Soil Biological Community and Chemical Characteristics of Evergreen and Deciduous Shrub Islands in an Alpine Shrub Meadow


 Aim Global change is driving shrub expansion in alpine and Arctic tundra, increasing the size and number of shrub islands. This phenomenon can directly affect the vegetation microclimate and indirectly affect biogeochemical cycling. However, it is unclear how shrub island formation, with a shift in dominant vegetation from graminoids to deciduous or evergreen shrub islands, affects the alpine biochemical cycling of carbon, nitrogen, phosphorus and decomposer diversity and activity. Methods Two dominant shrub species in the alpine shrub meadow in the eastern Tibetan Plateau were selected. Rhododendron lapponicum (RL) represents evergreen shrubs, and Sorbus rufopilosa (SR) represents deciduous shrubs. Soil fauna community diversity, soil microbial biomass, extracellular enzyme activity and enzyme stoichiometry of the organic soil layer (OL) and mineral soil layer (ML) were determined. Results Presence of shrub islands significantly affected nutrient cycling and microbial processes. Compared with the SR shrub island, the RL shrub island featured lower soil total nitrogen and microbial biomass carbon; lower total microbial phospholipid fatty acid, Gram-negative bacteria and total bacteria contents; and higher enzyme activities of β-glucosidase, cellobiohydrolase, β-N-acetylglucosaminidase, peroxidase and polyphenol oxidase. The OL response was greater than the ML response, especially in terms of enzyme activities. Conclusion Evergreen and deciduous shrub islands in an alpine shrub meadow had differences in soil biological communities and nutrient cycling, we suspect the OL was more sensitive than the ML to the shrub expansion in the alpine tundra.

Understanding the potential effects of these shrub islands on the soil biological 92 community and chemical characteristics is essential for predicting future alpine shrub 93 meadow functions. 94 The shrub island effect is widely recognized in many ecosystems, and this 95 phenomenon can directly affect the soil microclimate, soil nutrients and decomposers 96 (Blok et al., 2010; Ward et al., 2018). In winter, taller and richer shrubs can 97 accumulate deeper snow than less shrubby tundra areas, and snow, as an insulator, can 98 increase the soil temperature (i.e., the "snow-shrub hypothesis", Sturm et al., 2001). In 99 summer, the leaves of shrubs form a sheltered area that reduces the soil temperature  shrubs is relatively conservative, the relative growth rate is low, the litter quality is 152 poor (lower leaf N and P contents and higher lignin contents), and the production of  In this study, we propose the following scientific questions: What are the impacts of 163 different shrub islands on soil biological community biomass and diversity and on C, 164 N, and P cycling in alpine shrub meadow areas? What are the differences between 165 organic and mineral layers in the soil? Therefore, we investigated the organic and mineral soil layers in dominantly evergreen and deciduous shrub islands  This plant has a relatively conservative nutrient utilization and low relative growth 197 rate and the following nutrient compositions: leaf N 8.3 g kg -1 , leaf P 1.03 g kg -1 , 198 cellulose 191.8 g kg -1 , and lignin 353 g kg -1 . The organic layer is 4-5 cm. SR is a 199 deciduous shrub with a height of 2.7-4 m. It has odd pinnate compound leaves that are

Soil chemical analyses
All of the samples from the RL and SR shrub islands in each transect were 227 characterized in terms of physicochemical properties. The samples were dried in an 228 oven at 105°C for 48 h and weighed to calculate the measured soil moisture content.

229
The pH of the mineral layer and organic layer was measured by using ratios of soil to 230 deionized water of 1:2.5 and 1:5 (g/v). soil total C was determined by dichromate 231 oxidation-ferrous sulfate titration, total N content was determined using the 232 semi-micro-Kjeldahl method, and the total P content was determined by   Enzyme extraction and assays 268 We determined the activities of six extracellular enzymes associated with carbon,

296
The division of dominance of each type of group: the dominant group was the one in 297 which the number of individuals accounted for more than 10% of the total catch, 298 groups with 1%-10% were common, and groups with less than 1% were rare (Liu et   The shrub species had a significant effect on soil TN (P < 0.05), while the soil layer 334 had a very significant effect on soil organic C, TN, TP, C/N, and N/P (P < 0.01, Fig. 1).

335
The TN in the organic layer of the SR shrub island was the highest (mean value: SR: 336 22.40 g kg -1 , RL: 15.96 g kg -1 , and CK: 11.48 g kg -1 , followed by that in the RL shrub 337 island and the control site. There was no significant difference in the TC content 338 between the RL and SR shrub island soil layers, while the TN, TC and TP in the 339 organic layer soil of the control site were higher than those in the mineral layer. The 340 C/P of the organic layer in the RL shrub island was higher than that in the control group and SR shrub island (Fig. 1).

343
A total of 222 soil fauna individuals belonging to 1 phylum, 5 classes, 10 orders, and 344 33 families were collected from two soil layers in three shrub islands (Table 1) Simpson dominance index, Pielou evenness index, and Margalef richness index values did not differ significantly between the SR and RL soil layers (Table. 2) 371 Soil microbial community biomass 372 The shrub species and soil layer had significant effects on the bacteria content, G + and 373 Gproportions, and microbial biomass (P < 0.05, Fig. 2) The bacteria, G + , G -, and 374 total PLFA contents in the soil layers of the RL and SR shrub islands were higher than 375 those of the control site, and the bacteria, G + , G -, G + :Gand total PLFA values in the 376 soil layers of the SR shrub island were higher than those of the RL shrub island. The showed a downward trend from the organic layer to the mineral layer (Fig. 2).

382
Tree species and soil layer had extremely significant effects on MBC and MBN (P < 383 0.01, Fig. 4). The MBC and MBN in the organic layer under RL shrub island and SR 384 shrub island were higher than those in the control site, but there was no difference in 385 the mineral layer. The MBN and MBC in the organic layer of shrub species were more 386 abundant than those in the mineral layer and showed a downward trend (Fig. 3).

387
Enzyme activity and enzyme stoichiometry 388 The BG, CBH, NAG, and PPO activities in the soil layers in the RL shrub island were 389 higher than those in the SR shrub island, while the AP activity in the soil layers in the 390 SR shrub island (mean value: soil organic layer: 23.91; mineral soil layer: 4.30 µmol 391 h -1 g DOM -1 ) was higher than that in the RL shrub island (mean value: OL: 18.92 ML: 392 1.16 µmol h -1 g DOM -1 ). The PPO activity in the soil layers was the highest in the RL 393 shrub island and the lowest in the SR shrub island, while the POD activity in the 394 mineral layer was the highest in the control site and the lowest in the SR shrub island.

395
The AP, BG, LAP, and NAG activities in the organic layer under the RL and SR shrub 396 islands were higher than those in the mineral layer, but the PPO activity showed the 397 opposite pattern. In addition, the AP, BG, CBH and LAP activities in the organic layer 398 and mineral layer in the control site were not significantly different (Fig. 4).

399
The shrub species and soil layer had extremely significant effects on EEAN:P, EEAC:N, 400 EEAC:P, and vector angle (P < 0.01), and the interaction of shrub species and soil layer 401 had significant effects on EEAC:N and EEAC:P (P < 0.05, Fig. 6). EEAN:P in the organic 402 layer was the highest in the RL shrub island, followed by the SR shrub island, and the 403 lowest in the control site. The EEAN:P in the organic layer in the RL and SR shrub 404 islands was higher than that in the mineral layer, but there was no significant 405 difference in the control group (Fig. 5).  TC. The C/N ratio was negatively correlated with J, D, H, total PLFA, G + , TB and TF 420 (Fig. 7).  The RL shrub island had a higher C/P ratio, while the SR shrub island and the control 515 site had lower C/P ratios, showing that the RL shrub island had a higher carbon 516 sequestration capacity. The lower litter quality, the unique soil biocommunity 517 structure, and the lower microbial biomass in the RL shrub island make the material cycle slower, so nutrient conservation promotes carbon sequestration (Coleman, 1983).
Some studies have shown that Gbacteria are associated with simple C compounds 520 (alkyl), while G + bacteria are more associated with the more complex form C 521 (carbonyl) and that the G -/G + ratio has the potential to serve as a useful indicator of 522 the relative C availability of soil bacterial communities in organic soils and as a rough the G -/G + in the SR shrub island was higher than that in the RL shrub island, showing 525 that there were more abundant decomposable substrates or higher carbon availability 526 in the SR shrub island.

527
The enzyme stoichiometry, i.e., EEAC:P and EEAN:P, of the RL shrub island was higher 528 than that of SR shrub island, which showed that the microorganisms in the soil in the 529 RL shrub island were more limited by carbon and nitrogen than phosphorus. The