Der biologische Aufschluss lignocellulosehaltiger Biomasse mit Hilfe von Mikroorganismen oder ihrer Enzyme ist im Vergleich zu bekannten physikochemischen Verfahren umwelt- und ressourcenschonend. Der Einsatz geeigneter bakterieller oder pilzlicher Hydrolasen und Oxidoreduktasen in isolierten Form bedarf jedoch, aufgrund der noch zu geringen katalytischen Effizienzen und der nach wie vor zu hohen Herstellungskosten der Enzyme, weiterer Optimierung. Vor diesem Hinter-grund besteht, neben dem Ansatz einer gentechnischen Verbesserung der En-zym-Eigenschaften mittels protein oder metabolic engineering, die Möglichkeit einer prozesstechnischen Optimierung der Enzym-Präparate und ihrer Einsatzbe-dingungen. Dem letzteren Ansatz widmete sich die vorliegende Arbeit, in der ein bisher kommerziell nicht erhältliches Glycosidase-Gemisch aus dem ascomycetalen Schimmelpilz Penicillium janthinellum, der für seine hohen β-Glycosidase-Aktivitäten bekannt ist, im Zusammenhang mit dem enzymatischen Aufschluss faserreicher Substrate (Lignocellulosen) untersucht wurde. Ein Schwerpunkt lag dabei auf der kombinierten Anwendung des Glycosidase-Präparats mit zwei pilzlichen Peroxidasen (Mangan-Peroxidase, MnP und Dye-decolorizing-Peroxidase, DyP). Darüber hinaus wurden vergleichende Untersuchungen zu biologischen Aufschlussverfahren unter Einsatz pilzlicher Glycosidasen und/oder mikrobieller Vorkulturen durchgeführt. Die untersuchten Lignocellulose-Substrate (Hölzer, strohähnliche Materialien) stammten aus der gemäßigten und tropischen Klimazone (Europa bzw. Kambodscha), und wurden in den Experimenten in zerkleinerter Form, allerdings ohne weitere Vorbehandlung, eingesetzt. Die Konzentration niedermolekularer Zucker (insbesondere Monosaccharide), die während des enzymatischen Aufschlusses aus den Substraten freigesetzt wurden sowie der Biogasertrag, der mittels anaerober Fermentation aus den enzymatisch und/oder mikrobiell vorbehandelten Substraten erzielt wurde, dienten zur Beurteilung der Effektivität der jeweiligen Vorbehandlung. Außerdem wurde die in den Experimenten verwendete mikrobielle Vorkultur soweit molekularbiologisch untersucht, dass die Bakterienart identifiziert werden konnte, die maßgeblich am Aufschluss der lignocellulosehaltigen Biomasse beteiligt war. Die enzymatische Umsetzung der ausgewählten lignocellulosehaltigen Substrate mit Hilfe des Glycosidase-Gemisches aus P. janthinellum verlief erfolgreich und ist vergleichbar mit Ergebnissen, die laut Literatur unter Zuhilfenahme der effektivsten industriellen Cellulase-Präparate erzielt worden sind. Es wurden vorrangig Glucose und Xylose aus den verschiedenen Zellwand-Polysacchariden freigesetzt, wobei die Umsetzung von Cellulose und Hemicellulosen im Holz tropischer Laubbäume effizienter verlief als im Holz europäischer Laubbäume. Der Gehalt an Lignin und organischen Extraktiven beeinflusste – abgesehen von einigen artenspezifischen Inhibitoren – nur geringfügig den enzymatischen Aufschluss der Polysaccharid-Komponenten. Die Vorbehandlung mit dem Glycosidase-Präparat aus P. janthinellum führte zu einer Verbesserung der Biogasbildung und zum Ausbleiben der für faserreiche Substrate typischen Lag-Phase während der ers-ten Tage der anaeroben Vergärung der Lignocellulose aus Triticum sp. (Weizen-stroh) und Pinus sylvestris (Kiefernspäne). Dabei erhöhte sich der finale Biogasertrag innerhalb des Untersuchungszeitraums signifikant. Die genannten positiven Effekte einer enzymatischen Vorbehandlung könnten sich in kontinuierlich betriebenen großtechnischen Biogasanlagen als nützlich erweisen: Zum einen ließen sich die Gaserträge deutlich erhöhen und zum anderen könnte die erforderliche Verweilzeit des Substrates im Bioreaktor (Fermenter, Faulturm) und somit das benötigte Anlagenvolumen reduziert werden. Eine vorausgehende Oxidation des im Substrat enthaltenen Lignins mit Hilfe der MnP erwies sich in der nachfolgenden Behandlung mit Glycosidasen als förderlich hinsichtlich der Freisetzung von Zuckern aus dem Holz von Fagus sylvatica (Rotbuche). Verglichen mit der häufig verwendeten Malonsäure war die Citronensäure, ein pilzlicher Metabolit des Intermediär-Stoffwechsels (Zitronensäurezyklus), ein wirksamerer Mangan-Chelator für diese Voroxidation mittels MnP. Dies hing möglicherweise mit der höheren chemischen Reaktivität der Citronensäure zu-sammen, was eine verstärkte Bildung chemischer Radikale zur Folge hatte. Eine enzymatische Vorbehandlung mittels DyP und dem Glycosidase-Gemisch in einer Reaktionskaskade wirkte sich ebenfalls positiv auf die Biogasbildung, in diesem Fall aus Bagasse von Saccharum officinarum (Zuckerrohr), aus. Dabei kam es wahrscheinlich auch zu einer partiellen Oxidation und Zerstörung des Lignins und damit zu einer Verbesserung der Zugänglichkeit der Zellwand-Polysaccharide. Im Ergebnis konnten Cellulose und Hemicellulosen in späteren Phasen der anaeroben Vergärung von den entsprechenden Mikroorganismen (Bakterien, Archaeen) besser verwertet werden. Der Voraufschluss mit Glycosidasen führte hingegen in der initialen Phase der anaeroben Vergärung zu positiven Effekten bezüglich der Biogasbildung, indem die bereitgestellten Einfachzucker (z.B. Glucose, Xylose) rasch in Methan umgewandelt wurden. Beim Vergleich verschiedener biologischer Aufschlussverfahren erwies sich eine kombinierte Vorbehandlung des Substrates („Stroh“ von Miscanthus × giganteus), bestehend aus einer Vorhydrolyse durch das Glycosidase-Gemisch und einer Vorfermentation mit einer Mischkultur gärender Mikroorganismen, als der effektivste Weg. Durch die kombinierte biologische Vorbehandlung konnte ein ähnlich hoher Methanertrag wie für Maissilage (das derzeit optimale Substrat in Biogasanlagen) erreicht werden. In der entsprechenden mikrobiellen Vorkultur wurde ein Bacillus-Vertreter aus dem so genannten Bacillus-subtilis-Artkomplex (Bacillus subtilis species-complex) mittels klassischer mikrobiologischer und molekularbiologischer Analysen als möglicher „abbaurelevanter Organismus“ identifiziert. / The biological disintegration of lignocellulosic biomass by microorganisms and their enzymes is – in comparison to established physical and chemical approaches –environmentally friendly and sustainable. The broad use of isolated bacterial or fungal hydrolases and oxidoreductases requires, however, still substantial optimization because of too low catalytic performance and too high production costs for the enzymes. Against this background, there is the possibility, besides genetic improvement of enzyme properties by protein and metabolic engineering, to optimize the process performance of enzymes as well as the reaction conditions. The latter approach has been subject of the present dissertation, in the course of which a non-commercial preparation of glycosidases from the ascomycetous mold Penicillium janthinellum, which is well-known for its high β-glycosidase activities, was used for the enzymatic disintegration of fiber-rich substrates (lignocelluloses). Experimental work focused on the combined action of the glycosidase mixture with two fungal peroxidases (manganese peroxidase, MnP and dye-decolorizing peroxidase, DyP). Furthermore, comparing studies were carried out regarding enzymatic/biological lignocellulose disintegration by isolated fungal glycosidases and/or microbial precultures. Lignocellulose substrates studied (wood, straw-like materials) originated from temperate and tropic climate zones (Europe and Cambodia, respectively) and were used after chopping in all experiments without further pretreatment. The concentration of low-molecular mass sugars (in first place monosaccharides), which were being released from the substrates during enzyme action as well as the biogas yield that was achieved via fermentation of enzymatically or microbiologically pretreated samples, were taken into consideration to evaluate the efficacy of respective treatments. Moreover, the microbial preculture used in the above experiments was analyzed on the molecular level to an extent that it was possible to identify a bacterial key species that was involved in the degradation of lignocellulosic biomass. The enzymatic treatment of selected lignocellulosic substrates with the glycosidase mixture of P. janthinellum was successful and the results are – according to literature data – comparable to results reported for the best industrial cellulase preparations. In first place, glucose und xylose were released from different cell-wall polysaccharides, and the conversion of cellulose und hemicelluloses in the wood of tropical broad-leaved trees was more efficient than in wood of respective European trees. The content of lignin and organic extractives only slightly affected the enzymatic disintegration of polysaccharide components (apart from a few species-specific inhibitors). Substrate pretreatment with the glycosidase preparation of P. janthinellum resulted in an enhancement of biogas formation and in the disappearance of the lag-phase being characteristic for the conversion of fiber-rich substrates during the first days of anaerobic treatment of lignocelluloses from Triticum sp. (wheat straw) und Pinus sylvestris (wood shavings). In this context, the final biogas yields significantly increased in the course of the experiments. The observed positive effects of enzymatic pretreatment may be beneficially ap-plied in continuously working biogas plants. That way, on one hand, the gas yields could be considerably enhanced and on the other hand, the required retention time of the substrates in the bioreactor (fermenter, digestion tower) and hence the required reactor volume could be reduced. The preceding oxidation of substrate-bound lignin with MnP turned out to be beneficial for the subsequent glycosidase treatment with respect to the release of sugars from beech wood (Fagus sylvatica). In comparison to widely used malonic acid, citric acid – a ubiquitous fungal metabolite of the intermediary metabolism (tricarboxylic acid cycle) –proved to be the more effective manganese chelator for the pre-oxidation of lignin by MnP. Probably this corresponds to the higher chemical reactivity of citric acid, which entails a forced formation of chemical radicals. Enzymatic substrate pretreatment with DyP and the glycosidase mixture within a reaction cascade had also a positive effect on the formation of biogas, in this case, from bagasse of Saccharum officinarum (sugar cane). During the respective treatment, the lignin might partially be oxidized as well and thereby, the availability of cell-wall polysaccharides was improved for hydrolase action. As the result, microorganisms (bacteria, archaea) consumed cellulose and hemicelluloses more efficiently during later phases of anaerobic fermentation. On the other hand, glycosidase pretreatments had positive effects in the initial phase of fermentation, regarding biogas formation from the ‘made-available’ monosaccharides (e.g. glucose, xylose) that were immediately converted into methane. When comparing different biological methods to disintegrate lignocellulose, pre-hydrolysis with a glycosidase mixture combined with fermentative pretreatment proved to be the most effective option (demonstrated by the example of ‘straw’ from Miscanthus × giganteus). That way, a similarly high methane yield could be achieved as with maize silage (for the time being, the most suitable substrate used in biogas facilities). In the respective microbial preculture, a Bacillus species from the Bacillus subtilis species-complex was identified as a relevant potential degrader microbe by classic microbiological and molecular analyses.

No description available.

No description available.

No description available.

The West Indies have during recent years been experiencing large quantities of washed up Sargassum seaweed on beaches and coastlines causing many issues for local communities such as unpleasant odours, making beaches inaccessible and largely affecting tourism and fisheries. The issue has caused money loss for the communities both in cleaning costs and decreasing incomes from tourism and fishing. In an attempt to turn this destructive debris into a useful resource, this study aimed to determine the potential for Sargassum that has been anaerobically digested to be used as a liquid fertiliser on Grenada. Sargassum seaweed was collected from the Atlantic coast of Grenada and transported and fed to a biogas plant where the algae was anaerobically digested. The residues were collected and fed to Pak Choi crops in a cultivation experiment where the fertiliser was compared with other treatment methods in order to determine its efficiency as a fertiliser. Interviews of farmers on Grenada were also conducted during the project to examine the potential and usefulness of the biofertiliser on the local Grenada farms. It was found that the digestate product from the anaerobic digestion showed good potential as a biofertiliser, producing larger and healthier crops than the other methods of treatments, among them artificial fertiliser. A potential market for the product was also identified as all the interview subjects reported that artificial fertilisers can be hard or impossible to find on Grenada sometimes of the year, due to the island's high reliance on shipping containers for import, leading to a demand of alternative fertilisers. / Under de senaste åren har algblomningarna ökat runt om i Västindien och numera är flertalet stränder  fyllda av Sargassotång. Ett av länderna som har påverkats negativt är Grenada. Stora mängder Sargassum når östkusten av Grenada under större delen av året till skillnad från tidigare år då algerna enbart sköljdes in under vissa perioder och i mindre kvantiteter. Algmassorna leder till lokal övergödning i de drabbade områdena, vilket i sin tur leder till minskad syrekoncentration i vattnet och en stor påverkan på djur och växter i ekosystemet. I Grenada påverkar makroalgerna även inkomstkällan för fiskare då de inte kan arbeta när det är som mest alger eftersom båtarna inte kan ta sig förbi massorna i vattnet. Även lokalbefolkning och turister påverkas då nedbrytningen av algerna skapar en motbjudande lukt vilket leder till att människor flyttar från dessa områden och att turistnäringen minskar. Grenada har ett behov av att minska mängden alger längs stränderna och samtidigt finns en möjlighet att ta tillvara en lokal bioresurs. En möjlighet är att använda algmassan till produktion av biogas och biogödningsmedel eftersom nedbrytning av makroalgeralger både kan producera metangas samt att  återstående restbiomassa innehåller de viktigaste näringsämnen som växter behöver för att växa. Därför  har algerna använts i ett försök på True Blue Bay Boutique Resorts anläggning på Grenada för att producera biogas och gödsel med hjälp av en mindre biogasanläggning. Anläggningen matas med biomassa och därefter har en gas och en restprodukt producerats. Restprodukten har därefter används som gödningsmedel i ett odlingsförsök med Pak Choi för att fastlägga dess kvalitet som växtnäring. Plantor gödslades med restprodukter från biogasanläggningen, både när anläggningen matats med grisgödsel och hushållskompost i början samt när födan successivt övergick till alger. Andra plantor fick konstgödsel eller färska alger. Utöver det fanns det en grupp med plantor vilka inte tog del av någon form av gödsel för att fungera som en kontrollgrupp. Vid slutet av odlingsförsöket skördades Pak Choi grödorna och produktivitet och kvalitet på grödorna bedömdes baserat på olika kriterier. Resultaten visade att de växter som blivit gödslade med resterna från biogasanläggningen var störst och av bäst kvalitet. Därutöver har en undersökning genomförts för att få en större förståelse för hur tillgången på alger ser ut och för att få en bild över hur lantbrukares behov av gödningsmedel ser ut på Grenada. Lantbrukare har tillfrågats om tillgång och efterfrågan på olika sorters gödsel, hur de använts och om de har några erfarenheter från Sargassotång. Deras åsikter har varierat, men en gemensam nämnare är att alla är intresserade av lokalproducerade gödningsmedel. Konstgödsel är kostsamt och det kan vara svårt eller omöjligt att få tag på under vissa delar av året, då landet förlitar sig på import. Genom detta har en potentiell marknad för gödselmedel producerat med alger fastställts.

Denna rapport presenterar resultat och analys av prestandan i ett pilotförsök om temperaturstegsrötning (TPAD). Pilotförsöket genomfördes på uppdrag av Käppalaförbundet. Resultaten jämfördes med en nedskalad mesofil process som simulerade den nuvarande fullskaliga rötningsprocessen på Käppala  avloppsreningsverk. Syftet var att utvärdera om TPAD kunde erbjuda fördelar jämfört med det mesofila systemet. TPAD har tidigare visat stor potential i pilotstudier och till viss del i fullskaliga implementeringar på avloppsreningsverk runt om i världen. Då prestanda och beteende vid rötningsprocesser dock är starkt beroende av lokal slamkomposition och processparametrar, behövdes en skräddarsydd pilotstudie för att utvärdera TPADs applicerbarhet vid Käppala avloppsreningsverk specifikt. Sammanfattningsvis visade TPAD något bättre metanutbyte än det mesofila kontollförsöket (MAD), och VSD ökade markant. Dessa fördelar erhölls trots den lägre totala retentionstiden för TPAD. Kvävemineraliseringen ökade dock också märkbart vilket potentiellt kan medföra ökade kostnader associerade med rening av kväve i rejektvattnet för avloppsreningsverket. Avvattningsprov på labbskala visade ett omfattande utsläpp av TSS i rejektvattnet efter det termofila rötningssteget. Detta förbättrades dock avsevärt (om än inte i samma utsträckning som MAD) efter det mesofila skedet. Mycket preliminära resultat indikerade att TPAD hade en utmärkt inneboende hygieniseringsförmåga tack vare det termofila skedet, och resulterade i ett slutgiltigt rötslam som uppfyllde hygieniseringsriktlinjer enligt Revaq. TPAD verkade uppvisa hög robusthet, utan någon uppenbar syrakollaps trots hög belastning i det termofila steget. Framtida stresstester föreslås för att tvinga fram en termofil syrakollaps, vilket skulle kunna ge en syra/gas-fasad TPAD, med potentiellt ytterligare ökad prestanda enligt mycket av den befintliga litteraturen. / This report analyzes the performance of a pilot scale temperature phased anaerobic digestion process (TPAD) undertaken on commission from the Käppalaförbundet wastewater treatment plant. Results from the newly initiated TPAD pilot were compared to those of a scaled down mesophilic process simulating the current full scale digestion used at Käppala, to evaluate whether TPAD could provide benefits over the mesophilic system. TPAD had previously showed great promise at pilot and full scale at other plants around the world, but given that anaerobic digestion performance and behavior are highly dependent on local sludge composition and process parameters, a bespoke pilot was needed to evaluate TPAD at Käppala WWTP specifically. In summary, the TPAD exhibited slightly better methane yields than the mesophilic control, and showed better removal of volatile solids. Such benefits were seen despite the lower overall retention time of the TPAD. Nitrogen mineralisation however also increased, potentially imposing increased costs associated with sludge liquor nitrogen purification. Dewaterability tests showed the thermophilic stage of TPAD releasing large amounts of problematic colloidal material, which however was reduced by the subsequent mesophilic stage. Preliminary results indicated the TPAD had an excellent inherent hygienization ability owing to the thermophilic stage, producing a final digestate which fulfilled Revaq hygienization guidelines. The TPAD also seemed to exhibit great robustness, with no acid collapse in the thermophilic stage apparent despite high loads and short retention times. Future stress tests are proposed to test an acid-gas phased TPAD, with potentially further increased performance as per much of the existing literature.

The prevalence of antibiotics on farms for therapeutic and prophylactic use in animals can cause negative effects on biomethane production during anaerobic digestion. Previous literature has found decreased biomethane production rates from a variety of antibiotics, but biogas inhibition differs between studies of continuous and batch reactors and the type of antibiotic studied. Cephalosporin drugs are the most common antibiotic class used to treat mastitis in dairy cows and can retain most of their bioactivity after excretion. Ceftiofur is a commonly used cephalosporin drug but no previous study investigating the effect of Ceftiofur on biomethane during continuous anaerobic digestion has been performed. The aim of this study was to examine the effect on biomethane production when manure from cows treated with Ceftiofur was anaerobically digested. Laboratory sized anaerobic digesters (AD) were run at thermophilic (55°C) temperatures and a 10 day hydraulic retention time. Manure from cows treated with Ceftiofur were fed to the antibiotic treatment reactors for 50 days. The reactor performance was measured by i) biomethane production, ii) waste stabilization in terms of solids and chemical oxygen demand, iii) change in mass of Ceftiofur and iv) change in concentration of antibiotic resistant genes, specifically cfx(A), mef(A), and tet(Q). There was statistically significant decrease in cumulative gas production due to the addition of Ceftiofur into the reactors, but no significant difference between treatments in waste stabilization in terms of percent volatile solids (VS) and total chemical oxygen demand (TCOD) reduction. Anaerobic digestion decreased the amount of Ceftiofur in manure, and the amount of Ceftiofur in the reactors reduced over the time of the experiment. Change in antibiotic resistant genes (ARGs) were gene dependent over time. Concentrations of tet(Q) reduced significantly between feed and effluent of both treatments, and cfx(A) reduced significantly for the control treatment but not the Ceftiofur treatment. Concentrations of mef(A) increased over time in both treatments. Overall, the addition of Ceftiofur in continuously operated anaerobic digesters negatively affected biomethane production, a value-added product responsible for on-farm renewable energy. However, anaerobic digestion does decrease the mass of Ceftiofur within manure, thereby reducing the environmental loading from run-off from farms. / Master of Science / Anaerobic digestion is a biological treatment technology used on farms to treat manure. It can be used to reduce potential environmental damage from contaminants and manure, homogenize manure for fertilizer, and produce methane gas for renewable energy. An emerging challenge in manure management is the presence of antibiotics such as ceftiofur used in animal production to prevent and treat illnesses. When antibiotics are used on livestock, they are excreted from the animal in manure. When the manure is added to the digester, the antibiotic molecules within the manure can kill the bacteria responsible for manure homogenization and gas production. Ceftiofur is a type of cephalosporin antibiotic used to treat dairy cows for mastitis, a bacterial infection of the udder. When the cows are treated with Ceftiofur, it can remain in the excreted manure and enter the digester. The use of antibiotics on farms is also leading to a global phenomenon known as antibiotic resistance. The bacteria that are exposed to antibiotics can develop mutations to become immune to the antibiotic, and can spread the mutations through antibiotic resistant genes (ARGs). ARGs can spread to bacteria which have never been exposed to antibiotics, making them resistant. This causes a significant concern in regards to disease treatment across the world as the efficacy of antibiotics is threatened. Understanding how ARGs move and how they can be eliminated is crucial to preventing global antibiotic resistance. The purpose of this study was to assess the effect of anaerobic digestion on Ceftiofur and ARGs. Four continuous lab-scale anaerobic digesters, two using control manure and two using manure from cows treated with Ceftiofur, were run at 55˚C for a period of 50 days. Over time, the reactor with manure from cows treated with the Ceftiofur antibiotic produced less gas as compared to the control digesters. The amount of Ceftiofur within the digesters decreased over time, demonstrating anaerobic digestion’s ability to degrade the antibiotic molecule. The effect of anaerobic digestion on the ARG concentration was gene specific. The concentration of the tet(Q) gene, a gene responsible for resistance against the very common antibiotic tetracycline, was reduced by anaerobic digestion. These results demonstrate that anaerobic digestion is a technology which can reduce the environmental impact of manure from Ceftiofur-treated cows. This shows that manure treatment can be a first step in combating antibiotic resistance across the globe.

This research was designed to investigate the effect of cations on activated sludge characteristics and also to determine their influence on digestion performance. For this purpose, cations in solution and in floc were evaluated along with various activated sludge characteristics and the collected waste activated sludge underwent both anaerobic and aerobic digestion. It was found that large amounts of biopolymer (protein + polysaccharide) remained in the effluent of WWTP that received high influent sodium but had low iron and aluminum in floc. However, sludges from plants with high sodium and high iron and aluminum dewatered well and produced high quality effluents, suggesting that iron and aluminum have significant positive effects on floc properties. Following anaerobic digestion, a significant increase in solution protein occurred and correlations between solution protein, ammonium production, percentile volatile solids reduction and iron in floc were obtained. These data indicate that iron-linked protein is released to solution when iron is reduced and its degradation is responsible for volatile solids reduction in anaerobic digestion. In aerobic digestion, polysaccharide in solution increased along with calcium, magnesium and inorganic nitrogen. This implies that divalent cation-bound biopolymer might be the primary organic fraction that is degraded under aerobic digestion. Combined (anaerobic/aerobic) digestion was performed and produced further volatile solids destruction with discrete cation and biopolymer response during each phase of digestion. These results support the theory that two types of organic matter with different cation bindings are present in floc and each type is degraded under different digestion processes. / Master of Science

De nos jours, les cultures mixtes sont considérées comme une sérieuse alternative aux cultures pures pour les procédés de biotechnologie. En effet, les cultures mixtes peuvent fonctionner en réacteur continu, dans des conditions non-stériles et traiter une grande variété de substrats organiques. La principale restriction de l'utilisation de ces bioprocédés en cultures mixtes réside dans leur instabilité liée à la présence de voies métaboliques non désirées résultant d'interactions microbiennes complexes. Notamment, le rôle des bactéries de faible abondance reste à être élucidé. Ce travail a donc consisté, dans un premier temps à déterminer le rôle des bactéries minoritaires dans la production d'hydrogène par voie fermentaire en utilisant un chémostat alimenté en continu avec un milieu à base de glucose. Sept inocula ont été cultivés dans les mêmes conditions opératoires. De façon remarquable, Clostridium pasteurianum a été retrouvé comme espèce dominante de l'écosystème six fois sur sept. Seules la nature et la diversité des espèces minoritaires variaient d'un écosystème à l'autre. Ainsi, il a été montré que la structure des communautés microbiennes a une influence significative sur la production de bio-hydrogène. Au sein de ces communautés, les bactéries en proportion minoritaires jouent un rôle clé en orientant le métabolisme globale de l'écosystème. La deuxième étape de ce travail a consisté à utiliser certaines de ces espèces minoritaires comme Ingénieurs Ecologiques des Ecosystèmes microbiens (IEEM). Pour cela, la structure d'une communauté microbienne productrice d'hydrogène a été modifiée artificiellement en introduisant des souches bactériennes exogènes aux fonctions redondantes et/ou complémentaires des souches indigènes. Les résultats en réacteur batch ont montré que les performances de production d'hydrogène pouvaient être améliorées jusqu'à un facteur 3,5 par l'ajout de certaines souches. Dans l'ensemble, les résultats obtenus ne peuvent être expliqués par de simples interactions trophiques et suggèrent la présence de mécanismes d'interactions de coopération entre microorganismes. De plus, sous des conditions opératoires plus favorables (inoculum, milieu), l'insertion de certaines espèces minoritaires a permis plutôt de stabiliser le métabolisme de l'écosystème microbien sans pour autant en affecter favorablement la production d'hydrogène. Dans tous les cas, les interactions compétitives n'ont pas été favorables à la production d'hydrogène. Enfin, des essais en réacteur continu ont montré que le mode d'implantation des souches peut être un facteur primordial pour l'utilisation d'IEEM. En conclusion, ce travail a montré la potentialité d'utiliser des bactéries exogènes, en proportions minoritaires, comme facteurs biotiques pour stabiliser et/ou orienter les métabolismes microbiens vers des fonctions d'intérêt au sein des cultures mixtes microbiennes. / Nowadays mixed cultures are considered as a serious alternative to pure cultures in biotechnological processes. Mixed cultures can be operated continuously, under unsterile conditions and from various organic substrates. One of the most constraints remains the chronic instability of the mixed culture processes due to the presence of unwanted metabolic pathways resulting from complex microbial interactions. More particularly the role of bacteria in low abundance remains to be elucidated. Therefore this work consisted initially to determine the contribution of sub-dominant bacteria to fermentative hydrogen production using a chemostat continuously fed with a glucose-based medium. Seven inocula were grown under the same operating conditions. Interestingly, Clostridium pasteurianum was found as dominant in six assays on seven at steady state. Only the minority bacterial population differed with regards to their identity and diversity. Acting as true keystone species, these minority bacteria impacted substantially the metabolic network of the overall ecosystem despite their low abundance. In a second step, this work consisted in using some of these minority species as Ecological Engineers of Microbial Ecosystem (EEME). In order to study this aspect, the structure of a hydrogen-producing microbial community has been artificially modified by adding exogenous bacterial strains with redundant functions and/or complementary native strains. Results in batch reactors have shown that the hydrogen production performances could be improved to a 3.5 factor by the addition of certain strains. Results obtained can not be explained by simple trophic interactions and suggest the presence of interaction mechanism of cooperation among microorganisms. Moreover, under more favourable operating conditions (inoculum, culture medium), the addition of certain species in low abundance could stabilize the metabolism of microbial ecosystem without necessarily favourably affect the hydrogen production. In all cases, competitive interactions were not favourable for hydrogen production. Trials were then realised in continuous reactors. These trials have shown that the method used to implant strains in reactors could be a key factor for using the EEME.As a conclusion, this study has shown the potential to use exogenous bacteria, in minority proportions, as biotic factors to stabilised and/or guides microbial metabolisms to functions of interest within microbial mixed cultures.

Dans le futur, différentes sources d'énergies renouvelables comme les énergies de seconde génération produites à partir de déchets lignocellulosiques seront nécessaires pour palier à l'épuisement des énergies fossiles. Parmi ces énergies de seconde génération, le biohydrogène, le méthane et l'hythane produits à partir de procédés fermentaires anaérobies représentent des alternatives prometteuses. Cependant la production de biohydrogène et de méthane à partir de résidus lignocellulosiques est limitée par leurs structures récalcitrantes et une étape de prétraitement en amont des procédés fermentaires est souvent nécessaire. Ce travail a pour but d'étudier l'impact des facteurs biochimiques et structurels des résidus lignocellulosiques sur les performances de production d'hydrogène et de méthane, pour pouvoir par la suite développer des stratégies de prétaitements adaptées. Tout d'abord, sur un panel de vingt substrats lignocellulosiques, les potentiels hydrogène et méthane ont été corrélés aux paramètres biochimiques et structurels. Les résultats ont mis en évidence que le potentiel hydrogène est uniquement corrélé positivement à la teneur en sucres solubles. La production de méthane quant à elle est négativement corrélée à la teneur en lignine et, à un moindre degré, à la cristallinité de la cellulose, mais positivement à la teneur en sucres solubles, holocelluloses amorphes et protéines. Par la suite, des stratégies de prétraitements ont été établies pour améliorer la production d'hydrogène et de méthane. Le couplage prétaitements alcalins/enzymatique ainsi que les prétraitements à l'acide dilué, efficaces pour solubiliser les holocelluloses en sucres solubles ont été appliqués en amont de la production d'hydrogène. En combinant le pretraitement alcalin avec une hydrolyse enzymatique, le potentiel hydrogène des tiges de tournesol fut multiplié par quinze. En revanche, suite aux prétraitements acides, la production d'hydrogène fut inhibée à cause de la libération de sous-produits (furfural, 5-HMF et composés phénoliques) engendrant un changement d'espèces bactériennes vers des espèces non productrices d'hydrogène. Pour la production de méthane, cinq prétraitements thermo-chimiques (NaOH, H2O2, Ca(OH)2, HCl and FeCl3) efficaces pour délignifier ou solubiliser les holocelluloses ont été étudiés. Parmi ces prétraitements, la meilleure condition fut 55°C à une concentration de 4% NaOH pendant 24 h, résulant en une augmentation du potentiel méthane variant de 29 à 44 % en fonction des tiges de tournesol. Cette condition fut par la suite validée en réacteurs anaérobies continusavec une augmentation de 26.5% de la production de méthane. Un procédé à deux étages couplant la production d'hydrogène en batch suivi de la production de méthane en continu fut aussi étudié. Néanmoins, aucune différence significative en termes d'énergie produite ne fut observée entre les procédés à deux étages (H2/CH4) et à un étage (CH4). / In the future, various forms of renewable energy, such as second generation biofuels from lignocellulosic residues, will be required to replace fossil fuels. Among these, biohydrogen and methane produced through fermentative processes appear as interesting candidates. However, biohydrogen and/or methane production of lignocellulosic residues is often limited by the recalcitrant structure and a pretreatment step prior to fermentative processes is often required. Up to date, informations on lignocellulosic characteristics limiting both hydrogen and methane production are limited.Therefore, this work aims to investigate the effect of compositional and structural features of lignocellulosic residues on biohydrogen and methane performances for further developping appropriate pretreatments strategies. Firstly, a panel of twenty lignocellulosic residues was used to correlate both hydrogen and methane potentials with the compositional and structural characteristics. The results showed that hydrogen potential positively correlated with soluble carbohydrates only. Secondly, methane potential correlated negatively with lignin content and, in a lesser extent, with crystalline cellulose, but positively with the soluble carbohydrates, amorphous holocelluloses and protein contents. Pretreatments strategies were further developed to enhance both hydrogen and methane production of sunflower stalks. Dilute-acid and combined alkaline-enzymatic pretreatments, which were found efficient in solubilizing holocelluloses into soluble carbohydrates, were applied prior to biohydrogen potential tests. By combined alkaline-enzymatic pretreatment, hydrogen potential was fifteen times more than that of untreated samples. On the contrary, hydrogen production was inhibited after dilute-acid pretreatments due to the release of byproducts (furfural, 5-HMF and phenolic compounds) that led to microbial communities shift toward no hydrogen producing bacteria. Similarly, methane production, five thermo-chemical pretreatments (NaOH, H2O2, Ca(OH)2, HCl and FeCl3) found efficient in delignification or solubilization of holocelluloses, were considered. Among these pretreatments, the best conditions were 55°C with 4% NaOH for 24 h and led to an increase of 29-44 % in methane potential of sunflower stalks. This pretreatment condition was validated in one stage anaerobic mesophilic continuous digester for methane production and was found efficient to enhance from 26.5% the total energy produced compared to one stage-CH4 alone. Two-stage H2 (batch) / CH4 (continuous) process was also investigated. Nevertheless, in term of energy produced, no significant differences were observed between one-stage CH4 and two-stage H2 /CH4.