Smolt outmigration timing in Norway

Aim - Accurate predictions about transition timing of salmon smolts between freshwater and marine environments are key to effective management. We aimed to use available data on Atlantic salmon smolt migration to predict the emigration timing in rivers throughout Norway.

Location - In this study, we used data outmigration timing data of Atlantic salmon (Salmo salar) smolts from 41 rivers collected from 1984-2018 to make a predictive model for the timing of out-migrating salmon smolts along 12 degrees latitude.

Methods - Smolt migration data were collated from rivers where smolts are monitored with traps, video, and tagging and matched to river-specific metadata. Using a mixed effect generalized additive model, we tested for effects of spatial location, air temperature, river flow, and counting methods on the timing of 25% smolt emigration from rivers.

Results - After accounting for spatial effects and repeated measurements (across years and among rivers), air temperature and counting method were significant drivers of the estimated timing of smolt emigration. In-sample predictions yielded strong correlation with observed values, as did 10-fold cross-validation. Out-of-sample predictions suggested that the previous national estimates underestimated the migration timing in southern populations (linear model intercept = 39.73 days).

Conclusion - Model-derived estimates of run timing can be used to more accurately predict the timing of outmigration in order to better calibrate environmental flows and regulate management of marine industries such as aquaculture that may affect migration success at this particularly sensitive life stage.

Methods
Data Collection

Smolt migration data

The goal of our study was to collate available smolt migration data from Norway, in an attempt to make a predictive model of the timing of smolt migration in this area. Data were extracted from three sources. (1) published scientific articles, (2) Norwegian reports, and (3) unpublished data available from the authors’ research institutions. Data were updated from previous compilations using the same methodology (Ugedal et al. 2014). The database does not contain daily counts but summarizes the timing of smolt emigration by percentiles, recording the dates of 25% passage, 50% passage, and 75% passage (Table 1).

Smolt migration was monitored in 348 river years, comprising 47 rivers between 58.02 and 70.50 degrees latitude from 1984-2018 (Figure 1). Monitoring was conducted using different methods of observation: traps (N = 252 river years), video counting (N = 84 river years), and tagging (N = 12 river years). Note Eio and Vigda are the only two rivers that had multiple counting methods, so they are counted multiple times. The placement of these monitoring tools was not consistent among rivers nor was the timing of deployment standardized.

River morphology data

Morphological data from the river catchments was downloaded from NEVINA (http://nevina.nve.no/). This includes elevation data from the catchment, land composition (e.g. percent of catchment covered by agriculture, forest, lake, and urban areas), air temperature throughout the year (summer, winter, July, August temperatures). In addition, modelled average discharge, average rainfall, and average air temperature was extracted from each of the catchments from the same database.

Annual environmental data

Seasonal water temperature measurements data were not available for the majority of rivers. However, air temperature data were collected by monitoring stations throughout Norway, for which historical data are freely available through an API using the esd package in R (Benestad et al. 2009). We downloaded temperature records from all available stations. A generalized additive model was fit to each individual year to explain the air temperature recorded by the longitude, latitude, and altitude of the station using the gam function in the mgcv package (Wood 2017). The gam models were then carried forward to predict the air temperatures for each river in each year using coordinates of the rivermouth, and average grade of the river using the predict.gam function. Air temperature for each river in each year was summarized as the average between January 1 and March 31.

Modelled water discharge data were available for each river from the NorKyst800 model.

The Norwegian river discharges were modelled by the NVE (Norwegian Water Resources and Energy Directorate) using a distributed version of the HBV-model with 1 km horizontal resolution (Beldring et al. 2003 and Huang et al., 2019). We summarised water discharge for each river in each year by extracting the first day of the year when the flow first hit 10 and 25 % of the maximum flow from March 1 to July 18, which was considered to be the maximal likely window for onset of smolt migration. We used different temporal windows for temperature and flow because temperature should control physiological readiness (proximate cause) and flow should drive the exact timing of migration (ultimate cause). Both extractions gave very similar model outputs.

研究背景与目标——精准预测大西洋鲑（Salmo salar）降海幼体（salmon smolts）在淡水与海洋环境间的迁移时间，是开展有效渔业管理的核心前提。本研究旨在利用已公开的大西洋鲑降海幼体迁移数据，构建挪威全境河流中鲑鱼幼体入海迁移时间的预测模型。

研究区域——本研究使用了1984年至2018年间采集自41条河流的大西洋鲑降海幼体入海迁移时间数据，以构建覆盖12个纬度跨度区域的鲑鱼幼体入海迁移时间预测模型。

研究方法——本研究从采用诱捕笼、视频计数与标记法监测降海幼体迁移的河流中整理得到幼体迁移数据，并匹配各河流专属的元数据。本研究采用混合效应广义可加模型（Generalized Additive Mixed Model, GAMM），检验空间位置、气温、河流流量与计数方法对河流中25%降海幼体完成迁移的时间的影响。

研究结果——在控制空间效应与重复测量（跨年度及跨河流）的影响后，气温与计数方法是影响降海幼体迁移时间估算结果的显著驱动因子。样本内预测结果与观测值呈现极强的相关性，十折交叉验证（10-fold cross-validation）结果同样如此。样本外预测结果显示，此前的全国性估算低估了南部种群的迁移时间（线性模型截距=39.73天）。

研究结论——基于模型得到的种群洄游时间估算结果，可用于更精准地预测鲑鱼幼体入海迁移时间，从而更好地调控环境流量，并对水产养殖等可能影响该敏感生命阶段迁移成功率的海洋产业开展规范化管理。

研究方法
数据收集
1. 降海幼体迁移数据
本研究的核心目标为整理挪威境内已公开的大西洋鲑降海幼体迁移数据，以构建该区域内幼体迁移时间的预测模型。数据来源分为三类：(1) 已发表的学术论文；(2) 挪威官方报告；(3) 作者所在研究机构公开的未发表数据。本研究沿用Ugedal等人2014年的研究方法，对既往汇编的数据进行了更新。本数据库未收录单日计数数据，而是以百分位数汇总幼体入海迁移时间，记录了25%、50%及75%个体完成迁移的日期（表1）。

1984年至2018年间，研究团队共对47条纬度介于58.02°至70.50°的河流开展了348个河年度的幼体迁移监测（图1）。监测采用了多种观测手段：诱捕笼法（252个河年度）、视频计数法（84个河年度）以及标记法（12个河年度）。需说明的是，仅埃约河（Eio）与维格达河（Vigda）两条河流采用了多种计数方法，因此被重复计入统计。各河流的监测工具布设位置并不统一，监测部署时间也未实现标准化。

2. 河流形态学数据
本研究从NEVINA数据库（http://nevina.nve.no/）下载了各河流流域的形态学数据，包括流域海拔数据、土地利用构成（如流域内农业、森林、湖泊与城镇用地占比）以及全年气温数据（夏季、冬季、7月及8月气温）。此外，本研究还从同一数据库中提取了各流域的模拟平均径流量、平均降雨量与平均气温数据。

3. 年度环境数据
多数河流未记录季节性水温数据。不过，挪威全境的监测站点均采集了气温数据，其历史数据可通过R语言的esd包（Benestad等，2009）提供的应用程序编程接口（API）免费获取。本研究下载了所有可用站点的气温记录，针对每一年份分别拟合广义可加模型（Generalized Additive Model, GAM），利用mgcv包中的gam函数，以监测站点的经度、纬度与海拔为自变量解释实测气温。随后，本研究利用河口坐标与河流平均坡度，通过predict.gam函数，将上述GAM模型外推至每一条河流的每一年份，以预测其气温。本研究将每条河流每年的气温汇总为1月1日至3月31日的平均气温。

模拟径流量数据可通过NorKyst800模型获取。挪威河流径流量由挪威水资源与能源管理局（NVE, Norwegian Water Resources and Energy Directorate）采用水平分辨率为1km的分布式HBV模型模拟得到（Beldring等，2003；Huang等，2019）。本研究针对每条河流的每一年份，提取3月1日至7月18日期间径流量首次达到最大流量10%与25%的日期，以此汇总径流量数据，该时段被认为是降海幼体迁移启动的最大可能窗口。本研究为气温与径流量设置了不同的时间窗口，原因在于气温调控幼体的生理准备状态（近因），而径流量则决定迁移的具体时间（远因）。两种数据提取方式得到的模型输出结果高度相似。