Biogeographical transect studies in the high elevation mountain areas of Myanmar
Summary This study has its focus on elevational biodiversity patterns of ferns and fern allies in three mountain areas of the southeastern periphery of the Himalayan arc in northern Myanmar where the two large floristic realms of Eurasia meet, the Holarctic realm in the north and the Palaeotropic r...
|PDF Full Text
No Tags, Be the first to tag this record!
This study has its focus on elevational biodiversity patterns of ferns and fern allies in three mountain areas of the southeastern periphery of the Himalayan arc in northern Myanmar where the two large floristic realms of Eurasia meet, the Holarctic realm in the north and the Palaeotropic realm in the south. Because complex geographical configuration of mountains leads to high biological and climatic diversity, thus the elevational gradient represents a model system to assess the underlying mechanism of larger scale species richness and diversity patterns. Moreover, diversity pattern represents the regional evolutionary process of species and their ecological sorting,
the following hypotheses are underlined to meet the objectives of this study to link species diversity patterns on geographical space and evolutionary time connecting with environmental influences.
(1) Species richness pattern in Myanmar differs from other gradient within Himalaya and climate variables are the influencing factors for elevational richness pattern.
(2) The absence of phylogenetic diversity pattern is typical for ferns in accordance with lack of spatial barrier with respect to their long distance dispersal rate and reproductive behavior.
(3) Complexity of hump shaped pattern gradually turning to linear pattern along latitudinal gradients and the parameters of the climate-richness relationship derived from a subset of those elevational gradients should be able to describe the climate-richness relationship at any point along the latitudinal gradient. Data collection was carried out during three expeditions of together 33 weeks in the field, with support of up to 45 local helpers and team members of the Myanmar partner institutions. For the first time ever worldwide, aiming at a complete inventory of all species of seed-plants, pteridophytes and bryophytes, samples were recorded plot-based in three transects of four plots of 400 m² at every 200 meters in elevation between the foot of the slopes and summit of mountain up to 4200 m. In total, 14,485 seed plant specimens, 3,978 pteridophyte specimens and 160 gross samples of bryophytes were collected; most of them in three sets. Additionally, two transects of nine automatically recording climate stations were established between 400 and 3200 m. This first plot based inventory of ferns of in total 132 vegetation plots revealed 299 species from 72 genera and 24 families from Hponyinrazi and Hponkanrazi and Natma Taung gradient. Out of 299 species, 125 are new to Myanmar, in other words the species were previously only known from neighboring countries. The taxonomic composition is strikingly similar to what is known from the very well-known Neotropical regions (Kluge et al 2006), which may be explained by the ability of long distance dispersal of fern spores and the relatively old age of major fern
lineages (Smith 1972). One species is new to science, which was found only five times along the gradient in Northern Myanmar, Selliguea kachinensis Hovenkamp, S. Linds., Fraser-Jenk., sp. nov. Additionally a new combination Selliguea erythrocarpa (Mett. ex Kuhn) Hovenkamp was described, based on the specimens from ou collection. With respect to the new species it is important to mention that the generic placement of this species is rather unclear, because morphological characteristics are close to either Selliguea Bory or Arthromeris (T. Moore) J. Sm. However, an erection of a new genus was avoided due to the general taxonomic uncertainty of this
group of polpodial ferns. These data contribute to the upcoming Flora of Myanmar, but also set the basis for further taxonomic research in one of the least known areas worldwide (Chapter --> 3, 4). Another groundbreaking contribution to the diversity knowledge of the mountain forests of northern Myanmar is the unexpected great
richness of angiosperm species with maximum morphospecies counts of nearly 400 on plots of the submontane belt. It seems probable, but has to be tested based on determined species sets, that these forests are among the richest worldwide.
Testing of above-mentioned main hypotheses was accomplished within three work packages (WP).
(WP 1) The assessment of elevational richness showed that angiosperm patterns differ not only from other groups such as ferns and epiphytes but also from the interpolated result of other studies of the central Himalaya. In contrast to plant-list based distribution graphs, the real distribution of angiosperms do not show the widely known hump-shaped diversity, but have a more or less linear decline with increasing elevation. This pattern was mainly driven by trees, shrubs, and palms, whereas especially epiphytes clearly peaked around 1200 m a.s.l., and grasses peaked at the high elevations above treeline. Climatic variables and especially temperature and precipitation play a key role for explaining the species richness patterns of all considered life forms and underpin the climatic dependence of lifeforms. Also for traits there was a clear finding: tree leaf size reduces with
elevation, and so did other leaf properties as well. All these patterns can be related to climatic adaptations, especially to frost and drought stress. Not surprisingly, elevational fern species richness showed along both study gradients a hump shaped richness pattern. This pattern was mainly driven by the epiphytic
lifeform, their elevational richness trend was more pronounced than for terrestrial species. This is especially due to the higher sensitivity to the steep change of environmental conditions and their exposed position especially in the upper tree layer. Comparing both gradients, the elevational peak of richness was situated at the northern gradient at a considerably lower elevation, and moreover, the absolute species richness was higher at the northern gradient. This last finding is especially counterintuitive on the first glance, as species richness should increase when coming closer to the equator. However, both results can be explained by the climatic setting
of the gradient: (1) Along the northern gradient, temperature levels are situated at lower elevations at the southern gradients, thus peak of richness is at about 16°C mean annual temperature at both gradients. (2) The higher temperature was assumed to lower the ambient humidity that the highest species occurred under the environmental condition with moderate temperature and high humidity. Additionally, due to the close vicinity of the southern gradient to the dry zones of Central Myanmar, species richness is generally lower here, showing the strong demand of humidity for maintenance and lifecycle of ferns. These first findings show the strong
environmental relationships of species richness of ferns and confirm what is known from other tropical mountains in the Americas and Africa (Chapter --> 2, 3).
(WP 2) Phylogenetic diversity pattern present a certain degree of variation in phylogenetic overdispersion and environmental clustering; however the overall pattern does not deviate from the random assemblage. Additional four gradients, i.e. Taiwan and Japan: Kyushu, Nishikoma and Hokkaido were added to the study to have a wide range of climate and geographic variations. Phylogenetic richness among the species presents similar pattern with species richness in general, however they are not identical, for instance the decreasing in phylogenetic richness in a species rich assemblage could be a consequence of the less phylogenetic distances between assigned species. In accordance with their reproductive behavior and long distance dispersal rate, the relationship between phylogenetic diversity and species richness might not be different from the
randomness in contrast with the angiosperms. In contrast with angiosperms from the previous studies, the phylogenetic diversity of pteridophytes reflects more to stochastic events as the deterministic processes limit the regional species pool in general, however for ferns, the geographic barrier such as dispersal limitation and,
seasonal and temporal barrier for fertilization disregards to a certain extent. Thus the equilibrium of competitive interaction and environmental filtering represent the species-neutral interactions and the pteridophytes dispersal might have been overriden the vicariance rate. However the trait patterns of some gradients such as
Natma Taung and Hponyinrazi from Myanmar, and Nishikoma from Japan have uncoupled with the phylogenetic pattern, thus the environmental divergence of those gradients are strong to lead the trait having clustered assemblage. Therefore, adaptations in ferns occur convergently in diverse phylogenetic assemblage (Chapter--> 5).
(WP 3) The best species richness model was developed with a set few climatic variables such as temperature and cloud cover. However the climate variables explain more than 60% of the local variation and combinations of small scale local factors could enhance the model prediction power. Putting the local results from Myanmar in a wider context, it is generally accepted that the total richness within each gradient should decline towards high latitudes as reported for most groups of organisms due to the general temperature driven nature of species richness. Second, for the same reason the frequently reported and for Myanmar confirmed hump shape pattern of species richness (mentioned above), the peak of richness should shift towards lower elevations for gradients at increasing latitude. Both trends combined should result in a pattern, where the symmetric hump in tropical regions gradually turned into a linear trend in temperate regions, which means the respective models were reduced in complexity. Relating species richness patterns to macroclimatic conditions along both gradients types, elevation and latitude, respectively, hypothesize main trends, which were tested on eight elevation
gradients between 4°S and 43.3°N. The combination of eight full elevational gradients in East Asia from Indonesia at the equator via Taiwan and Myanmar to temperate Japan, all sampled with the same standardized sampling protocol, offer the unique possibility to test this hypothesized interlinked trend. Additionally, by applying general additive models, the best fitting climatic variables could be assessed together with their relative contribution.
Moreover, since 'best fitting' does not necessarily mean 'good predicting', we applied the technique of leave-one-out-cross-validation (LOOCV) to find the best combination of climatic variables, which predicted the left out gradients. A reasonable result from these analyses can be extrapolated into the whole region covered by the analyzed gradients (East Asia).
As a result we could confirm the simple model of a shift in species richness trends from unimodal to linear towards temperate regions and suggest that over this broad spatial scale temperature is the most decisive factor. Confirming this, the highest predictability following our macroclimatic models was indeed temperature in
combination with cloud cover as humidity related variable, together predicting about 60% of species richness variation. It is interesting in this regard that explained variance is rather similar for the model based on generalized linear model result (GLM- model) and best model based on leave one out cross validation result
(LOOCV-model) (also somewhat lower for the latter ones), but that absolute numbers of species were more realistically modelled by LOOCV, since GLM predicted up to 380 species per plot, which is by far nonsense.
There are two interesting conclusions to be drawn: First, not precipitation as a humidity variable is included in the models, as most frequently used in species richness-climate models, but rather cloud cover. This indeed better reflects the physiological specifications of ferns, since not absolute water input, but balanced humidity is crucial. Second: In contrast to a large bulk of publications, which regularly report higher fits for richness–climate relationships, our best model is only able to predict a little bit more than half of the variation in local fern species richness in East and Southeast Asia. The remaining variation in local fern species richness is
most likely due to small scale factors, which are hardly accounted for by macroclimatic factors (Chapter 6).
The outcome of this study has contributed to comprehensive national conservation plans; (1) the preparation of the national biodiversity strategy and action plan, (2) the nomination dossier of UNESCO neutral heritage site in northern Myanmar.