USGS Home
Contact USGS
Search USGS

In contrast, fish were nearly absent from no-fish wetlands in both years. Given the predominance of fathead minnows in fish-treatment wetlands, and their near absence in no-fish sites, we believe it is appropriate to attribute all "fish effects" to fathead minnows. Effects of Fish:

Results of MANOVA performed on the four categories of amphibians indicated significant fish effects in all 3 study years (1996 F=23.24; 4,13 df; P<.01, 1997 F=8.31; 4,13 df; P<.01, and 1998 F=12.29; 3,14 df; P<.01), but reflected no significant history or interaction effects (all P>.05) (Table 1). Given these large differences in amphibians among wetland groups, we ran separate ANOVAs on our 4 groups of amphibians. Abundance of adult tiger salamanders was lower in wetlands with minnows in 1997 (F=15.24; 1,16 df; P<.01) and 1998 (F=13.54; 1,16df; P<.01), but these differences were not evident during 1996 (P=.31) (Figure 3). Adult frogs were less abundant in wetlands with fish in 1997 (F=4.79; 1,16 df; P<.05) but not in 1996 (P=.24). It is interesting to note that we captured no adult frogs during 1998. Trends in larval amphibians were generally consistent; wetlands with minnows contained fewer larval tiger salamanders during all study years (1996 F=60.80; 1,16 df; P<.01, 1997 F=30.96; 1,16 df; P<.01, and 1998 F=28.86; 1,16 df; P<.01) as well as fewer larval frogs (1996 F=3.58; 1,16 df; P<.05, 1997 F=9.41; 1,16 df; P<.01). One exception was larval frogs which did not differ between wetlands with and without minnows in 1998 (F=.57; 1,16 df; P>.05).

Abundance of several important groups of aquatic invertebrates was consistently lower in wetlands with fathead minnows. Column sample data reflected lower abundance of large cladocerans (1996 F=26.87; 1,16 df; P<.01, 1997 F=9.29; 1,16 df; P<.01, 1998 F=51.98; 1,16 df; P<.01), calanoid copepods (1996 F=6.85; 1,16 df; P<.02, 1997 F=22.01; 1,16 df; P<.01, 1998 F=8.25; 1,16 df; P=.02), ostracods (1996 F=5.12; 1,16 df; P<.04, 1997 F=11.38; 1,16 df; P<.01, 1998 F=5.66; 1,16 df; P=.03), and aquatic insects (1996 F=23.60; 1,16 df; P<.01, 1997 F=5.13; 1,16 df; P<.04, 1998 F=17.96; 1,16 df; P<.01) (Figure 4a, Figure 4b). Significant differences were not observed for small cladocerans, cyclopoid copepods, or amphipods during any year (all P>.05) (Figures 4a,b).

Activity trap data also indicated that wetlands with fathead minnows had fewer large cladocerans (1996 F=61.03; 1,16 df; P<.01, 1997 F=74.36; 1,16 df; P<.01, 1998 F=7.18; 1,16 df; P=.02), as well as fewer calanoid copepods (1996 F=6.54; 1,16 df; P<.03, 1997 F=5.43; 1,16 df; P<.04, 1998 F=9.36; 1,16 df; P<.01) (Figure 5a). Total aquatic insects were also less abundant in fish-treatment wetlands in 1997 (1996 F=7.41; 1,16 df; P<.02) and 1998 (F=16.14; 1,16 df; P<.01), but similar differences were not evident during 1996 (P=.84) (Figure 5b). Based on our activity trap data, abundance of small cladocerans, cyclopoid copepods, and amphipods were not consistently reduced by fathead minnows (all P>.05) in any study year. Ostracods were captured at lower numbers in wetlands with fish during 1998 (F=18.63; 1,16 df; P<.01), but not in 1996 or 1997 (both P>.25).

Throughout 1996-1998, fathead minnows had consistent influences on turbidity, chlorophyll a, and total phosphorus in our study wetlands. Turbidity was higher in wetlands with fathead minnows (1996 F=17.26; 1,16 df; P<.01, 1997 F=16.51; 1,16 df; P<.01, 1998 F=5.6; 1,16 df; P<.05), as was chlorophyll a (1996 F=15.53; 1,16 df; P<.01, 1997 F=17.17; 1,16 df; P<.01, 1998 F=25.96; 1,16 df; P<.01) and total phosphorus (1996 F=11.67; 1,16 df; P=.01, 1997 F=5.20; 1,16 df; P<.04, 1998 F=13.74; 1,16 df; P<.01) (Figure 6). In contrast, Kjeldahl nitrogen showed no relation with fish presence/absence in any study year (P>.05) (Figure 6).

Our aquatic plant MANOVA indicated no significant fish, history, or interactions in any study year (all P>.05) (Table 2), thus we performed no further analyses with these data.

Effects of Wetland History:
Wetland history had little influence on wetland characteristics and no significant differences were detected for any variable in multiple years. Larval tiger salamanders were less abundant in restored sites than in natural wetlands in 1997 (F=7.95; 1,16 df; P<.02) (Figure 3). Column samples indicated that calanoid copepods were less abundant in restored wetlands in 1996 (F=7.07; 1,16 df; P<.02) and ostracods were less abundant in natural in wetlands in 1996 (F=5.58; 1,16 df; P<.05) (Figures 4a,b). No comparable history effects were found using activity trap data (all P>.05) (Figure 5a, Figure 5b). Total phosphorus concentrations were higher in restored wetlands in 1996 (F=7.85; 1,16 df; P=.01), but similar difference were not detected in other study years (P>.05) (Figure 6). Turbidity, chlorophyll a, and Kjeldahl nitrogen showed no relation to wetland history in any study year (all P>.05) (Figure 6).

Fish-History Interactions:
Few significant fish-history interactions were evident indicating that for the most part, wetlands responded similarly to presence of fathead minnows, regardless of wetland history. For amphibians, the only significant interaction we observed was with adult tiger salamanders in 1996 (F=6.59; 1,16 df; P<.03) (Figure 3). Here, amphibians were less abundant in restored wetlands with fish than in restored-fishless sites (P<.02). Column sample data also identified significant interactions with calanoid copepods in 1997 (F=9.10; 1,16 df; P<.01) and amphipods in 1996 (F=4.92; 1,16 df; P<.05); in both cases these groups were most abundant in restored, fishless sites (Figure 4a, Figure 4b). Large cladocera were sparse in wetlands with fathead minnows and least abundant in restored-fish sites during 1996 (activity trap data) (F=12.31; 1,16 df; P<.01) (Figure 5a). In 1996, amphipods were more abundant in restored-no fish wetlands relative to all other treatment groups (column sample data) (P<.05). Finally, in 1997, turbidity (F=4.69; 1,16 df; P<.05) and chlorophyll a (F=4.87; 1,16 df; P<.05) (P>.05) were higher in sites with fathead minnows and highest values came from restored-fish wetlands (Figure 6).