Biodiversity and Crop Yield in Sustainable Agriculture Systems

ENVX1002 | Project 1

Published

March 22, 2024

Introduction

As the global population continues to grow, sustainable agriculture is essential for food security and safeguarding the environment (Frison 2011). In 2021, a comprehensive data set was compiled, on sustainable farming and its impact on biodiversity and agricultural yield (Jones et al 2021). For our analysis, Corn is chosen as the focus crop. The variables selected include both categorical (nominal) factors including various farming practices, and numerical (continuous) variables, specifically biodiversity abundance and crop yield. As a wildlife conservation student, i am drawn to this data set due to its relevance to my field of study. By selecting this data set, I hope to refine my data analysis skills, which will be valuable to my career.

Data Analysis

Corn

The global data set contains 4076 observations sourced from 237 articles, 55 observations specifically focused on corn.

Code
library(readxl)
Biodiversity <- read_excel("ENVX1002_Biodiversity_farms.xlsx", sheet = "Analysis")
Corn_Data <- Biodiversity[grepl("corn", Biodiversity$crops_all_common_C, ignore.case = TRUE), ]
length (Corn_Data$crops_all_common_C) #Number of observations of corn 
[1] 55

Comparator and Intervention

In this study, the ‘comparator’ variable, refers to the traditional farming methods and plots that typically have less natural habitat integrated into them. On the other hand, the ‘intervention’ variable refers to diverse farming techniques that integrate more flora and fauna, such as agroforestry.

Code
comparator_categories <- (Corn_Data$System_C)
corn_data_subset <- data.frame(
  comparator_categories = Corn_Data$System_C,
  crop = rep("Corn", nrow(Corn_Data))
)

library(knitr)
corn_table <- table(corn_data_subset$comparator_categories)
corn_table_df <- as.data.frame(corn_table)
colnames(corn_table_df) <- c("Comparator Categories", "Number of observations")
kable(corn_table_df, caption = "Table 1. Observations of Corn in Different Comparator Categories")
Table 1. Observations of Corn in Different Comparator Categories
Comparator Categories Number of observations
Monoculture 47
Simplified other 8
Code
intervention_C <- (Corn_Data$System_I)
intervention_dataset <- data.frame(intervention_C = Corn_Data$System_I)
crop2 = rep("Corn", nrow(Corn_Data))
intervention_table <- table(intervention_dataset$intervention_C)
intervention_DF <- as.data.frame(intervention_table)
colnames(intervention_DF) <- c("Intervention Categories", "Number of observations")
kable(intervention_DF, caption = "Table 2. Observations of Corn in Different Intervention Categories")
Table 2. Observations of Corn in Different Intervention Categories
Intervention Categories Number of observations
Associated plant species 34
Combined practices 8
Crop rotation 4
Intercropping 9

Biodiversity

Biodiversity was measured using a range of metrics, specifically abundance and species richness. The average values of biodiversity are represented in figures 1 and 2. The data in comparator systems appears to be positively skewed and demonstrates a wide spread ranging from 0.86 to 140.3. The central tendency is marked by a median of 8.9, which is lower than the mean of 20.3. Contrastingly, the intervention systems distribution appears to be bimodal, with peaks observed at low values and again at 80-100. Thee data displays a broad range of values, ranging from 0.08 to 116.3, central tendency of 57.65 which was higher than the mean of 46.

Biodiversity in Comparator Systems
Code
Biodiversity_Comparator <- Corn_Data$B_value_C
summary(Biodiversity_Comparator)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
  0.8644   2.1400   8.9238  20.3031  27.4000 140.3000 
Code
hist(Biodiversity_Comparator, 
     main = "Figure 1. Histogram of Biodiversity mean in Comparator systems", 
     xlab = "Biodiversity Mean Value", 
     ylab = "Number of observations ",
     col = "mediumseagreen", 
     border = "black")

Biodiversity in Intervention Systems
Code
Biodiversity_Intervention <- Corn_Data$B_value_T 
summary(Biodiversity_Intervention)
     Min.   1st Qu.    Median      Mean   3rd Qu.      Max. 
  0.07734   2.82708  57.65000  45.97239  83.27000 116.30000 
Code
hist(Biodiversity_Intervention, 
     main = "Figure 2. Histogram of Biodiversity mean in Intervention systems", 
     xlab = "Biodiversity Mean Value", 
     ylab = "Number of observations ",
     col = "mediumseagreen", 
     border = "black")

Yield

The yield data in the comparator system demonstrated a wide spread, with values ranging from 7161 to 14968 kg/ha. The central tendency is represented by the median of 12065, which is higher than the mean of 11277 and indicates a slightly right-skewed distribution. In contrast, the yield data in the intervention system displays a narrower spread ranging from 14170 to 16065. Both the median and mean are 15387, indicating a symmetrical distribution (Figure 3).

Yield in Comparator Systems
Code
#Remove NA from the dataset and convert the values to numbers
Yield_C <- Corn_Data$Yield_value_C
Yield_C[Yield_C == "NA"] <- NA
Yield_C<- na.omit(Yield_C)
Yield_C <- as.numeric(Yield_C)
summary(Yield_C)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
   7161    7613   12065   11277   14581   14968 
Yield in Intervention Systems
Code
#Remove NA from the dataset and convert the values to numbers
Yield_I <- Corn_Data$Yield_value_T
Yield_I[Yield_I == "NA"] <- NA
Yield_I<- na.omit(Yield_I)
Yield_I <- as.numeric(Yield_I)
summary(Yield_I)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  14710   14710   15387   15387   16065   16065 
Code
# Load necessary library
library(ggplot2)
# Creating a dataframe
yield_data <- data.frame(System = c(rep("Yield_C", length(Yield_C)), rep("Yield_I", length(Yield_I))),
  Yield = c(Yield_C, Yield_I)
)

ggplot(yield_data, aes(x = System, y = Yield, fill = System)) +
  geom_boxplot() +
  scale_fill_manual(values = c("Yield_C" = "lightgreen", "Yield_I" = "mediumseagreen")) +
  labs(x = "System", y = "Yield (Kg/ha)", title = "Yield Comparison Between Comparator and Intervention Systems")

Biodiversity and Yield

This scatter plot appears to have a slightly positive trend between biodiversity mean value and yield, however the relationship is not linear.

Code
library(ggplot2)
Corn_Data$Yield_value_C <- as.numeric(as.character(Corn_Data$Yield_value_C)) #Remove NA from the dataset and convert the values to numbers
ggplot(data = Corn_Data, aes(x = B_value_C, y = Yield_value_C)) +
  geom_point(color = "mediumseagreen", alpha = 0.7) +
  labs(title = "Scatter Plot of Biodiversity Mean Value vs. Yield",
       x = "Biodiversity Mean Value",
       y = "Yield") +
  scale_y_continuous(breaks = seq(0, ceiling(max(Corn_Data$Yield_value_C, na.rm = TRUE)), by = 1000)) +
  theme_minimal()

Discussion

The analysis of agricultural systems effect on biodiversity and crop yield reveals interesting insights. Both comparator and intervention systems exhibit wide ranges in biodiversity, indicating a diverse variety of life. In particular, our findings suggest that diverse farming practices may foster greater ecosystem diversity, highlighting the importance of implementing these approaches. In relation to crop yield, diversed farming methods displayed a narrower spread of yield, which indicates consistency. Our investigation into the relationship between biodiversity and yield uncovered a slight trend. However, it’s important to note that the relationship is not particularly strong. This finding aligns with research conducted by Jones (2022), indicating that while biodiversity may have some influence on yield, other factors are likely to play significant roles.Future research could benefit from repeated studies, to deepen our understanding of these complex interactions and prioritising sustainable environments.

Conclusion

In conclusion, our analysis reveals the complex relationships between agricultural practices, biodiversity and crop yield. Intervention systems offer more consistent crop yields, and emphasizes the importance of adopting sustainable farming practices for biodiversity. Further research is recommended with a focus on factors that may influence crop yield.

References

The following journal articles were used as background research into our topics of interest. These articles were accessed via reputable journal sites, and are validated by peer reviewers and experts in the field. Throughout selection, careful consideration was given to ensure the chosen articles were free from bias or conflicts of interests, enhancing their credibility and reliability. Additionally, the websites, OpenAi and Quarto were used in assistance for coding, themes and aesthetics.

  • Frison, E.A., Cherfas, J. and Hodgkin, T. (2011) ‘Agricultural biodiversity is essential for a sustainable improvement in food and nutrition security’, Sustainability, 3(1), pp. 238–253. doi:10.3390/su3010238.
  • HTML theming (no date) Quarto. Available at: https://quarto.org/docs/output-formats/html-themes.html#overview (Accessed: 21 March 2024).
  • Jones, S.K. et al. (2021) ‘A global database of diversified farming effects on biodiversity and yield’, Scientific Data, 8(1). doi:10.1038/s41597-021-01000-y.
  • Jones, S.K. et al. (2023) ‘Achieving win-win outcomes for biodiversity and yield through diversified farming’, Basic and Applied Ecology, 67, pp. 14–31. doi:10.1016/j.baae.2022.12.005.
  • OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat