KOPS - Das Institutionelle Repositorium der Universität Konstanz

Impacts of invasive amphipods on the local benthic fauna and leaf litter decomposition

Impacts of invasive amphipods on the local benthic fauna and leaf litter decomposition


Dateien zu dieser Ressource

Prüfsumme: MD5:e7c58f0915de87a64d79026daca39418

HESSELSCHWERDT, John, 2009. Impacts of invasive amphipods on the local benthic fauna and leaf litter decomposition

@phdthesis{Hesselschwerdt2009Impac-7984, title={Impacts of invasive amphipods on the local benthic fauna and leaf litter decomposition}, year={2009}, author={Hesselschwerdt, John}, address={Konstanz}, school={Universität Konstanz} }

Einfluss invasiver Amphipoden auf die einheimische benthische Lebensgemeinschaft und den Abbau von Falllaub Hesselschwerdt, John application/pdf Impacts of invasive amphipods on the local benthic fauna and leaf litter decomposition 2011-03-24T17:39:03Z deposit-license In the last decades, the Ponto-Caspian amphipod Dikerogammarus villosus has invaded most European waterways while displacing native benthic invertebrates. D. villosus is rather described as a predator than a shredder and even preys directly on other amphipods. In 2002, D. villosus was introduced to Lake Constance and was expected to quickly outcompete the smaller shredder Gammarus roeselii, which has dominated the amphipod fauna so far. As expected, the density of G. roeselii in Lake Constance declined very quickly but has now remained stable for seven years. During the winter G. roeselii even dominates the stony littoral. In this thesis, I have been able to clarify the mechanisms behind this coexistence in Lake Constance.<br /><br />Enclosure experiments in the littoral revealed different substrate preferences for both amphipod species. D. villosus strongly preferred stones, especially when covered with the also invasive mussel Dreissena polymorpha. In contrast, G. roeselii did not show any preferences at all. In the field, D. villosus was found distributed mostly on stones and in lower densities within the macroalgae Chara, according to its preference. G. roeselii was only found in Chara and on leaf litter and was nearly completely absent on stones, contrary to its preference. Thus, the stronger D. villosus displaced G. roeselii from stones but G. roeselii successfully occupied the substrates less preferred by D. villosus.<br /><br />Aside from habitat segregation, G. roeselii indicated active enemy avoiding behavior by reacting to olfactory traces. In choice experiments within a Y-maze, G. roeselii as well as D. villosus revealed the capability to perceive the kairomone of the other amphipod species and the kairomone of another invasive crustacean, the American crayfish Orconectes limosus. Both amphipods displayed an avoiding reaction against the other species but did not react to the kairomone of conspecifics. Thus, both amphipods seem to avoid other crustaceans. This is a reasonable reaction for G. roeselii but is counterproductive for D. villosus since G. roeselii represents prey. D. villosus did not change its reaction after being trained to become familiar with G. roeselii as prey for seven days. Therefore it is a reasonable assumption that D. villosus has probably not adapted to local amphipods in Central Europe. Repetition of the experiment in several years could clarify whether D. villosus is able to adapt to the new situation.<br /><br />The mainly discussed mechanism for the displacement of other amphipods by D. villosus is direct predation. Since D. villosus was found together with G. roeselii in Chara, I assumed a substrate specific predation and conducted predation experiments on all dominant substrates in Lake Constance: stones (with and without D. polymorpha), sand, Potamogeton perfoliatus, Chara, and an artificial substrate mimicking the structure of Chara: fish net). In most substrates G. roeselii suffered strongly from predation by D. villosus. It was only inside Chara and the corresponding artificial substrate that I was unable to detect any predation on G. roeselii at all. D. villosus seems to be hindered within the fine structure of these substrates. D. villosus is bigger and has more protruding pereiopods than G. roeselii. Thus, G. roeselii can escape the predation inside Chara. G. roeselii now benefits from the strong efforts that were made to reoligotrophicate Lake Constance after a phase of eutrophication in the 1980s. During this period, numbers of Chara were strongly reduced, but began to increase again as phosphate loads declined. But Chara is still lacking in most winters and leaves G. roeselii without shelter over winter.<br /><br />In aquaria experiments, D. villosus shifted its predatory behavior when the water temperature dropped below 6°C reducing feeding on G. roeselii; below 4°C it stopped preying on the latter at all. This result is supported by field data around a warm water outlet of a thermal bath on Lake Constance: in the surrounding area (4°C) both species occurred in equal densities, while at the outlet (mean 5.5°C with higher peaks) D. villosus almost completely displaced G. roeselii. In Lake Constance, Chara and low temperatures provide refuge for G. roeselii over the whole year and secure coexistence with the invasive D. villosus. But this situation could be at risk if global warming elevates the water temperature over winter by 2°C, as predicted by 2050.<br /><br />In literature, D. villosus is described as predator rather than shredder. I tested the feeding rates of both amphipods on several common leaf species. D. villosus displayed much lower feeding rates on leaves than G. roeselii and did not feed on beech at all. Therefore, reduced densities of the previously dominant shredder G. roeselii could affect leaf litter decomposition in Lake Constance.<br /><br />To gain deeper insight into leaf litter decomposition within the lake, I conducted a decomposition experiment in winter 2005/06. Litter bags filled with leaves of beech (Fagus sylvatica) and black alder (Alnus glutinosa) were placed in the littoral of Lake Constance at depths of 0.5, 2, 5, and 10 m and on the shoreline for up to 107 days. Naturally shed leaves were collected with nets under trees one month prior to the experiment and never touched the ground. In the lake, beech decomposed much slower than alder and than beech in a previous study in Lake Constance at higher temperatures. The invertebrate community in the bags was highly affected by the water depth but was abundant from the first day in the lake. The dominating shredders were the amphipods G. roeselii and D. villosus and the isopod Asellus aquaticus. At most depths, G. roeselii significantly outnumbered D. villosus, which was able to withstand very low temperatures of 3.5°C for long periods at a depth of 0.5 m. D. villosus seemed to have no strong impact on the litter decomposition over winter in Lake Constance. Decomposition rate and all other leaf parameters were not affected by water depth. The core processes of leaf decomposition are conducted by microorganisms. In my experiment, the fungal biomass displayed an interesting pattern on alder leaves: directly after exposure in the lake, biomass increased rapidly by day six, even almost reaching the maximum biomass. It subsequently declined back to the start value and then slowly rose again by the end of the experiment. This pattern could be observed in all depths, even on the swampy shore. I interpret this pattern as a reaction to preattached fungi on the leaves while still on the tree. After a short outbreak of the terrestrial fungi, they degrade after being submersed for too long; after a certain amount of time, a stable aquatic fungal community is established. This hypothesis is strongly supported by DGGE analyses with sequencing of several strains of this samples.<br />In these analyses, several terrestrial fungi were found during the initial days of incubation (data by Sven Boekhoff and not included in the thesis). The bacterial biomass remained much lower than fungal biomass over the whole period of exposure. Antagonism between fungi and bacteria has been reported in literature several times. Bacteria can hinder fungal growth unless the fungi inoculate the leaves before the bacteria. In this event, the bacterial community cannot establish itself very well. This corresponds to our findings of preattached terrestrial fungi and a smaller bacterial biomass. It is important to use naturally inoculated leaf material that has not been stored too long for experiments analyzing the natural development of the decomposition process. 2012-02-29T23:25:05Z eng Hesselschwerdt, John 2009

Dateiabrufe seit 01.10.2014 (Informationen über die Zugriffsstatistik)

Dissertation_Hesselschwerdt.pdf 155

Das Dokument erscheint in:

KOPS Suche


Mein Benutzerkonto