Bacterial community of cushion plant Thylacospermum ceaspitosum on elevational gradient in the Himalayan cold desert

Abstract

Although bacterial assemblages are important components of soils in arid ecosystems, the knowledge about composition, life-strategies, and environmental drivers is still fragmentary, especially in remote high-elevation mountains. We compared the quality and quantity of heterotrophic bacterial assemblages between the rhizosphere of the dominant cushion-forming plant Thylacospermum ceaspitosum and its surrounding bulk soil in two mountain ranges (East Karakoram: 4850-5250 m and Little Tibet: 5350-5850 m), in communities from cold steppes to the subnival zone in Ladakh, arid Trans-Himalaya, northwest India. Bacterial communities were characterized by molecular fingerprinting in combination with culture-dependent methods. The effects of environmental factors (elevation, mountain range, and soil physico-chemical parameters) on the bacterial community composition and structure were tested by multivariate redundancy analysis and conditional inference trees. Actinobacteria dominate the cultivable part of community and represent a major bacterial lineage of cold desert soils. The most abundant genera were Streptomyces, Arthrobacter, and Paenibacillus, representing both r- and K-strategists. The soil texture is the most important factor for the community structure and the total bacteria counts. Less abundant and diverse assemblages are found in East Karakoram with coarser soils derived from leucogranite bedrock, while more diverse assemblages in Little Tibet are associated with finer soils derived from easily weathering gneisses. Cushion rhizosphere is in general less diverse than bulk soil, and contains more r-strategists. K-strategists are more associated with the extremes of the gradient, with drought at lowest elevations (4850-5000 m) and frost at the highest elevations (5750-5850 m). The present study illuminates the composition of soil bacterial assemblages in relation to the cushion plant T. ceaspitosum in a xeric environment and brings important information about heterotrophic bacteria in Himalayan soil.

Keywords: Himalayas; Ladakh; heterotrophic microbial community; life strategy; mountains; subnival soil.

Figure 1

db-RDA analyses of the bacterial dissimilarity matrices (Bray–Curtis) and vector-fitting of the environmental variables calculated separately for Nubra. Communities were grouped to the centroid by elevation and cushion interaction.

Figure 2

db-RDA analyses of the bacterial dissimilarity matrices (Bray–Curtis) and vector-fitting of the environmental variables calculated separately for Tso Moriri transects. Communities were grouped to the centroid by elevation and cushion interaction.

Figure 3

Environmental factors predicting variation in bacterial assemblages in Nubra and Tso Moriri. We identified the differences with multivariate regression tree which hierarchically splits bacterial dissimilarity matrix into more homogenous subsets according to the elevation, cushion, and soil physico-chemical parameters. The dissimilarity matrix was calculated from SSCP profiles data with Bray–Curtis index and the tree was pruned according the 1-SE rule.

Figure 4

Characteristics of heterotrophic bacterial communities at two localities Nubra and Tso Moriri in all studied elevations. Total count of bacteria^*10⁸/g DW of soil; CFU, culturable cell population density after 8 days of cultivation^*10⁶/g DW of soil; C/T, percentage of CFU from total bacterial counts; r-strateg—percentage of bacteria that produce visible colonies within 72 h.

Figure 5

Conditional inference trees testing the effect of cushion, elevation, and soil physico-chemical parameters on the total bacterial counts. Hierarchical splitting of the data was based on selecting predictors that best distinguish a variable's responses and dividing samples into two groups according to the predictor's splitting value. Significant predictors are shown in oval windows, while the number of releves in each group (n) and their mean species richness (y) in green boxes.

Figure 6

The conditional inference tree analysis of CFU (for details see Figure 4).

Figure 7

The conditional inference tree analysis of proportion of cultivable bacteria to total bacterial count (for details see Figure 4).

Figure 8

The conditional inference tree analysis of r-strategist (for details see Figure 4).

Figure 9

The course of the mean daily temperature at 5 cm aboveground at the lowest studied elevation (Nubra 4850 m) and the highest (Tso Moriri 5850 m) during years 2008–11.

Figure 10

Physico-chemical properties (Ca²⁺, Mg⁺, K+, organic matter⁻₃, PO³⁻₄, pH) of bulk and rhizosphere soil at two localities Nubra and Tso Moriri in all studied elevations.