Environmental changes in biodiversity hotspot ecosystems of
South Ecuador: RESPonse and feedback effECTs (FOR2730)

Abstract:

Human influence and its impacts are perceptible in all ecosystems resulting in land transformation, changes in global biogeochemistry, climate change, and loss of biological diversity. Mapping the spatial and temporal patterns of human influence is essential to address land use management and conservation programs. In this study, we tailored the Human Footprint index (HF) developed at global level to evaluate the spatial and temporal patterns of human pressure in South Ecuador for 1982, 1990 and 2008. Landscape and ecosystem levels were analyzed to identify the contribution of different human proxies to the HF.We also used the HF to evaluate the effectiveness of protected areas to reduce human pressure in the surrounding landscape. We found that levels of human pressure increased and the wildest areas decreased since 1982. We identified important “hotspots of changes” in the seasonally dry forests in the western part and the premontane evergreen forest in the eastern part of the study area. Our results show that each human proxy contributes in a different way to the observed values of HF in the studied ecosystems. Finally, we found that Podocarpus NP, the most important protected area in our study region, seems to be partially effective in reducing human pressure inside and in the buffer zones where only a low increase in HF was detected. However, the HF values observed in the surrounding landscape were higher than those observed in the buffer zone and inside the protected area. We demonstrated that HF could be a useful regional evaluation tool to facilitate conservation planning.

Abstract:

In einem tropischen Bergregenwaldgebiet in Ecuador, haben verschiedene Forschungsarbeiten gezeigt, dass auch Mikronährelemente einen Einfluss auf die mikrobielle Aktivität haben können. Mittels eines Inkubationsversuches mit einer Streuauflagenprobe aus dem Untersuchungsgebiet, wurde analysiert ob eine multiple Mikronährlimitierung von Mikroorganismen vorliegt. Dabei wurde die Streuprobe in einem sechs Tage dauernden Inkubationsversuch täglich mit verschiedenen Mikronährelementen (Cu, Mo, Mn, Zn, P)gedüngt. Danach wurde die Nährlösung abgesaugt und die daraus entstandenen Eluate mithilfe eines Flammen-Atomabsorptionsspektrometer auf die verbliebenen Metallgehalte analysiert. Zuverlässige Werte konnten nur für Cu, Mn und Zn gemessen werden, die Werte für Mo und P lagen unter der Nachweis- oder Bestimmungsgrenze. Bei der Kupfermessung konnte die höchste Kupferkonzentration in der Cu-gedüngten Probe gemessen werden, doch es wurde schon weniger Cu ausgewaschen, wenn zusätzlich mit einem anderen Mikronährelement gedüngt wurde als in den Kontrolldüngungen. Für Mn wurde in der Kontrolllösung die höchste Mn-Konzentration gemessen, in allen anderen Proben mit zusätzlicher Düngung wurde Mn stärker zurückgehalten. Die gemessene Zn-Konzentration bei Düngung mit Cu, Mn, Mo und P lag fast bei null. Zink wurde sehr stark zurückgehalten. Auch in der Zink-Gedüngten Probe wurde Zn in den ersten drei Tagen stark zurückgehalten und wurde erst danach stärker ausgewaschen als in den Kontrollproben. Aufgrund der Resultate ist zu erkennen, dass auch Mikronährelemente eine limitierende Wirkung auf Mikroorganismen haben können und die Mikroorganismen im untersuchten Gebiet durch verschiedene Mikronährelemente limitiert sind.

Abstract:

Growth limitation induced by Al toxicity is believed to commonly occur in tropical forests, although a direct proof is frequently lacking. To test for the general assumption of Al toxicity, Al, Ca, and Mg concentrations in precipitation, throughfall, stemflow, organic layer leachate, mineral soil solutions, stream water, and the leaves of 17 native tree species were analyzed. We calculated Al fluxes and modeled Al speciation in the litter leachate and mineral soil solutions. We assessed potential Al toxicity based on soil base saturation, Al concentrations, Ca:Al and Mg:Al molar ratios and Al speciation in soil solution as well as Al concentrations and Ca:Al andMg:Al molar ratios in tree leaves. High Al fluxes in litterfall (8.77±1.3 to 14.2±1.9 kg ha?1 yr?1, mean ± SE) indicated a high Al circulation through the ecosystem. The fraction of exchangeable and potentially plant-available Al in mineral soils was high, being a likely reason for a low root length density in the mineral soil. However, Al concentrations in all solutions were consistently below critical values and Ca:Al molar and the Ca2+:Alinorganic molar ratios in the organic layer leachate and soil solutions were above 1, the suggested threshold for Al toxicity. Except for two Al-accumulating and one non-accumulating tree species, the Ca:Al molar ratios in tree leaves were above the Al toxicity threshold of 12.5. Our results demonstrate high Al cycling through the vegetation partly because of the presence of some Al accumulator plants. However, there was little indication of an Al toxicity risk in soil and of acute Al toxicity in plants likely reflecting that tree species are well adapted to the environmental conditions at our study site and thus hardly prone to Al toxicity.

Abstract:

The forests of South Ecuador have high conservation value because they are highly threatened, but also possess high diversity and endemism levels. However, as the process of conservation is inherently spatial, the little available information about the factors that promote vulnerability of these forests as well as the spatial distribution of the conservation targets have produced some conservation pitfalls. The main objective of this dissertation was to generate spatial information about the threats and biological values occurring in this region to prioritize areas for conservation. The second chapter analyzes the deforestation and fragmentation patterns in the region since the 1970s using aerial photographs and satellite images to identify areas with remaining forest, fronts of deforestation, annual deforestation rates and the dynamics in the composition and configuration of the landscape of South Ecuador. We recorded annual deforestation rates of 0.75% (1976 – 1989) and 2.86% (1989 – 2008) for two consecutive survey periods. Also, we found that South Ecuador is experiencing an ongoing fragmentation process due to an increase in the number of patches, a decrease in mean patch size and an increase in the isolation of forest fragments. This study also contributes to a better understanding of forest change dynamics in the tropics. We found that substantial portions of natural forests are being degraded or converted into pastures and that main fronts of deforestation are located in the lowest areas in the premontane evergreen forest. The third chapter describes the temporal and spatial patterns of human pressure, as it is one of the main factors that influence the effectiveness of conservation strategies. At local scale, we adapted at local scale the Human Footprint Index (HF) developed by Sanderson et al. (2002), to evaluate spatial changes in HF during a 26 year period at both landscape and ecosystem levels. This information allowed us to identify “hotspots of change” and the wildest areas remaining in order to evaluate how different human proxies contribute to HF and to demonstrate how effective the most important protected areas have been in reducing human pressure inside and outside their boundaries. The findings show a noticeable 4 increase in human pressure levels in South Ecuador and a progressive reduction in the wildest areas. We also identified that the important “hotspots of changes” are located in the western region and the Rio Zamora river basin. The most impacted vegetation types were seasonally dry forest and shrubland. Here, population density is the human proxy with the highest contribution to the observed patterns. Finally, we found that Podocarpus National Park has been partially effective in reducing human pressure inside and outside its borders. HF levels have increased inside and outside the boundaries of the protected area, but the human pressure was always lower than that observed in the surrounding landscape. The fourth chapter analyzes the patterns of alpha and beta diversity of endemic plant species to evaluate the congruence of both patterns and to identify areas with the highest diversity of endemic plants in order to prioritize areas for conservation. We found that hotspots of alpha diversity are concentrated along the Andes, but this diversity was only slightly congruent with beta diversity patterns of endemic plants mostly concentrated in the western and eastern escarpments of the Andes, and in the Coastal and Amazon cordilleras. We also found that approximately 40% of the areas with the highest alpha and beta diversities have already disappeared due to deforestation and that only 30% is under protection in Ecuador. Thus, we propose 12 potential areas with a high priority of conservation mostly located in South Ecuador to improve the representativeness and complementarity of the current reserve network. Finally, the fifth chapter synthesizes the principal findings of this thesis highlighting the implications for conservation and suggesting potential areas to be preserved based on human pressure levels, remaining forest and alpha and beta diversity patterns of endemic plants.

Abstract:

Anthropogenic activities have produced changes in natural ecosystems worldwide. In tropical regions in South America, industrialization of cities and forest clearance via burning are the main activities releasing pollutants into the atmosphere and inducing changes in nutrient deposition patterns and climate of primary forests. Nitrogen (N) and phosphorus (P) are considered the main limiting nutrients of plant growth since their availability is vital for net primary productivity. Therefore, any change in N or P availability in soils would likely alter important mechanisms of forests dynamics such as growth and survival. Changes in soil pH (acidity), nutrient cycles and altered nutrient stocks affect N and P availability and affect various physiological processes of trees. Although low levels of nutrient deposition have been reported for montane forests in southern Ecuador (ca. 5 kg ha–1 for N, 0.49 kg ha–1 for P), even these levels are expected to lead to changes in forest structure and dynamics over the long term. The responsiveness of forest to changes in resource availability varies with forest age and successional status, as well as life stage of the plant. Young plants (tree seedlings) should be more responsive to changes in nutrient availability than mature forest trees. Further, any demographic response is more likely to be visible in seedlings before mature trees because of the more rapid dynamics of seedlings. Therefore, I studied the regeneration dynamics of montane forest to understand which nutrient-related processes are involved in the growth and establishment of seedlings at both the individual and community levels. The Ecuadorian NUtrient Manipulation EXperiment (NUMEX) has been designed to study the response of montane forest to moderate fertilization. The experiment has been set up over an elevation range across three main study sites (1000 m. a.s.l: Bombuscaro; 2000 m. a.s.l: San Francisco; 3000 m. a.s.l: Cajanuma) within the Podocarpus National Park and San Francisco Reserve. The factorial experiment consists of four blocks containing four experimental plots (N, P, NP and control) in every study site (16 plots per elevation). Fertilization has been done since 2008, adding moderate quantities of fertilizer (50 kg ha–1 y–1 of N and/or 10 kg ha–1 y–1 of P). Three different approaches were used to assess the seedling community and common species responses to fertilization. First, the seedling community was monitored in natural forest at 2000 m over three consecutive years (2011 – 2013). In 192 monitoring subplots (1m2 each) in San Francisco, all seedlings and saplings were mapped and tagged. Measurements of height, diameter, herbivory and leaf production were recorded for all individuals, and the number of recruited and dead seedlings was counted every year. Second, allocation patterns and stoichiometry of seedlings of the six most common species were determined across the elevation gradient. Naturally occurring seedlings of the following species were harvested: Clarisia racemosa and Pouteria torta in Bombuscaro, Graffenrieda emarginata and Palicourea angustifolia in San Francisco and Grafferieda harlingii and Hedyosmum purpuracens in Cajanuma. Morphology (biomass allocation, herbivory and foliar areas such as SLA, LAR and LA) and foliar stoichiometry (nutrient contents and N:P ratios) were recorded and analyzed. Third, a reciprocal transplantation experiment (STE) within the NUMEX experimental plots monitored seedlings of the most common species in Bombuscaro, (Pouteria torta), to assess specific responses in seedling performance over one year. These three levels of analysis showed differential responses of the species community and common species to fertilization. Contrary to lowland forests, the community seedlings in this montane forest showed a moderate response. The density of seedling individuals decreased following nutrient addition, as a result of lower recruitment in treatments plots, but mortality was unaffected. Consequently, density-dependent mechanisms activated by additional N uptake did not show evidence of being a main driver to changes in species density. The community of seedlings did not show significant growth in response to fertilization although seedlings were higher in the N treatment. However, N addition increased herbivory over all seedlings and plots, which could have masked other seedlings growth responses. Common species seem to be well adapted to the relatively poor soils since these species were not favored by nutrient addition showing no change in growth or leaf traits. Over the long-term, common species might lose their dominance by being less competitive than faster growing species. However, it would need a longer period of monitoring under continued fertilization to produce visible shifts in community composition. Species-specific responses are hard to determine in such species rich communities. Therefore, analyses of the six common species were conducted to complement the community-level study. Leaf morphology and foliar stoichiometry were assessed from harvested seedling from the experimental plots. Both N and P had effects on seedling traits, suggesting co-limitation of N and P in montane tree species in all elevations. However, foliar N:P ratios and the more frequently response to P addition indicated this nutrient might be more limiting than N along the gradient. Responsiveness of the studied species varied between the six species. Stronger foliar P concentration compared with N in all species but Pouteria torta suggested higher P consumption in montane forest species, similar to the responses of several other tropical plant species after P addition. Increased herbivory was only evident in the opposite responses of G. emarginata (N and N+P addition) and P. angustifolia (P addition) at 2000 m suggesting that plant selection by herbivores is driven by resource quality. Most of the species had higher belowground biomass in root fractions following fertilization, except for G. harlingii at 3000 m, the only species that showed significant increase in aboveground biomass after nutrient addition. Pouteria torta seedlings showed no significant changes in species performance after fertilization one year after establishment of the STE. Pouteria seems to be well-adapted to poor soils, since N and P addition did not alter foliar nutrient concentrations. Nevertheless, N and N+P addition significantly increased mortality and diameter growth rates. P addition resulted in higher leaf area loss and shifted carbon allocation to root growth. These responses indicated consequences in the competitive strength in the understory and recruitment success of Pouteria torta. However, the ambiguous response in some attributes (growth and herbivory) not related with mortality made it difficult to predict the future abundance of this species in long term. In conclusion, complementary studies demonstrated that specific nutrient limitation for N or P in montane species seem not to be a rule in rich species ecosystems although nutrient addition did alter some pattern of growth and survival. Thus, nutrient fertilization might affect various mechanisms and dynamics of plant communities, the balance of which will only play out over long time scales.

Abstract:

Tropische und subtropische Trockenwälder zählen zu den bedrohtesten tropischen Ökosystemen (MILES et al. 2006). Sie bieten für mehr Menschen Siedlungsraum als humide Waldgebiete und sind dadurch besonders stark dem menschlichen Einfluss ausgesetzt. Die vorliegende Arbeit beschäftigt sich mit der Entwicklung der Baumverjüngung in der tumbesischen Trockenwaldregion Südecuadors und Nordperus und dem Einfluss der dort üblichen Waldweide. In Reservaten in der Provinz Loja (Ecuador) und im angrenzenden Nationalpark Cerros de Amotape auf peruanischer Seite wurde in 288 Parzellen (je 4 m²) die Baumregeneration registriert. Die Parzellen waren im Trockenwald mit den Vegetationsformen halb-laubabwerfend und laubabwerfend verteilt; zudem wurde die Hälfte der Parzellen eingezäunt, um den Einfluss der Beweidung auf die Regeneration zu erfassen. Die Aufnahmen fanden in einem Zeitraum von zwei Jahren zweimal während der Trockenzeit und zweimal kurz nach der Regenzeit statt, beginnend mit einer Aufnahme in der Trockenzeit 2014. Es wurden die Individuen markiert und die Artnamen dokumentiert, Höhen und abgestorbene Individuen wurden aufgenommen. Aus den Daten wurden die Höhenzuwächse, die Mortalitätsraten sowie die Shannon- und Evennessindices berechnet. Insgesamt wurden 93 Arten aus 36 verschiedenen Familien aufgenommen. Die Ergebnisse zeigten für den Erhebungszeitraum eine höhere Individuenanzahl während der Regenzeit, die Anzahl der Arten war im Vergleich zur Trockenzeit jedoch geringer. Die Diversität unterschied sich nicht signifikant zwischen den Jahreszeiten. Bei Betrachtung der Aufnahmen stach das zweite Monitoring mit besonders hohen Individuenanzahlen heraus. Zwischen den Waldtypen gab es einen signifikanten Unterschied in der Höhe der Verjüngung, sowie im Höhenzuwachs von der dritten auf die vierte Aufnahme. Die offenen und eingezäunten Flächen unterschieden sich ebenso in den Höhen und dem Höhenzuwachs signifikant. Die Diversität unterschied sich mit und ohne Zäunung und bei den beiden Waldtypen nicht eindeutig. Die Ergebnisse können unter anderem durch die Struktur des Altbestandes, dessen Einfluss auf die Strahlung, die Temperatur und die Feuchtigkeit, und wiederum deren Einfluss auf die Verjüngung erklärt werden. Lokale und regionale floristische Diversität und Zusammenstellung, Wachstumsformen, Phänologie und Demographie der Verjüngungspflanzen sind in weiten Formen besonders abhängig von der Hydrologie. Beweidung andererseits lenkt die Ausbreitung Verbiss resistenterer Arten, welche durch Reduzierung der Weidetiere zur Walderhaltung beitragen können. Diese Studie soll als lokale Forschungsarbeit zum besseren Verständnis des Ökosystems Trockenwald beitragen. Die gesammelten Erkenntnisse sollen später der Entwicklung nachhaltiger Landnutzungssysteme dienen, welche den Fortbestand dieses gefährdeten Habitats sichern.

Abstract:

Tropical forests and the trees as their principal components have been investigated in detail. However, due to its complexity, their interactions, adaptations and response to climate variations require much more research. In this study, dendrochronological techniques were applied to evaluate the potential of tree-rings from tropical tree species as climate records. Two ecosystems with very distinct climate scenarios were selected from a dry and humid forest in southern Ecuador. A comparative analysis between these two forest types was performed by applying three dendrochronological methods. First, Tree Ring Width (TRW) measurements from tree species with distinct ring boundaries were dated to develop ring-width chronologies. Second, stable carbon isotopes (?13C) were measured from whole-wood and alpha-cellulose of dated annual tree-rings. Finally, concentrations of more than 23 chemical elements were determined from individual dated tree-rings after dissolving the wooden material in HNO3. The results showed the high potential of tropical tree species as climate archives, Bursera graveolens and Maclura tinctoria for the dry forest and Cedrela montana for the humid forest. Radial growth variations in tree species from the dry forest revealed a strong and reliable precipitation signal. Then, for these tropical regions, the first ring-width based wet-season precipitation reconstruction over the past century was developed, and spatial correlations unraveled a strong connection to the climatic conditions of the central Pacific precipitation and temperature variability. Interseries correlations of the TRW from the trees of the humid forest revealed a weak common signal. Stable carbon isotopes evidenced higher climate sensitivity than TRW measurements in the humid forest. However, to infer a reliable climate reconstruction from stable carbon isotopes, more ?13C time series were needed. ?13C values from whole-wood and alpha-cellulose reflected local and regional signals of precipitation and humidity. Meanwhile, nutrient concentration in the wood was higher in the dry forest, but common patterns and trends of nutrients were more distinct in the humid forest. For both study sites, two groups of nutrients with opposite radial distribution were identified (Group 1: Ca, Sr, Ba, Ga; and Group 2: K, P, Rb). In conclusion, TRW of tree species from the dry forest have a high paleoclimate potential, especially to reconstruct precipitation amounts in arid zones of southern Ecuador. Stable carbon isotopes constitute a promising tool to perform climatic reconstructions in both ecosystems. Finally, the valuable historical information of nutrient concentration evidenced in tree-rings opens promising ways to study tree growth dynamics especially in the humid forest.

Abstract:

Zwar lieferten die Analysen in dieser Arbeit keine eindeutigen Antworten auf die Fra-gestellung, dennoch bietet sie Ansatzpunkte für weitere Untersuchungen zu diesem Thema. Zum einen wäre die genaue Erforschung des Reliefs vor Ort sinnvoll. Dabei könnte einerseits der Verlauf der Fließgewässer überprüft werden, andererseits könn-ten die Einzugsgebiete auf Hangrutschungen untersucht werden, welche die Boden-verhältnisse stören können. Des Weiteren kann umliegende Vegetation, die die Bäume möglicherweise beeinflussen könnte, erforscht werden. Die aufgestellte Hypothese wird in dieser Arbeit vollständig abgelehnt, da keinerlei räumliche Muster in der topographischen Lage der Bäume erkennbar sind. Grund da-für sind vor allem eingeschränkten Möglichkeiten in ArcGIS die Topographie des Ge-bietes detailliert untersuchen zu können. Zum anderen scheint Lage der Bäume im Relief keine ausschlaggebende Ursache für deren Isotopenverhältnisse darzustellen. Sehr wahrscheinlich sind die lokalen Bodenverhältnisse entscheidend für das Isoto-pensignal der Bäume. Da die örtliche Bodenbeschaffenheit zum Teil durch Rutschkör-per gestört sein kann, ist eine Untersuchung dieser maßgeblich für zukünftige For-schungen.

Abstract:

Facing the ongoing loss of natural ecosystems, worldwide monitoring of biodiversity across different spatial scales is essential for conservation planning. Remote sensing (RS) has proven to be a cost-efficient tool to access environmental characteristics such as vegetation structure and associated distributions of animal species on a broad scale. Special emphasis is put on birds as indicators for biodiversity owing to their strong species–habitat relationship. So far, bird diversity was modeled ignoring that species–habitat relationships differ among feeding guilds. This is surprising, since habitat preferences strongly depend on diet specialization. Therefore, I investigated RS texture image based vegetation metrics to test whether the predictability of specialized avian feeding guilds including insectivores, frugivores and nectarivores is higher than of the less specialized omnivore guild and overall bird diversity. I used point count data of bird communities among 30 study sites in a complex tropical mountain forest ecosystem in south-eastern Ecuador to estimate (i) Shannon index and (ii) community composition as measures of ?-diversity and combined ?- and ?-diversity, respectively. In order to relate both diversity measures to RS metrics, I compared two high dimensional predictor sets – satellite images and airborne orthophotos – with structural indices derived from a discrete return airborne Lidar sensor. Partial least squares regression was used to unveil the predictive power of all fitted feeding guild models. For the comparability of all models, a sample size correction on species number per guild was applied. Shannon index predictability ranged between 37 % and 65 %; and best predictions were achieved for insectivores using metrics from satellite or Lidar images and nectarivores species using metrics from orthophotos. Community composition was generally better predicted than Shannon index with explained variations from 65 % to 85 %. Frugivore and nectarivore community compositions were best predicted using metrics from orthophotos, whereas the two other sensors best predicted omnivores. For both diversity measures, performance of satellite derived metrics revealed slightly better model results compared to other sensors emphasizing its applicability for the regarded study area. In conclusion, specialized feeding guilds were not consistently better predicted than omnivore or overall bird diversity; rather the study showed that model performances depended on the regarded diversity measure and RS image type. However, insectivores might be the best surrogate for overall diversity with high predictability in all compared models. In addition, the high explanatory power for community composition suggests that the measure should considered in avian diversity modeling for conservation planning.

Abstract:

Ant assemblages are sensitive to abiotic changes in the environment, therefore they are widely used as indicators of environmental changes. Previous studies demonstrated that abiotic changes with elevation and increased anthropogenic disturbance not only reduce species richness of ant assemblages, but also modify their trophic composition and nutrient use. In tropical ecosystems where nutrient availability may vary between dry and wet seasons, seasonal or interactive effects might play an important but still neglected role in shaping these patterns. Here I used standardized bait experiments in natural and disturbed sites along an elevation gradient in a tropical montane rainforest during the wet and dry season. In order to analyze the single and interactive effects of elevation, disturbance and season on species richness, nutrient use and trophic composition of ants, I used linear mixed effect models. Additionally, I used principal component analysis (PCA) to assess whether morphological traits of ants are linked to their nutrient use. Species richness decreased monotonically along the elevation gradient, with a stronger decline in the dry season. Forest disturbance had no significant effect on species richness. The relative use of most nutrients decreased with increasing elevation. Forest disturbance only affected the relative use of lipids by decreasing it compared to natural forests. However, my results revealed complex interactive effects of elevation, disturbance and season on species richness and the use of nutrients by ant assemblages. Furthermore, I found a shift from predominantly omnivore species to more predatory species with increasing elevation. PCA revealed a preference of lipid baits by species with morphological traits associated with predatory taxa. My findings highlight the importance of seasonality and mixed effects on the composition of ant assemblages and their nutrient use in a tropical montane forest. Additionally, these results highlight the value of disturbed forests within my study area, since they support similar species richness and trophic composition of ant assemblages compared to natural forests.

Abstract:

The worldwide demand for large scale biodiversity monitoring systems is challenging for recent remote sensing techniques. For monitoring changes in species as well as functional diversity the consideration of ecosystem processes is crucial. Herbivory as a focal function in tropical forest ecosystems is therefore being investigated in this study. Aim of the study is to reveal the potential of remotely sensed image textures for predicting herbivory. For this leaf area loss (LAL) due to herbivory has been quantified in different vegetation layers of a tropical mountain forest in the Ecuadorian Andes. The study area is a structurally highly complex region along an elevation gradient from 1000 to 3000 m a.s.l. with undisturbed primary forest and disturbed forest fragments within agriculturally used land. It has been proven that LAL has similar patterns as herbivore abundance over the elevation gradient showing it to be a possible proxy for herbivore abundance. Further results show a correlation of LAL data and canopy cover thus local forest structure. Correlations between LAL and image textures as a proxy for vegetation structure were stronger than with pixel wise derived spectral values only. Unveiling the high potential of image textures as a surrogate for herbivory and ecosystem function.

Abstract:

The South Ecuadorian Andes are one of the hottest global hotspots of biodiversity and currently threatened by land use and climate change. Forest structure and composition are crucial factors for understanding the capacity of forest support species in changing environments. In order to prioritize limited conservation resources a better understanding of tree diversity patterns is needed. The use of image texture measures, as a proxy for spatial and forest structure has shown useful possibilities in explaining patterns of tree diversity and species richness. My goal was to evaluate the performance of different texture measures on NDVI, EVI and two broad-band combinations on high resolution (0.3m) aerial photography to predict tree diversity. Bayesian Model Averaging (BMA) was used to relate in situ measurements of tree diversity to measures of image texture. Texture explained up to 44.7% of the variability of tree diversity with measures related to habitat heterogeneity, particularly variance, providing highest explanatory power. Image texture measures bear considerable potential for predicting tree diversity in the tropics and can contribute to improvements on conservation efforts and management planning.

Abstract:

The comparatively long research history within the study area allowed to develop well-grounded hypotheses on how the hydrological system within the study area should behave (Bogner et al., 2014; Boy et al., 2008; Bücker et al., 2011, 2010; Crespo et al., 2012, 2011; Fleischbein et al., 2006; Goller et al., 2005). Based on those earlier findings made within the study area and the hydrological research conducted elsewhere in tropical mountain cloud forests (see Bonell and Bruijnzeel, 2004; Bruijnzeel, 2004; Bruijnzeel et al., 2011 for recent overviews) provided the background to shape the research conducted within this dissertation around the focal point of global change impacts on the hydrological cycle inside the study area (see chapter 1.3).Separated into three chapters, each representing a scientific research paper, this thesis will address the issue on how to evaluate effects of land-use change on the hydrological cycle (chapter 2), how stable water isotopes are distributed over space and time within the study area (chapter 3) and how stable water isotopes can be used in hydrological models to assist this analysis of global change impacts in the future (chapter 4). All of the research conducted for this dissertation was performed in the Rio San Francisco catchment within South Ecuador.

Abstract:

The present study is composed of three interlinked research papers described in three respective chapters: Chapter 2: “Understanding uncertainties when inferring mean transit times of water trough tracer-based lumped-parameter models in Andean tropical montane cloud forest catchments” deals with the estimation of mean transit times and explores the most reliable transit time distribution functions for 32 sampled sites in the catchment, which included 18 soil water sites, 2 streamwater sites in the main river, 7 streamwater tributaries, 2 small creeks and 3 springwaters. For the analysis, seven lumped-parameters models were tested. Besides, for each simulation, uncertainties were accounted through applying the Generalized Likelihood Uncertainty Estimation approach (GLUE). Chapter 3: “Sampling frequency trade-offs in the assessment of mean transit times of tropical montane catchment waters under semi-steady-state conditions”. This chapter describes the effect of sampling resolution data on the results provided by lumped-parameter models. Based on the insights from results described in Chapter 2, only selected models and sampled sites were screened for intercomparison. The results of this topic were aimed to account the degree of reliability of mean transit time estimations. Chapter 4: “Multi-criteria assessment of water dynamics reveal subcatchment variability in a seemingly homogeneous montane rainforest catchment”. By considering a nested catchment approach (main river outlet and its seven tributaries), mean transit times and other diverse catchment characteristics data (e.g., hydrometric and topographic indices, physicochemical properties of stream waters) were used to perform diverse statistical and modelling techniques, such as bivariate and multivariate analysis, Principal Component Analysis (PCA), Cluster Analysis, EMMA, Hydrological and Chemical modelling. All of which allowed to gain insights on the key factors governing the flow processes in each analysed site.

Abstract:

In-Situ-Konzentrationsmessungen sind ein wichtiges Werkzeug in der Hydrologie. Die Ver-wendung entsprechender Geräte erlaubt die Aufnahme von hochfrequenten und gleichzeitig langfristigen Zeitreihen. Nitratkonzentrationen unterliegen in natürlichen Fließgewässern ta-geszeitabhängigen Schwankungen. In dieser Arbeit werden Konzentrationsmessungen aus dem Rio San Francisco in Ecuador aus-gewertet. Ergänzt werden diese durch Wettermessungen und Pegelstandsmessungen aus dem gleichen Gebiet. Der Nitratdatensatz verläuft über etwa viereinhalb Jahre und enthält viertel-stündliche Messungen. Diese werden aufbereitet, gefiltert, und nach Tageszyklen hin unter-sucht. Zum Einsatz kommt dabei eine hierarchische Clusteranalyse mit Dynamic-Time-War-ping-Distanzen. Um Zusammenhänge der Zyklen mit den anderen Variablen aufzudecken, wird eine Lineare Diskriminanzanalyse (LDA) mit einer Variablenauswahl nach Wilks-Lambda durchgeführt. Zusätzlich werden unabhängig von Tageszyklen kurzfristige Residuen der Nit-ratkonzentration per multimodal normalverteiltem Mischungsmodell untersucht. Die Clusteranalyse zeigt, dass es einen Cluster A mit einem regelmäßigen Tageszyklus sowie einen Cluster B mit unregelmäßigen Konzentrationsverläufen gibt. Die Trennung von A und B scheint mit der Lufttemperatur, der relativen Luftfeuchtigkeit, dem Niederschlag, und dem Pe-gel zusammen zu hängen. Ein LDA-Modell mit den Tagesvariablen Pegelamplitude, Nieder-schlagssumme, und mittlere relative Luftfeuchtigkeit ordnet erfolgreich die meisten Tage dem Cluster A zu, versagt jedoch bei der Zuordnung in Cluster B. Die Analyse mit dem Mischungsmodell ergibt keine eindeutigen Ergebnisse. Es scheint aber einen negativen Zusammenhang zwischen den Residuen der Nitratkonzentration mit der Luft-temperatur und der Globalstrahlung zu geben. Die Ergebnisse dieser Arbeit entsprechen teilweise jenen, welche auch in der Literatur aufzu-finden sind. Der Tagesverlauf mit einem Maximum am Morgen und einem Minimum am Abend scheint typisch für Fließgewässer zu sein. Jedoch findet sich in den Daten kein Hinweis auf eine saisonale Verschiebung des Zyklus.