites. The study site was a mangrove basin forest located on Marco Island in Collier County, Fla. At its northern end, the basin featured a monospecific black mangrove stand with a firm sandy soil and a thin detritus covering; this area will be referred to as the black mangrove zone. At its southern end, red mangrove (Rhizophora mangle L.) was mixed with black mangrove, creating a substrate of thick red mangrove detritus overlaying a layer of peat; this area will be referred to as the red/black mangrove zone. The geomorphology of the basin created a flooding regime (Ritchie 1990) that frequently produced salt marsh mosquitoes. The basin was topographically simple with a gentle sloping contour and scattered depressions of low relief (maximum depth 6.25 m) in red mangrove stands. The low relief allowed the basin to dry quickly, frequently exposing sites suitable for mosquito oviposition. The basin was usually dry from November to May except for occasional tidal inundations. Heavy rains and higher tides frequently flooded the basin from June to October. An isolated ovipositional event was used to identify the relationship of several environmental variables to Ae. taeniorhynchus oviposition. The study site had dried on 25 July 1987 following a prolonged submergence from 26 June to 24 July 1987. Because of the high hatching percentage of summer eggs («95% [Ritchie 1988]), most of the eggs present were laid between 25 and 28 July in a relatively discrete ovipositional event. By sampling on 28 July, the distribution of eggs could be matched with that of selected environmental variables to identify potential oviposition cues and indices. Parameters of interest were soil moisture, the abundance of the pulmonate snail, Melampus coffeus L. (a mangrove detritivore [Mook 1985] observed
0022-2585/91/0496-0500$02.00/0 © 1991 Entomological Society of America
Downloaded from http://jme.oxfordjournals.org/ by guest on June 6, 2016
FLOODWATER MOSQUITOES EXHIBIT some common ovipositional behaviors. Laboratory studies indicate that several species preferentially oviposit in response to substrate moisture (Knight & Baker 1962, Meek & Williams 1986). Field collected eggs frequently are distributed in horizontal bands that represent the preferred soil moisture zone (Lefkovitch & Brust 1968, Horsfall et al. 1973, Olson & Meek 1977, Novak 1981, Fallis & Snow 1983, Ritchie & Johnson 1991). Thick detritus (perhaps by limiting soil desiccation) also has been associated with high egg densities (Horsfall et al. 1973, Russo 1978, Ritchie & Johnson 1991). Specific plant-mosquito egg associations have been identified and used to indicate potential oviposition sites (Horsfall 1963, Scotton & Axtell 1979, Dale et al, 1986). Black mangrove (Avicennia germinans L.) basin forests are considered to be the typical oviposition site for Aedes taeniorhynchus Wiedemann in south Florida (Provost 1977). However, mangrove basin forests may feature other mangrove species (red mangrove, Rhizophora mangle L. and white mangrove, (Laguncularia racemosa Gaertn.) that can greatly affect the substrate (Twilley et al. 1986). The relationship of Ae. taeniorhynchus oviposition to these variables is poorly understood. Detritalrich areas of red mangrove basin forests may contain high densities of Ae. taeniorhynchus eggs (Ritchie & Johnson 1991) and eggshells (Ritchie & Johnson 1989). Egg and eggshell distribution were used to test if detrital-rich soil of red mangrove stands might serve as ovipositional foci within a black mangrove forest. Egg distribution was used to show ovipositional preferences for a discrete oviposition event. Because eggshells represent historic oviposition (Lopp 1957, S.A.R., unpublished data), their distribution was used to test if the distribution
July 1991
RITCHIE
&
JOHNSON:
Ae. taeniorhynchus
497
openings of 0.185 and 0.170 mm (Ritchie & Addison 1991). Data from each zone was pooled over the different sampling dates and the number of eggshells per cm3 of soil compared with a MannWhitney U test (Schlotzhauer & Littell 1987). Similarly collected 15-cm3 soil cores (n = 10 for each zone; collected on 5 May 1990) were analyzed for percent organic carbon using the ignition loss procedure (Dean 1974); data were transformed by arcsine (Zar 1974) and compared with an unpaired t test. Detritus Decomposition Rates. The decomposition rates of red and black mangrove leaves were compared in the field to see if differences could account for the differential accumulation of red and black mangrove litter. Similarly sized, succulent red and black mangrove leaves were collected from the forest floor and secured individually sideby-side to the soil with a tooth-pick. In trial 1, four leaf pairs were set at five sites from 22 to 25 July 1987. In trial 2, six leaf pairs each were placed at four sites from 28 July to 2 August 1987. A control leaf, placed in a fiber glass screen pouch (0.15-cm openings) to prevent direct access by snails, was placed at each test site. Because changes in water content affected leaf weight, leaf damage was arbitrarily scaled. Field observations indicated that skeletonization was initiated on the leaf surface (surface skeletonization), spread over the leaf and then deepened (complete skeletonization) until only the thickest leaf veins remained. This feeding pattern has been documented for M. coffeus on red mangrove leaves (Lopez et al. 1977). Leaf damage was scaled as follows: 0 (no damage), 1 (50% intact in each sample. Water table depth (cm) was measured for each core hole to provide an estimate of the relative elevation. The mean number of hatched eggs (log [x + 1] transformed) for each zone were compared by an unpaired t test. Stepwise linear regression (Schlotzhauer & Littell 1987) was used to find the variable^) that accounted for significant variability in the transformed number of eggs (transformed by log [x + 1]), snails and red and black mangrove leaves per sample. Because egg distribution was very clumped, egg occurrence also was used as the dependent variable. Dummy variables of 0 and 1 were used respectively to represent the absence and presence of eggs (Kleinbaum & Kupper 1978) and were regressed against the independent variables used in the stepwise regression. The program Surfer (Golden Software, Inc., Golden, Colo.) was used to construct surface contour plots, so the distribution of the variables in the grid could be compared visually. Kriging (inverse distance weighting power = 2) was used to estimate grid points from the 10 nearest data points. Cubic splining (X and Y expansion factor = 2) was used to smooth the surface contour. Eggshell density and soil organic carbon were sampled randomly along transects bisecting each zone. Soil samples were collected with modified plastic syringes: 3- and 15-cm3 soil cores were taken with 6- and 60-cm3 respectively. The black mangrove zone was sampled during August 1988 (n = 24) and December 1989 (n = 20) while the red/ black mangrove zone was sampled on May 1988 (n = 24), May 1989 (n = 24) and December 1989 (n = 20). Samples were processed for eggshells by selective sieving using nested brass sieves with mesh
OVIPOSITION
498
JOURNAL OF MEDICAL ENTOMOLOGY
Vol. 28, no. 4
Downloaded from http://jme.oxfordjournals.org/ by guest on June 6, 2016
Fig. 1. Surface contour plots for (A) the number of Ae. taeniorhynchus eggs and (B) red mangrove leaves >50% intact from 100 soil cores collected in a Florida mangrove forest on 28 July 1987 in a 9 by 60 m sampling grid. The boundary of the red/black mangrove zone and black mangrove zone is approximated by the presence of red mangrove detritus (Fig. IB).
mangrove zone and 4.2 and 40.2% in the black mangrove zone, respectively. Eggshell density (x ± SE) was significantly (t = 5.641, df = 106, P < 0.001) higher in the red/black mangrove zone (0.89 ± 0.10 shells/cm3) than the black mangrove zone (0.04 ± 0.02 shells/cm3). Stepwise regression analysis indicated that only red mangrove detritus contributed significantly (R2 = 0.16; df = 1, 98; t = 4.04; P < 0.001) to the variability of transformed number of eggs per sample; inclusion of the other variables (relative elevation and the number of black mangrove leaves and M. coffeus per sample) increased the R2 slightly (from 0.16 to 0.17). Egg occurrence also was significantly related to red mangrove leaf density (R2 = 0.287; df = 1,99; F = 39.926; P < 0.001); addition of the other variables increased the R2 slightly (R2 = 0.314; df = 4, 96; F = 11.003; P < 0.001). Additional detritus may have accounted for the significantly higher (t = 4.896, df = 18, P < 0.001) percentage of organic carbon (x ± SE) in the red/ black mangrove soil (24.4 ± 1.8%) than in the soil from the black mangrove zone (11.9 ± 2.1%). There was a positive relationship between relative elevation and the number of mosquito eggs. This analysis only involved samples positive for red mangrove litter (n = 60) because mosquito eggs were limited almost strictly to this habitat. The regression of transformed number of eggs per sam-
Fig. 2. Surface contour plots for (A) the number of black mangrove leaves >50% intact, (B) relative elevation in cm and (C) the number of M. coffeus /sample in the sample grid.
pie against relative elevation (EL) was not significant (P = 0.91). However, the quadratic model (using EL and EL2) was significant (R2 = 0.155; df = 2, 57; F = 5.22; P = 0.008), suggesting that oviposition was concentrated in a preferred soil moisture zone located at intermediate elevations. Similar results were found using egg occurrence as the dependent variable (R2 = 0.370; df = 2, 58; F = 17.03; P < 0.001).
July 1991
RITCHIE
&
JOHNSON:
Ae. taeniorhynchus
Discussion Aedes taeniorhynchus showed a clear ovipositional preference for the red/black mangrove zone, which eggshell distribution indicated was representative of previous oviposition events at our site. These data refine Provost's (1977) concept that black mangrove basin forests represent the primary domain of salt marsh mosquito production in southern Florida. In some black mangrove forests, patches of red mangrove create areas attractive to ovipositing Ae. taeniorhynchus. The larger amount of detritus within the red/black mangrove zone provided greater cover, which may protect eggs from desiccation. Surface detritus also may be more attractive to resting mosquitoes, with high oviposition rates simply an artifact of this behavior. Greater surface detritus and higher soil organic content also provided for a larger larval food supply. Several factors may have interacted to make the black mangrove substrate unattractive to ovipositing Ae. taeniorhynchus. Perhaps oviposition is associated directly or indirectly with the type of mangrove detritus. However, oviposition has been
499
observed in forests dominated by white or black mangrove (S.A.R., unpublished data), indicating that oviposition is independent of mangrove type. Perhaps the lack of extensive surface litter and the low soil organic content made the black mangrove zone unattractive for oviposition. If so, the mechanisms accounting for differential detritus and soil formation under red versus black mangroves could be the key to oviposition patterns at our study site. Several mechanisms reduce detritus in monospecific black mangrove forests. Monospecific black mangrove forests have lower litter fall and litter standing crop than do mixed red and black mangrove forests (Twilley et al. 1986). Black mangrove leaves are thinner and decompose faster than red mangrove leaves (Twilley et al. 1986). The pulmonate snail, M. coffeus, may be a significant factor in the rapid decomposition of black mangrove litter. Heald (1971) reported that these snails can "quickly reduce fallen black mangrove leaves to a bare skeleton". Other potential detrivores, such as amphipods (Heald 1971), also could be significant mangrove detritivores. Elevation within a mangrove forest may affect the distribution of detritivores and thus detritus accumulation. Tidal flushing significantly reduced detritus in mangrove basin forests (Twilley et al. 1986). Perhaps the impact of tidal flushing on litter standing crop is greatest in areas with low detrital cover (e.g., the black mangrove zone). Our results indicated the limitations of using egg and eggshell sampling to elucidate ovipositional preferences. Suggars et al. (1986) also found that the contagious distribution of floodwater mosquito eggs made the identification of the environmental parameters significantly associated with oviposition difficult. We overcame this statistical problem by using egg occurrence rather than egg density as an index of oviposition. Egg-shell density also can be used to delineate patterns of aedine oviposition, although data must be interpreted with caution. Eggshell distribution represents historical oviposition and may not reflect current patterns. Nonetheless, the economy of this method (Ritchie & Addison 1991) is allowing us to examine the relationship between Ae. taeniorhynchus oviposition and site characteristics at a large number of mangrove basin forests and test the validity of these results beyond the range of a single study site.
Acknowledgment The authors wish to acknowledge Brandt G. Watson of the Collier Mosquito Control District for his kind support. They also acknowledge J. H. Frank, James Heaney, and Jon Allen of the University of Florida and Daniel Kline and Richard Brenner of the United States Department of Agriculture—Medical and Veterinary Entomology Research Laboratory for their helpful criticism of the manuscript and Sarah Linney and Jessica Miller of the Conservancy, Inc. for helping to collect and process samples.
Downloaded from http://jme.oxfordjournals.org/ by guest on June 6, 2016
Because thin detritus covering in the monospecific black mangrove forest may limit oviposition, regression analysis was used to examine the relationship of black mangrove litter, relative elevation, and M. coffeus populations. The number of black mangrove leaves per sample (F = 5.06; df = 4, 96; P =
