What would current ecosystems be like without the impact of mankind? This question, which is critical for ecosystem management, has long remained unanswered due to a lack of present-day data from truly undisturbed ecosystems. Using mountaineering techniques, we accessed pristine relict ecosystems in the Peruvian Andes to provide this baseline data and compared it with the surrounding accessible and disturbed landscape. We show that natural ecosystems and human impact in the high Andes are radically different from preconceived ideas. Vegetation of these ‘lost worlds’ was dominated by plant species previously unknown to science that have become extinct in nearby human-affected ecosystems. Furthermore, natural vegetation had greater plant biomass with potentially as much as ten times more forest, but lower plant diversity. Contrary to our expectations, soils showed relatively little degradation when compared within a vegetation type, but differed mainly between forest and grassland ecosystems. At the landscape level, a presumed large-scale forest reduction resulted in a nowadays more acidic soilscape with higher carbon storage, partly ameliorating carbon loss through deforestation. Human impact in the high Andes, thus, had mixed effects on biodiversity, while soils and carbon stocks would have been mainly indirectly affected through a suggested large-scale vegetation change.

Aim Alpine ecosystems differ in area, macroenvironment and biogeographical history across the Earth, but the relationship between these factors and plant species richness is still unexplored. Here, we assess the global patterns of plant species richness in alpine ecosystems and their association with environmental, geographical and historical factors at regional and community scales. Location Global. Time period Data collected between 1923 and 2019. Major taxa studied Vascular plants. Methods We used a dataset representative of global alpine vegetation, consisting of 8,928 plots sampled within 26 ecoregions and six biogeographical realms, to estimate regional richness using sample‐based rarefaction and extrapolation. Then, we evaluated latitudinal patterns of regional and community richness with generalized additive models. Using environmental, geographical and historical predictors from global raster layers, we modelled regional and community richness in a mixed‐effect modelling framework. Results The latitudinal pattern of regional richness peaked around the equator and at mid‐latitudes, in response to current and past alpine area, isolation and the variation in soil pH among regions. At the community level, species richness peaked at mid‐latitudes of the Northern Hemisphere, despite a considerable within‐region variation. Community richness was related to macroclimate and historical predictors, with strong effects of other spatially structured factors. Main conclusions In contrast to the well‐known latitudinal diversity gradient, the alpine plant species richness of some temperate regions in Eurasia was comparable to that of hyperdiverse tropical ecosystems, such as the páramo. The species richness of these putative hotspot regions is explained mainly by the extent of alpine area and their glacial history, whereas community richness depends on local environmental factors. Our results highlight hotspots of species richness at mid‐latitudes, indicating that the diversity of alpine plants is linked to regional idiosyncrasies and to the historical prevalence of alpine ecosystems, rather than current macroclimatic gradients.

No abstract

We studied tree height in stands of high-Andean Polylepis forests in two cordilleras near Cuzco (Peru) with respect to variations in human impact and climatic conditions, and compared air and soil temperatures between qualitatively defined dry and humid slopes. We studied 46 forest plots of 100 m2 of five Polylepis species at 3560–4680 m. We measured diameter at breast height (dbh) and tree height in the stands (1229 trees in total), as well as air and soil temperatures in a subset of plots. The data was analyzed combining plots of given species from different sites at the same elevation (±100 m). There was no elevational decrease of mean maximum tree height across the entire data set. On humid slopes, tree height decreased continuously with elevation, whereas on dry slopes it peaked at middle elevations. With mean maximum tree heights of 9 m at 4530 m on the humid slopes and of 13 m at 4650 m on the dry slopes, we here document the tallest high-elevation forests found so far worldwide. These highest stands grow under cold mean growing season air temperatures (3.6 and 3.8°C on humid vs. dry slopes) and mean growing season soil temperatures (5.1 vs. 4.6°C). Mean annual air and soil temperature both decreased with elevation. Dry slopes had higher mean and maximum growing season air temperatures than humid slopes. Mean annual soil temperatures did not significantly differ and mean annual air temperatures only slightly differed between slopes. However, maximum air temperatures differed on average by 6.6 K between dry and humid slopes. This suggests that the differences in tree height between the two slopes are most likely due to differences in solar radiation as reflected by maximum air temperatures. Our study furthermore provides evidence that alpine Polylepis treelines grow under lower temperature conditions than global high-elevation treelines on average, suggesting that Polylepis species may have evolved special physiological adaptations to low temperatures.

Magnolia zenii is a critically endangered species known from only 18 trees that survive on Baohua Mountain in Jiangsu province, China. Little information is available regarding its molecular biology, with no genomic study performed on M. zenii until now. We determined the complete plastid genome of M. zenii and identified microsatellites. Whole sequence alignment and phylogenetic analysis using BI and ML methods were also conducted. The plastome of M. zenii was 160,048 bp long with 39.2% GC content and included a pair of inverted repeats (IRs) of 26,596 bp that separated a large single-copy (LSC) region of 88,098 bp and a small single-copy (SSC) region of 18,757 bp. One hundred thirty genes were identified, of which 79 were protein-coding genes, 37 were transfer RNAs, and eight were ribosomal RNAs. Thirty seven simple sequence repeats (SSRs) were also identified. Comparative analyses of genome structure and sequence data of closely-related species revealed five mutation hotspots, useful for future phylogenetic research. Magnolia zenii was placed as sister to M. biondii with strong support in all analyses. Overall, this study providing M. zenii genomic resources will be beneficial for the evolutionary study and phylogenetic reconstruction of Magnoliaceae.

Questions: Have millennia of human land use fundamentally altered the vegetation of a large proportion of the high Andean puna biome, with natural vegetation now restricted to inaccessible areas? Can inaccessible ledges be used as surrogates to infer the potential natural vegetation (PNV) in heavily impacted areas of the puna ecosystem of the high Andes? Is there a difference in plant community composition and diversity between the potential natural puna vegetation, represented by areas inaccessible to grazing and burning, and the anthropogenically disturbed vegetation found on nearby, but accessible, slopes?Location: Abra M alaga Private Conservation Area, Cusco, southern Peruvian Andes.Methods: Four study habitats were chosen that comprised ledges and slopes from within and outside of the conservation area. For each habitat, vegetation composition was recorded using eight to twelve 2 9 2-m 2 plots studied for species cover and abiotic variables.

Osmanthus (Oleaceae) is considered one of the most confusing genera with regards to circumscription and phylogenetic placement of taxa within the subtribe Oleinae of Oleaceae, with controversies mainly focused on the attribution of section Leiolea. In the present study, we analyzed 71 samples that represent the broad taxonomic, biogeographic, and morphological patterns in the subtribe Oleinae. Based on concatenation of four plastid genes (trnL-F, trnT-L, trnS-G, and matK) and comparison with the ITS region, Bayesian Inference, Maximum Likelihood and Maximum Parsimony phylogenies were inferred. Morphological character traits and geographical distributions of taxa were also studied. These results provide strong support for the segregation of the Leiolea clade (Osmanthus marginatus, O. matsumuranus and O. minor) of sect. Leiolea from Osmanthus. Therefore, the Leiolea clade is instated as the new genus Chengiodendron, with a taxonomic treatment provided. This study provides a clearer understanding of the phylogenetic relationships of genera in the subtribe Oleinae.

Human activity affects properties and development of ecosystems across the globe, to such a degree that it is now challenging to get baseline values for undisturbed ecosystems. This is especially true for soil development, which is potentially affected by land-use history and holds a legacy of past human interventions. Therefore, it is still largely unknown for most ecozones how soil would have developed ‘naturally’. Here, we show undisturbed soil development, i.e. the processes of weathering and accumulation of soil organic matter (SOM), by comparing pristine with grazed sites in the high Andes (4500 m) of southern Peru. We located study plots on a large ledge (0.2 km2) that is only accessible with mountaineering equipment. Plots with pristine vegetation were compared to rangeland plots that were presumably under relatively constant grazing management for at least four millennia. Vegetation change, induced by grazing management, led to lower vegetation cover of the soil, thereby increasing soil surface temperatures and soil acidification. Both factors increased weathering in rangeland soils. Formation of pedogenic oxides with high surface area explained preservation of SOM, with positive feedback to acidification. Higher contents of pyrophosphate extractable Fe and Al oxides indicated the importance of organo-mineral associations for SOM stabilization on rangeland sites, which are likely responsible for a higher degree of humification. This higher degree of humification induced melanization (darker colour) of the rangeland soils which, together with sparse vegetation cover, also feeds back to soil temperature. With this work, we present a conceptual framework of positive feedback links between human-induced vegetation change, soil development and accumulation of SOM, which is only possible due to the unique baseline values of a pristine ecosystem. Using ‘inaccessibility’ as a tool to quantify human impact in future interdisciplinary studies may push research forward on evaluating anthropogenic impact on Earth’s ecosystems.