Microgeographic adaptation corresponds with elevational distributions of congeneric montane grasshoppers

Local adaptation can occur at small spatial scales relative to the dispersal capacity of species. Alpine ecosystems have sharp environmental clines that offer an opportunity to investigate the effects of fine scale shifts in species’ niche breadth on adaptive genetic processes. Here we examine two grasshopper species endemic to the Australian Alps (Kosciuscola spp.) that differ in elevational niche breadth; one broader, K. usitatus (1400-2200m), and one narrower, K. tristis (1600-2000m). We examine signatures of selection with respect to environmental and morphological variables in two mountain regions using FST outlier tests and environmental association analyses (EAA) applied to Single Nucleotide Polymorphism data (K. usitatus: 9,017 SNPs, n = 130; K. tristis: 7,363 SNPs, n = 135). Stronger genetic structure was found in the more narrowly distributed K. tristis, which showed almost twice the number of SNPs under putative selection (10.8%) compared with K. usitatus (5.3%). When examining SNPs in common across species (n = 3,058), 260 SNPs (8.5%) were outliers shared across species, and these were mostly associated with elevation, a proxy for temperature, suggesting parallel adaptive processes in response to climatic drivers. Additive polygenic scores (an estimate of the cumulative signal of selection across all candidate loci) were non-linearly and positively correlated with elevation in both species. However, a steeper correlation in K. tristis indicated a stronger signal of spatially varying selection towards higher elevations. Our study illustrates that the niche breadth of co-occurring and related species distributed along the same environmental cline is associated with differences in patterns of microgeographic adaptation.

Methods
Sampling was undertaken for two weeks in late February to early March of 2017. Three transects ranging from 1400-2200m in elevation were sampled within each of two mountain regions, Thredbo and Guthega within Kosciuszko National Park in New South Wales. DNA was extracted from the tissue of one femur per collected specimen at the Diversity Arrays Technology sequencing facility, Canberra (DArTseq). A modified double-digest RAD sequencing protocol was used to generate data at DArTseq. SNPs were further filtered using the Radiator r package.

Usage Notes
Files: K.usitatus_9017_SNPs.ped, K.usitatus_9017_SNPs.map

Genotype information of 130 individuals of Kosciuscola usitatus analysed at 9,017 SNPs. The data was used to perform selection detection tests. The files are in PED and MAP format. MAP file= list of names of SNP markers

File: K.usitatus_3321_SNPs.str

File with 3,321 SNPs and 130 individuals, used to analyse pairwise Fst, genetic structure, isolation by distance and spatial autocorrelation analyses for Kosciuscola usitatus. The file is STRUCTURE formatted. Missing values are coded as -9

File: K.usitatus_envt_data.csv

Environmental data information used to detect signatures of selection (EAA analysis, pRDA and RDA).

pop= population information representing the area of collection for each individual

transect= transect information for each individual

location= Location

individual= unique name of individual

elevation= elevation in meter above sea level

FPC= Foliage Projective Cover (%)

precipseas = precipitation seasonality

roughness = terrain roughness

FZ = number of frost days

bd = soil bulk density

ph = soil ph     

femur length = measured in mm

sex = sex per individual, binary coded: male = 0, female = 1

Files: K.tristis_7363_SNPS.ped, K.tristis_7363_SNPS.map

Genotype information of a total of 135 individuals of Kosciuscola usitatus analysed at 7,363 SNPs. The data was used to perform selection detection tests. The files are in PED and MAP format. MAP file= list of names of SNP markers

File: K.tristis_3566_SNPS.str

File with 3,566 SNPs and 135 individuals, used to analyse pairwise Fst, genetic structure, isolation by distance and spatial autocorrelation analyses for Kosciuscola tristis. The file is STRUCTURE formatted. Missing values are coded as -9

File: K.tristis_envt_data

Environmental data information used to detect signatures of selection (EAA analysis, pRDA and RDA).

pop= population information representing the area of collection for each individual

transect= transect information for each individual

location= Location

individual= unique name of individual

elevation= elevation in meter above sea level

FPC= Foliage Projective Cover (%)

precipseas = precipitation seasonality

roughness = terrain roughness

FZ = number of frost days

bd = soil bulk density

ph = soil ph     

femur length = measured in mm

sex = sex per individual, binary coded: male = 0, female = 1

File: Whole_data_3058_SNPs.ped, Whole_data_3058_SNPs.map

Genotype information of 266 individuals (129 K. usitatus, 137 K. tristis) analysed at 3,058 SNPs. The data was used to perform selection detection tests. The files are in PED and MAP format. MAP file= list of names of SNP markers

File: Whole_data_2143_SNPs.str

File with 2,143 SNPs and 266 individuals, used to perform genetic structure analysis for the common dataset for the two species. The file is STRUCTURE formatted. Missing values are coded as -9

局域适应可发生于相较于物种扩散能力而言的极小空间尺度范围内。高山生态系统存在陡峭的环境梯度，为探究物种生态位宽度的精细尺度变化对适应性遗传过程的影响提供了绝佳研究契机。本研究针对澳大利亚阿尔卑斯山区特有的两种蚱蜢物种（Kosciuscola属，Kosciuscola spp.）展开，二者的海拔生态位宽度存在显著差异：其一为生态位较宽的K. usitatus（分布海拔1400~2200米），其二为生态位较窄的K. tristis（分布海拔1600~2000米）。本研究在两个山区区域内，针对环境与形态变量开展选择信号分析，具体采用FST离群值检测（FST outlier tests）以及环境关联分析（environmental association analyses，EAA），并将其应用于单核苷酸多态性（Single Nucleotide Polymorphism，SNP）数据：K. usitatus共包含9017个SNP位点，样本量n=130；K. tristis共包含7363个SNP位点，样本量n=135。

分布范围更窄的K. tristis呈现出更强的遗传结构，其检测到的潜在选择作用SNP位点占比（10.8%）几乎是K. usitatus（5.3%）的两倍。在跨物种共享的SNP位点（共3058个）中，有260个位点（占比8.5%）为两个物种共有的离群位点，这些位点大多与海拔相关——海拔可作为温度的替代指标，这表明二者在气候驱动下存在平行适应性进化过程。

加性多基因评分（即对所有候选位点的累积选择信号进行估算的指标）在两个物种中均与海拔呈非线性正相关。但K. tristis的相关曲线斜率更陡，表明其在高海拔区域受到的空间异质选择信号更强。本研究表明，沿同一环境梯度分布的同域近缘物种，其生态位宽度与微观地理适应模式的差异存在显著关联。

### 研究方法
采样工作于2017年2月末至3月初开展，历时两周。在新南威尔士州科修斯科国家公园内的两个山区区域（Thredbo与Guthega）各设置三条海拔跨度为1400~2200米的样带进行采样。研究人员在堪培拉的多样性阵列技术测序平台（Diversity Arrays Technology sequencing facility，DArTseq）从每份采集标本的单条股骨组织中提取DNA，采用改良的双酶切RAD测序方案在DArTseq平台生成测序数据，后续通过Radiator R包对SNP位点进行进一步过滤。

#### 使用说明
1. **文件：K.usitatus_9017_SNPs.ped、K.usitatus_9017_SNPs.map**
该文件包含130个Kosciuscola usitatus个体在9017个SNP位点上的基因型信息，用于开展选择检测分析。文件格式为PED与MAP格式，其中MAP文件为SNP标记名称列表。
2. **文件：K.usitatus_3321_SNPs.str**
该文件包含130个个体的3321个SNP位点数据，用于开展Kosciuscola usitatus的成对FST分析、遗传结构分析、距离隔离分析以及空间自相关分析。文件采用STRUCTURE格式，缺失值编码为-9。
3. **文件：K.usitatus_envt_data.csv**
该文件包含用于选择信号检测的环境数据（适用于EAA分析、pRDA与RDA分析），各字段说明如下：
- pop：种群信息，代表每个个体的采样区域
- transect：个体所属的样带信息
- location：采样位置
- individual：个体唯一标识名称
- elevation：海拔高度（单位：米，相对于海平面）
- FPC：植被投影盖度（Foliage Projective Cover，百分比）
- precipseas：降水季节性
- roughness：地形粗糙度
- FZ：霜冻日数
- bd：土壤容重
- ph：土壤pH值
- femur length：股骨长度（单位：毫米）
- sex：个体性别，采用二进制编码：雄性=0，雌性=1
4. **文件：K.tristis_7363_SNPS.ped、K.tristis_7363_SNPS.map**
该文件包含135个Kosciuscola tristis个体在7363个SNP位点上的基因型信息，用于开展选择检测分析。文件格式为PED与MAP格式，其中MAP文件为SNP标记名称列表。
5. **文件：K.tristis_3566_SNPS.str**
该文件包含135个个体的3566个SNP位点数据，用于开展Kosciuscola tristis的成对FST分析、遗传结构分析、距离隔离分析以及空间自相关分析。文件采用STRUCTURE格式，缺失值编码为-9。
6. **文件：K.tristis_envt_data**
该文件包含用于选择信号检测的环境数据（适用于EAA分析、pRDA与RDA分析），各字段说明与K.usitatus_envt_data.csv一致。
7. **文件：Whole_data_3058_SNPs.ped、Whole_data_3058_SNPs.map**
该文件包含266个个体（其中129个为K. usitatus，137个为K. tristis）在3058个SNP位点上的基因型信息，用于开展选择检测分析。文件格式为PED与MAP格式，其中MAP文件为SNP标记名称列表。
8. **文件：Whole_data_2143_SNPs.str**
该文件包含266个个体的2143个SNP位点数据，用于开展两个物种共享数据集的遗传结构分析。文件采用STRUCTURE格式，缺失值编码为-9。