In kun-ecology/SER: Short-period Environmental Regime

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(kableExtra)
library(knitr)
library(SER)
load("SER_tutorial.rda") 

This tutorial is about how to calculate short-period environmental regime (SER) with a R package SER. It provides basic information of how short-period environmental regime is developed and how it can be used for hydrological data as an example. It includes following parts:

• Overview: introduction about how these indices are constructed.
• Main arguments of the SER function: detail information about three arguments of the main function SER.
• Example: an example that shows how SER apply to hydrological data.

Overview

Environment regime (or environment pattern), unlike traditional environmental variables, is defined as the environmental dynamic characteristics during a given (either long or short) period, such as change rate, distribution and extreme events, which might be masked by using simple average or median values. As linking environmental changes to biotic responses is one of the central topics in ecology, incorporating environment regimes into ecological studies can greatly advance our understanding between biota-environment relations. For example, short-period flow regime indices (e.g., skewness, change rate in flow events) played a vital role in riverine algal community (Qu et al. 2019; Wu et al. 2018, 2019)

With the facilitation of science and technology, more and more high frequently (daily or even hourly) measured environmental variables (e.g., nutrient concentration, dissolved oxygen) are available nowadays. These high frequency measured data allow computing environment regime indices. However, an R package or function that can be easily used to calculate short-term environment regime is still missing.

Therefore, we developed the so called short-period environmental regime (SER) to characterize such frequently measured environmental data and make them suitable to link with spot shot biotic data. Such regime is inspired by Olden and Poff (2003), which developed a series of indices to identify the characteristics of flow variation in riverine study. In total, 11 elementary indices that concentrate more about environmental variation in short-period were developed (Table 1). It should be noted that short-period defined here is case specific. For example, for some study, changes in days are the interested period, while for other, months/years may be the targeted time interval.

Table 1. Detail information of the 11 elementary indices

kable(indice_info) %>% 
  kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"),full_width = T,position = "center") %>% 
  #row_spec(c(1,6,11),bold = T) %>% # specifying formate of the 3 aspects of flow events
  row_spec(0, bold = T,font_size = 16) %>% # specifying the format of header
  pack_rows("Magnitude of environmental data",1,4) %>% 
  pack_rows("Frequency of environmental data",5,8) %>% 
  pack_rows("Rate of change of environmental data",9,11) %>% 
  footnote(symbol =c("Units of some indices are dependent on types of environmental data.","Currently, the function is developed with regarding days as the interested period, more features are under development.",
  "N can be defined as any interested time period, by default N=BetwSamT (days between two successive sampling dates).") )

Main arguments of the SER function

• env_date: a vector containing dates that a environmental variables were measured, make sure it is in date format before passing it to the function.
• env_val: a vector containing environmental values for corresponding dates in env_date, make sure it is in numeric format.
• sample_date: sampling dates, a vector containing dates for biological sampling. Date for initiation of the experiment (measurements) should be included as the first element of the vector.
• days_bf:a numeric vector representing N days (months or years) before the sampling date; if days_bf = NULL (by default), then days_bf = days between two successive sampling dates.
• type: a character vector indicating whether the 'days_bf' is "day", "month", or "year" specific.
• include_sample.date: TRUE or FALSE (by default) indicates whether environmental data at sample.date is included or not during the calculation.
• include_successivee: TRUE or FALSE (by default) indicates whether environmental data between two successive sampling dates is used for calculating 'BetwSamT' environmental regime.
• simplifye: TRUE (by default) or FALSE indicates whether the simplified result is returned. Set simplify = TRUE if you want to check the intermediate results.

Examples

Below is an example showing how short-period environmental regime can be used for discharge data. Short-period environmental regime (in this case it is short-period hydrological regime) in days between two successive sampling dates (Table 2) were developed based on the 11 elementary indices. In total, 11 short-period hydrological indices were developed. The indices' names were constructed as the combination of short time period and names of elementary indices, for example, BetwSamT.MA1 and BetwSamT.RC stand for mean of the daily average flow and mean rate of change in days between two successive sampling days respectively.

Table 2. Short-period hydrological indices

kable(hydro_info) %>% 
  kable_styling(bootstrap_options = c("striped", "hover", "condensed", "responsive"),full_width = T,position = "center") %>% 
  #row_spec(c(1,6,11),bold = T) %>% # specifying formate of the 3 aspects of flow events
  row_spec(0, bold = T,font_size = 16) %>% # specifying the format of header
  pack_rows("Magnitude of flow events",1,4) %>% 
  pack_rows("Frequency of flow events",5,8) %>% 
  pack_rows("Rate of change of flow events",9,11)

It should be noted that the time period for estimating high/low flow pulse is 2016-2018, which is the time period interested in the unpublished work. When applying function with other types of data, such time period is automatically defined as the period of the input data env_date.

There are two data sets:

• hydro_df is a dataframe, i.e., discharge data for three streams for the period of 2016-2018. There are two columns, i.e., Date (should be in date format) and Discharge (should be numeric data; unit: L/s). Below is the structure of the data.

str(hydro_df)

• sample_date is a vector that contains 13 date elements, first of which is the initiate date of the experiment while the rest 12 are the sampling dates.

library(SER)
str(sample_date)

Calculation of SER with discharge data.

hydro_ser <- SER(hydro_df$Date, hydro_df$Discharge, sample_date,days_bf = NULL)
str(hydro_ser)
head(hydro_ser)

Summary

• In short, by summarizing time-series environmental variables into indices covering different facets i.e., magnitude, frequency, and rate of change of time-series data in any focused period, which are typically masked by using simple average or median values in conventional way, SER helps linking time-series environmental data to discrete biological sampling. Here, we only used stream discharge data to estimate short-period hydrological regime, but other types of environmental data e.g., nutrient concentration, light, dissolved oxygen and etc can also be used to develop case-specific environmental regime. Besides, two studies (Guo et al. 2020, 2021; Wu et al. 2022) using SER have been published, you may check them for how to interpret these calculated indices and how to include them in following-up analyses.

References

Olden, Julian D, and NL Poff. 2003. “Redundancy and the Choice of Hydrologic Indices for Characterizing Streamflow Regimes.” River Research and Applications 19 (2): 101–21.

Yueming Qu, Naicheng Wu, Björn Guse, Kristė Makarevičiūtė, Xiuming Sun, and Nicola Fohrer. 2019. “Riverine Phytoplankton Functional Groups Response to Multiple Stressors Variously Depending on Hydrological Periods.” Ecological Indicators 101: 41–49.

Naicheng Wu, Yueming Qu, Björn Guse, Kristė Makarevičiūtė, Szewing To, Tenna Riis, and Nicola Fohrer. 2018. “Hydrological and Environmental Variables Outperform Spatial Factors in Structuring Species, Trait Composition, and Beta Diversity of Pelagic Algae.” Ecology and Evolution 8 (5): 2947–61.

Naicheng Wu, Hans Thodsen, Hans Estrup Andersen, Henrik Tornbjerg, Annette Baattrup-Pedersen, and Tenna Riis. 2019. “Flow Regimes Filter Species Traits of Benthic Diatom Communities and Modify the Functional Features of Lowland Streams-a Nationwide Scale Study.” Science of the Total Environment 651: 357–66.

Kun Guo, Naicheng Wu, Paraskevi Manolaki, Annette Baattrup-Pedersen, and Tenna Riis. "Short-period hydrological regimes override physico-chemical variables in shaping stream diatom traits, biomass and biofilm community functions." Science of The Total Environment 743 (2020): 140720.

Kun Guo, Naicheng Wu, Wei Li, Annette Baattrup-Pedersen, and Tenna Riis. "Microbial biofilm community dynamics in five lowland streams." Science of The Total Environment 798 (2021): 149169.

Naicheng Wu, Yaochun Wang, Yixia Wang, Xiuming Sun, Claas Faber, and Nicola Fohrer. "Environment regimes play an important role in structuring trait‐and taxonomy‐based temporal beta diversity of riverine diatoms." Journal of Ecology 110, no. 6 (2022): 1442-1454.

kun-ecology/SER documentation built on Dec. 2, 2023, 10:13 p.m.