Proceedings of the Third International Conference on Invasive ...

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementDethier 2006). In addition, it can provide substantialaboveground structure not normally present in somehabitats. These structural and biogeochemical modificationshave direct and indirect effects on o<strong>the</strong>r plant and animalspecies as mentioned earlier. In our study system, Englishcordgrass differentially modifies four habitat types, resultingin variable, community-wide consequences (Hacker andDethier 2006). For example, we know that <strong>the</strong> accumulation<strong>of</strong> sediment by cordgrass differs between <strong>the</strong> four habitats,with mudflats experiencing <strong>the</strong> greatest sediment accretion(~30 cm) and low and high salinity marshes, <strong>the</strong> least (~10-15 cm) (Hacker and Dethier 2006). Cordgrass growing incobble beaches accretes much less sediment (~20 cm) thanmudflats. A likely explanation for this difference is waveenergy; water movement is lower in mudflats but higher incobble beaches.In addition to accreting sediment, thus changing <strong>the</strong>structural characteristics <strong>of</strong> <strong>the</strong> substrate, cordgrass causes anumber <strong>of</strong> chemical transformations. We found changes insediment water content, redox potential (a proxy for oxygencontent), and salinity in invaded versus native areasdepending on community type (Hacker and Dethier 2006).For example, cordgrass caused a decline in sediment watercontent in mudflats and high salinity marshes suggesting that<strong>the</strong> elevated root mat had less tidal inundation, betterdrainage, and/or more water uptake by cordgrass. In cobblebeaches, <strong>the</strong> opposite was seen; root mat sediments hadhigher water content than <strong>the</strong> unmodified cobble sediments.In low salinity marshes, <strong>the</strong>re was no change in sedimentwater content with <strong>the</strong> presence <strong>of</strong> cordgrass compared tonative vascular plants. Cordgrass generally increases oxygencontent in <strong>the</strong> sediments <strong>of</strong> all <strong>the</strong> communities. This may berelated, in part, to better drainage seen in some <strong>of</strong> <strong>the</strong>communities but is more likely due to oxygen leakage fromroot aerenchyma (Maricle and Lee 2002). Finally, surfacesalinities were generally lower in sediments with cordgrassalthough <strong>the</strong>y did not change in low salinity sites with nativeplant cover. These results suggest that cordgrass shades <strong>the</strong>sediment surface thus decreasing water evaporation and saltaccumulation compared to unvegetated areas although thishypo<strong>the</strong>sis was untested. We have not investigated cordgrasseffects on nutrient cycling but it is clear that cordgrass is amajor carbon source unlike any o<strong>the</strong>r in <strong>the</strong>se communitiesand likely modifies microbial and macr<strong>of</strong>aunal resource use.English cordgrass modifications have significantconsequences for community structure. We compared nativevascular plant and algal abundance in areas with and withoutcordgrass in all communities (Hacker and Dethier 2006). Inthis study, we found that in mudflats, cobble beaches, andhigh salinity marshes, cordgrass invasion caused an increasein native vascular plant cover and decline in algal cover. Byelevating sediments, increasing oxygen content, anddecreasing salinity, cordgrass clearly provides a moresuitable habitat for native vascular salt marsh plants but aless suitable habitat for algae, which require greater tidalinundation to avoid desiccation. In low salinity marshes,native vascular plants declined precipitously, presumably aconsequence <strong>of</strong> cordgrass competitive dominance. We alsocompared marine invertebrates in mud and cobble sedimentsversus adjacent cordgrass patches. Cobble areas withoutcordgrass had oligochaetes, bivalves, and a variety <strong>of</strong>crustaceans; cobble with cordgrass had less infauna because<strong>of</strong> dense root mat formation. Uninvaded mudflats <strong>of</strong>ten hadabundant clams and a variety <strong>of</strong> polychaete worms, whilethose with cordgrass had fewer infauna but more epifaunalcrustaceans such as amphipods (presumably because <strong>of</strong> <strong>the</strong>three–dimensional structure provided by <strong>the</strong> vegetation).Invertebrate studies (Zipperer 1996; O’Connell 2002) inareas invaded by Spartina alterniflora in Willapa Bay, WA,similarly found that certain taxa are excluded by cordgrass(esp. polychaetes and bivalves) while o<strong>the</strong>rs are increased(esp. dipteran larvae and spiders). Although we have notquantitatively measured possible changes in epifaunalcommunities, we have observed a general increase in highermarsh epifauna such as grasshoppers, spiders, snails, mice,and marsh wrens within cordgrass meadows and a decline inshorebirds, some species <strong>of</strong> snails, mussels, and oysters.Cordgrass removal and consequences for post-removalrestorationRemoval <strong>of</strong> English cordgrass, involving mowing andherbicide applications, began in 1997 and has caused amodest 10-20% decline as <strong>of</strong> 2002 (Hacker et al. 2001;Dethier and Hacker 2004). Although local eradication hasoccurred at some sites with minimal treatment repetition,most sites have required repeated removal spanning multipleyears to achieve eradication. To investigate <strong>the</strong> factorshindering removal success, we conducted a multiple sitestudy that linked removal data with ecological factors andremoval methodologies. We found that removal successdepended on removal regime and community type (Dethierand Hacker 2004). The bulk <strong>of</strong> cordgrass decline was due toconsistent, multi-year removal (once per year) in certaincommunities. For example, cobble beaches, high salinitymarshes, and mudflats showed <strong>the</strong> most promising responseto consistent removal even when years <strong>of</strong> removal effortwere similar to low salinity marshes. When removal wasintermittent ( 2 years missed), low salinity marshes andmudflats were <strong>the</strong> least responsive to removal.The pattern <strong>of</strong> regrowth with removal regime points to<strong>the</strong> resiliency <strong>of</strong> cordgrass. If photosyn<strong>the</strong>sis is notcontinually interrupted to slowly kill <strong>the</strong> plant, removalsuccess is compromised due to its ability to regrow. In a set<strong>of</strong> manipulative experiments, we found that biomass gainsunder intermittent removal greatly outweigh losses accruedunder consistent removal both because <strong>of</strong> <strong>the</strong> highlyproductive nature <strong>of</strong> <strong>the</strong> species, <strong>the</strong> benefits <strong>of</strong> competitiverelease, and <strong>the</strong> habitat modifications that facilitate fur<strong>the</strong>rgrowth (Reeder and Hacker 2004).- 213 -