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

Abstract:

Tropical regions are currently undergoing remarkable rates of land use change accompanied by altered litter inputs to soil. In vast areas of Southern Ecuador forests are clear cut and converted for use as cattle pastures. Frequently these pasture sites are invaded by bracken fern, when bracken becomes dominant pasture productivity decreases and the sites are abandoned. In the present study implications of invasive bracken on soil biogeochemical properties were investigated. Soil samples (0–5 cm) were taken from an active pasture with Setaria sphacelata as predominant grass and from an abandoned pasture overgrown by bracken. Grass (C4 plant) and bracken (C3 plant) litter, differing in C:N ratio (33 and 77, respectively) and lignin content (Klason-lignin: 18% and 45%, respectively), were incubated in soils of their corresponding sites and vice versa for 28 days at 22 C. Unamended microcosms containing only the respective soil or litter were taken as controls. During incubation the amount of CO2 and its d13C-signature were determined at different time intervals. Additionally, the soil microbial community structure (PLFA-analysis) aswell as the concentrations of KCl-extractable C and N were monitored. The comparison between the control soils of active and abandoned pasture sites showed that the massive displacement of Setaria-grass by bracken after pasture abandonment was characterized by decreased pH values accompanied by decreased amounts of readily available organic carbon and nitrogen, a lower microbial biomass and decreased activity as well as a higher relative abundance of actinomycetes. The d13C-signature of CO2 indicated a preferential mineralization of grass-derived organic carbon in pasture control soils. In soils amended with grass litter the mineralization of soil organic matter was retarded (negative priming effect) and also a preferential utilization of easily available organic substances derived from the grass litter was evident. Compared to the other treatments, the pasture soil amended with grass litter showed an opposite shift in the microbial community structure towards a lower relative abundance of fungi. After addition of bracken litter to the abandoned pasture soil a positive priming effect seemed to be supported by an N limitation at the end of incubation. This was accompanied by an increase in the ratio of Gram-positive to Gram-negative bacterial PLFA marker. The differences in litter quality between grass and bracken are important triggers of changes in soil biogeochemical and soil microbial properties after land use conversion.

Abstract:

The genus Eois (Larentiinae) comprises an important part of a megadiverse assemblage of geometrid moths in the mountain rainforests of southern Ecuador. Owing to the diversity of morphotypes found in this genus, the abundance of many of its component species, and the close relationships with larval food plants in the genus Piper, the genus Eois has been chosen as a target for a more detailed investigation of local species diversity, species delimitations, and ecological segregation. COI barcode sequences (676 bp) were generated from approximately 350 individuals to establish an inventory of DNA barcodes from one study area (Reserva Biológica San Francisco). This inventory will be used to match larvae to known adult samples and will be of vital help to extend the currently very limited knowledge about food plant relationships and host specialization. Analyses of barcoding sequence divergence show that when a conservative morphology-based approach in species delimitation is used a few cases arise were maximum within-species distances exceed minimum betweenspecies distances. However, there is a varying degree of evidence for the presence of cryptic species in all of these critical cases, as gained from analyses of elongation factor 1-alpha sequences, phylogenetic analysis and morphological reexamination. Up to 33 previously unrecognized species could be identified with the help of COI barcodes and morphological evidence in addition to at least 10 entirely new species, thereby increasing the number of Eois morphospecies in that small area of Andean mountain forests from 99 to 142. Notably there are no cases where two or more previously known morphospecies had to be lumped (i.e. there was no cryptic polymorphism). Morphospecies were either confirmed or evidence for cryptic species was found. We acquired elongation factor 1-alpha sequences and extended COI sequences of approximately 100 individuals representing at least 70 Eois species amounting to a combined sequence dataset of ~2200 bp. Maximum parsimony and Bayesian inference of phylogeny have been employed to estimate relationships within Eois. Phylogenetic trees reveal that all included wing-pattern types except for one evolved only once. The exception to this rule occurs in the most basal clade whose species closely resemble those in a clade higher up in the tree. This may represent an ancestral character state or the result of convergent evolution. Additionally it could be determined that species known to feed on Piper are spread over most of the major clades within Eois, indicating that the host plant relationship with Piper is indeed a trait found all over the neotropical members of the genus and not just certain subclades. Within the Larentiinae Eois has traditionally been placed close to the Eupitheciini, but there is presently no support for such a placement in our phylogenetic analyses.

Abstract:

During four months of field surveys at the Reserva Biológica San Francisco in the south Ecuadorian Andes, caterpillars of 59 Geometridae species were collected in a montane rainforest between 1800 and 2800m altitude and reared to adults. The resulting data on host plant affiliations of these species was collated. The preimaginal stages of 58 and adult stages of all 59 species are depicted in colour plates. Observations on morphology and behaviour are briefly described. Five species, documented for the first time in the study area by means of larval collections, had not been previously collected by intensive light-trap surveys. Together with published literature records, life-history data covers 8.6% of the 1271 geometrid species observed so far in the study area. For 50 species these are the first records of their early stages, and for another 7 the data significantly extend known host plant ranges. Most larvae were collected on shrubs or trees, but more unusual host plant affiliations, such as ferns (6 geometrid species) and lichens (3 geometrid species), were also recorded. Thirty-four percent of the caterpillars were infested by wasp or tachinid parasitoids.

Abstract:

Abstract. Available proxy records witnessing palaeoclimate of the tropical Andes are comparably scarce. Major implications of palaeoclimate development in the humid and arid parts of the Andes are briefly summarized. The long-term behaviour of ENSO has general significance for the climatic history of the Andes due to its impact on regional circulation patterns and precipitation regimes, therefore ENSO history derived from non-Andean palaeo-records is highlighted. Methodological constraints of the chronological precision and the palaeoclimatic interpretation of records derived from different natural archives, such as glacier sediments and ice cores, lake sediments and palaeo-wetlands, pollen profiles and tree rings are addressed and complementary results concerning former climatic conditions are discussed in terms of possible implications of former atmospheric circulation patterns and main climatic forcing factors. During the last years, increasing tree-ring information is getting available from the tropical Andes, providing high-resolution climate-sensitive records covering the past centuries for the study of climate variability.

Abstract:

Abstract Mountain pastures dominated by the pasture grass Setaria sphacelata in the Andes of southern Ecuador are heavily infested by southern bracken (Pteridium arachnoideum), a major problem for pasture management. Field observations suggest that bracken might outcompete the grass due to its competitive strength with regard to the absorption of photosynthetically active radiation (PAR). To understand the PAR absorption potential of both species, the main aim of the current paper is to (i) parameterize a radiation scheme of a two-big leaf model by deriving structural (LAI, leaf angle parameter) and optical(leaf albedo, transmittance) plant traits for average individuals from field surveys, (ii) to initialise the properly parameterized radiation scheme with realistic global irradiation conditions of the Rio San Francisco Valley in the Andes of southern Ecuador, and (iii) to compare the PAR absorption capabilities of both species under typical local weather conditions. Field data show that bracken reveals a slightly higher average leaf area index (LAI) and more horizontally oriented leaves in comparison to Setaria. Spectrometer measurements reveal that bracken and Setaria are characterised by a similar average leaf absorptance. Simulations with the average diurnal course of incoming solar radiation (1998-2005) and the mean leaf-sun geometry reveal that PAR absorption is fairly equal for both species. However, the comparison of typical clear and overcast days show that two parameters, (i) the relation of incoming diffuse and direct irradiance, and (ii) the leaf-sun geometry play a major role for PAR absorption in the two-big leaf approach: Under cloudy sky conditions (mainly diffuse irradiance), PAR absorption is slightly higher for Setaria while under clear sky conditions (mainly direct irradiance), the average bracken individual is characterized by a higher PAR absorption potential. (~74 MJ m-2 a-1) . The latter situation which occurs if the maximum daily irradiance exceeds 615 W m-2 is mainly due to the nearly orthogonal incidence of the direct solar beam onto the horizontally oriented frond area which implies a high amount of direct PAR absorption during the noon maximum of direct irradiance. Such situations of solar irradiance favouring a higher PAR absorptance of bracken occur in ~36% of the observation period (1998-2005). By considering the annual course of PAR irradiance in the San Francisco Valley, the clear advantage of bracken on clear days (36% of all days) is completely compensated by the slight but more frequent advantage of Setaria under overcast conditions (64% of all days). This means that neither bracken nor Setaria show a distinct advantage in PAR absorption capability under the current climatic conditions of the study area.

Abstract:

The drivers of spatial throughfall heterogeneity are still not fully understood. At an undisturbed forest site in the Ecuadorian Andes with ca. 2600 mm of annual rainfall we determined the accuracy of throughfall measurements by comparing Hellmann-type funnel gauges with troughs. At the same undisturbed and a managed, selectively-logged forest site we determined spatial variability of throughfall, temporal stability of spatial variability and the controls of spatial throughfall variability using a 4-year dataset in weekly resolution. There were no systematic differences between the collected volumes of funnel gauges and troughs. Based on the statistical distribution of annual throughfall volumes, a high number of 27 funnel-type rainfall collectors were required in the undisturbed forest and 20 in the managed forest to estimate throughfall with an error of 10% and a confidence interval of 95%. Spatial throughfall variability in the studied forests was high, markedly stable during 4 years and similar in six selected rain events suggesting that a stable canopy structure controlled throughfall variability. After mathematically eliminating the canopy influence, no meteorological variable had a significant effect on throughfall variability. We conclude that the high spatial variability of throughfall in the study forest, mainly controlled by a long-term stable canopy structure, contributes to the creation of different ecological niches which are a prerequisite for the enormous biological diversity of the north Andean forests.

Abstract:

We studied the spatial heterogeneity of tree diversity, and of forest structure and productivity in a highly diverse tropical mountain area in southern Ecuador with the aim of understanding the causes of the large variation in these parameters. Two major environmental gradients, elevation and topography, representing a broad range of climatic and edaphic site conditions, were analyzed. We found the highest species richness of trees in valleys <2100 m. Valleys showed highest values of basal area, leaf area index and tree basal area increment as well. Tree diversity also increased from ridges to valleys, while canopy openness decreased. Significant relationships existed between tree diversity and soil parameters (pH, total contents of Mg, K, Ca, N and P), and between diversity and the spatial variability of pH and Ca and Mg contents suggesting a dependence of tree diversity on both absolute levels and on the small-scale heterogeneity of soil nutrient availability. Tree diversity and basal area increment were positively correlated, partly because both are similarly affected by soil conditions. We conclude that the extraordinarily high tree species richness in the area is primarily caused by three factors: (1) the existence of steep altitudinal and topographic gradients in a rather limited area creating a small-scale mosaic of edaphically different habitats; (2) the intermingling of Amazonian lowland plant species, that reach their upper distribution limits, and of montane forest species; and (3) the geographical position of the study area between the humid eastern Andean slope and the dry interandean forests of South Ecuador.

Abstract:

Since 1998, members of the DFG investigation group 402 (Since 2003 DFG investigation group 816) devoted themselves to the analysis of a local mountainous rainforest system in the Cordillera Real in southern Ecuador. The ?upper treeline? was subject to a closer study because of the atypical lowering of the local treeline ecotone at 2600 to 3350 m a.s.l. and the missing scientific findings on this topic. Within this thesis, methods of vegetation research as well as climatologic and pedologic analysis were used. Vegetation was taken up at ten different treeline sites ranging from the northern boundary of the Andes Depression near Cuenca in southern Ecuador to Huancabamba in the north of Peru. At each investigation area, eight vegetation transects (100 m2 each) were analysed and trees with a trunk diameter of >5 cm were recorded to species lists. The coverage of climatic parameters was carried out following the North-South-gradient at six sites. Three of them were analysed in detail by means of comparative micro-climatic measurements between forest and Páramo. The pedologic research was carried out at three lower-located sites of the upper treeline ecotone in the centre of the investigation area. The chapter of results is ordered into three sections, namely, structures of vegetation and its characteristics, climate and soil. In the chapter of botanical geography, syntaxonomic similarity of each transect and site was calculated by means of statistical classification methods and forest species with their particular populations were presented in detail. Results show that the upper treeline of the Andes Depression is completely different from the outer areas of the Depression in floristic terms. In the centre of the Andean Depression, which is rich in endemic species, up to 66 tree species form part of the forest border line (concerning an area of only 800 m2!). This number falls drastically at the northern edge of the area of examination, where the upper treeline rises to barely more than 4000 m a.s.l. Apart from the reduction of species with rising sea level, this fact can be postulated by distribution of Polylepis and different treeline structures. Monotypic Polylepis forests containing less species only exist at the northern edge of the Andes Depression and at the most southern study site in the West of Huancabamba. At all other sites, polytypic forests containing many different species are characteristic of the local treeline ecotone. The sites? taxonomic differences can be confirmed by ordination methods, which allow a clear floristic separation between Polylepis-free areas with populations of lowland taxa and Polylepis-forests. The different combinations of species are mainly due to variations in altitude, temperature, the number of humid months, and latitude. Nevertheless, the existence of an enormous range of species itself builds the precondition for a huge potential and a variable mixture of available taxa. The analysis of climate data shows that soil as well as air temperatures in the centre of the study area are clearly above 5.5° C, the postulated threshold value for tree growth. Therefore, temperature cannot be considered to cause the lowering of the local treeline ecotone of the Andean Depression compared to the Northern and Central Andes. Instead, quasi-permanent East winds, high amounts of precipitation throughout the year and maybe even extreme incidents of radiation lead to the depression of the local treeline ecotone system. While high global radiation leads to radiation stress especially for tree seedlings situated outside closed forest stands, strong East winds additionally constrain the establishment of forests in the higher-located ridge areas of Páramo formations. High precipitation leads to a shortage of nutrients in the more gently inclined ridges where the interflow is lower compared to steeper slopes. The results of chemical soil analysis show clearly that the three examined sites are characterized by low pH-values as well as a lack of nutrients. Furthermore, the upper soil layers of the Páramo areas show higher aluminium toxicity due to the type of litter decomposed by Páramo plants, a fact hindering the settling of young trees. Among the natural factors which prevent the growth of forests in higher areas are high rates of radiation, an extreme wind speed, as well as enormous rainfalls. Other sites restricted in forest growth are caused by human influence. The high diversity of species is remarkable in the upper treeline ecotone in the centre of the Andes Depression, a fact that can be ascribed to the lack of nutrients (strongly competitive species are locked out, cf. soil nutrient hypothesis) on the one hand. On the other hand, high precipitation amounts are also responsible for the lowering of the treeline. The result is the combination of a clearly marked forest treeline depression combined with an astonishing treeline complexity which is still an almost unknown paradox.