Gerd Mayer

Autofluorescence imaging, an excellent tool for comparative morphology

Here we present a set of methods for documenting (exo‐)morphology by applying autofluorescence imaging. For arthropods, but also for other taxa, autofluorescence imaging combined with composite imaging is a fast documentation method with high‐resolution capacities. Compared to conventional micro‐ and macrophotography, the illumination is much more homogenous, and structures are often better contrasted. Applying different wavelengths to the same object can additionally be used to enhance distinct structures. Autofluorescence imaging can be applied to dried and embedded specimens, but also directly on specimens within their storage liquid. This has an enormous potential for the documentation of rare specimens and especially type specimens without the need of preparation. Also for various fossils, autofluorescence can be used to enhance the contrast between the fossil and the matrix significantly, making even smallest details visible. ‘Life‐colour’ fluorescence especially is identified as a technique with great potential. It provides additional information for which otherwise more complex methods would have to be applied. The complete range of differences and variations between fluorescence macrophotography and different types of fluorescence microscopy techniques are here explored and evaluated in detail. Also future improvements are suggested. In summary, autofluorescence imaging is a powerful, easy and fast‐to‐apply tool for morphological studies.

MOUTHPARTS OF THE PONTO-CASPIAN INVADER DIKEROGAMMARUS VILLOSUS (AMPHIPODA: PONTOGAMMARIDAE)

Abstract The pontogammarid amphipod Dikerogammarus villosus, originally a Ponto-Caspian faunal element, has, in the recent 15-20 years, successfully invaded various aquatic systems in Europe including Lake Constance. In these rivers and lakes it had and still has severe ecological impact on native macro-invertebrates, often eliminating the native and earlier established gammaridean species. In order to test the hypothesis that the mode of food acquisition of D. villosus is of significance for this phenomenon, we focused on the mouthparts of D. villosus, i.e., mandibles, the two pairs of maxillae and the maxillipeds using SEM. Contrary to expectations, provoked by field and laboratory observations, the results of this study show that the mouthparts of D. villosus are not highly specialized just for carnivory and predation. Indeed, the stout mandibles, with their well-developed incisors enable to kill even prey with robust integument, but other modes of feeding are possible. On the maxillulae, maxillae, and maxillipeds we found setae that can be used, together with the gnathopods and the antennae, for filtering suspended algae and other small particles from the respiration current. The same structures are involved in collecting detritus. In contrast, D. villosus does not possess any specific tools for scraping periphyton from the substrate. Feeding on macrophytes may be possibly but not very effective because the surfaces of the molars are not well suited for grinding such plant material. It is shown that D. villosus is neither a shredder, as traditionally predicated for most gammarideans, nor is it a specialized carnivore, as predation experiments proposed, but rather unspecialized. Its ability to be carnivorous and to use a wide spectrum of other food may be an important reason for the success of this invader, being an advantage compared to mainly herbivorous gammarideans, which have been eliminated in many places by D. villosus.

Towards autonomous vision self-calibration for soccer robots

The ability to autonomously adapt to variations in the environmental conditions is a very useful feature for mobile robots. Of particular interest in robotic soccer are self-calibrating vision systems that automatically adapt to local lighting conditions. The paper presents a method for autonomous calibration of a color classificator used for color blob-based image segmentation and landmark and object recognition. The experimental results demonstrate significantly improved robustness of visual processing.

Influence of substrate preference and complexity on co-existence of two non-native gammarideans (Crustacea: Amphipoda)

Substrate choice, swimming activity and risk to predation by burbot (Lota lota) of the well established Gammarus roeselii and the invader Dikerogammarus villosus were studied in mixed and single-species aquarium experiments. We used stones, gravel and aquatic weeds (Elodea, Chara) as substrates. We hypothesized that both species have different substrate preferences and that substrate affects the predation risk. We also assumed that presence of D. villosus influences substrate preference and predation risk of G. roeselii since the invader is known to affect the behavior of other gammarids. Adults of D. villosus in single species experiments and juveniles in mixed and single species experiments were evenly distributed over the different substrates but adults in mixed species experiments were more likely to prefer stone substrate. In contrast, adults and juveniles of G. roeselii clearly preferred aquatic weeds independent of the presence/absence of the invader. Both species preferred substrates with fissured surface over substrates with smooth surface. Gammarus roeselii was observed swimming more often than D. villosus in the open water but its swimming activity was lower when its preferred substrate was present compared with its swimming activity if non-preferred substrates were present. Predation rate of burbot on D. villosus was comparatively low and independent of the substrate. Burbot consumed many more G. roeselii than D. villosus, both in mixed and single species experiments. But when the preferred substrate of G. roeselii (weeds) was used in the experiments, predation rate of burbot on G. roeselii was somewhat lower than that when non-preferred substrates were present. The results of the experiments support our hypothesis that the gammarids studied here have different substrate preferences and that presence of the preferred substrate can affect predation risk. However, there is no evidence that presence of D. villosus affected substrate choice or predation risk in G. roeselii. We consider that differences in use of spatial niches permit co-existence of G. roeselii and D. villosus in the wild when substrates are diverse. The fact that G. roeselii than D. villosus is more often observed swimming in the open water may explain its higher risk of being captured by fish.

Playing Robot Soccer under Natural Light: A Case Study

The recent debate in the RoboCup middle-size community about natural light conditions shows that a more in-depth analysis of the problems incurred by this is necessary in order to draft out a focused and realistic roadmap for research. Based on real-world images taken under varying lighting conditions, we performed descriptive and statistical analysis of the effects on color-based vision routines. The results show that pure color-based image processing is not likely to perform well under varying lighting conditions, even if the vision system is calibrated on a per-game base. We conclude that color-based vision has to be combined with other methods and algorithms in order to work robustly in more difficult environments with varying illumination.

Imaging and Documenting Gammarideans

We give an overview of available techniques for imaging and documenting applied to gammarideans and discuss their advantages and disadvantages. Although recent techniques, such as confocal laser scanning microscopy (cLSM), focused ion beam scanning electron microscopy (FIB SEM), or computed microtomography (μCT), provide new possibilities to detect and document structures, these high-tech devices are expensive, and access to them is often limited. Alternatively, there are many possibilities to enhance the capabilities of established techniques such as macrophotography and light microscopy. We discuss improvements of the illumination with polarized light and the possibilities of utilizing the autofluorescence of animals such as the gammarideans. In addition, we present software-based enhancing tools such as image fusion and image stitching.

Mouthpart Morphology of Gammarus roeselii Compared to a Successful Invader, Dikerogammarus villosus (Amphipoda)

Abstract Until about 20 years ago, the well-established gammarid amphipod Gammarus roeselii, originating from the Balkan Peninsula, was widespread and often the dominant amphipod species in rivers, canals, and big lakes of Germany and adjacent countries. Since 1989 Dikerogammarus villosus, a Ponto-Caspian pontogammarid, has successfully invaded the aquatic systems of mid and western Europe reducing and even eliminating native and earlier established gammarideans including G. roeselii. Field observations and laboratory experiments indicate that these two gammaridean species have different feeding habits. In our SEM study we wanted to test the hypothesis that different modes of food acquisition are expressed in differences in the morphology of the mouthparts and other structures involved in feeding, such as the antennae and gnathopods, possibly are indicative of the success of D. villosus over G. roeselii. Although both gammarideans have tools for omnivorous feeding, comparisons indeed revealed obvious differences in appendages involved in food acquisition. In G. roeselii, the molar surfaces of the mandibles are structured like rasps, suggesting they are well suited for grinding firm plant material, whereas those of D. villosus are only weakly structured. In G. roeselii, the basipodal endites of the maxillulae are armed with chisel-like cuspidate setae, and the maxillipeds bear flattened cuspidate setae, possibly a specialization of these mouthparts for scraping off adherent food from the substrate. We did not find such modifications on the maxillulae and maxillipeds of D. villosus. Compared to D. villosus, the setation of the antennae and the gnathopods in G. roeselii is sparse and short. Therefore, filter feeding, one of the feeding modes of D. villosus, seems to play a minor role in the nutrition of G. roeselii.

Visual robot detection in robocup using neural networks

Robot recognition is a very important point for further improvements in game-play in RoboCup middle size league. In this paper we present a neural recognition method we developed to find robots using different visual information. Two algorithms are introduced to detect possible robot areas in an image and a subsequent recognition method with two combined multi-layer perceptrons is used to classify this areas regarding different features. The presented results indicate a very good overall performance of this approach.

Improving Vision-Based Self-localization

After removing the walls around the field, vision-based localization has become an even more interesting approach for robotic soccer. The paper discusses how removal of the wall affects the localization task in RoboCup, both for vision-based and non-visual approaches, and argues that vision-based Monte Carlo localization based on landmark features seems to cope well with the changed field setup. An innovative approach for landmark feature detection for vision-based Monte Carlo Localization is presented. Experimental results indicate that the approach is robust and reliable.

Hierarchical behavior organization

In most behavior-based approaches, implementing a broad set of different behavioral skills and coordinating them to achieve coherent complex behavior is an error-prone and very tedious task. Concepts for organizing reactive behavior in a hierarchical manner are rarely found in behavior-based approaches, and there is no widely accepted approach for creating such behavior hierarchies. Most applications of behavior-based concepts use only few behaviors and do not seem to scale well. Reuse of behaviors for different application scenarios or even on different robots is very rare, and the integration of behavior-based approaches with planning is unsolved. This paper discusses the design, implementation, and performance of a behavior framework that addresses some of these issues within the context of behavior-based and hybrid robot control architectures. The approach presents a step towards more systematic software engineering of behavior-based robot systems.