• 1 Aims & Objectives
  • 2 Which Ordination is Best?
  • 3 Comparing Ordination Algorithms
    • 3.1 Base plot
    • 3.2 Base plot with transect lines
    • 3.3 Base plot with zones
      • 3.3.1 Convex Hull Key Points
    • 3.4 Base plots with species
      • 3.4.1 Interpreting species and quadrat relationships
  • 4 Let’s do the time warp again…
    • 4.1 All quadrats across years
    • 4.2 Convex hulls across years
    • 4.3 Convex hulls across years and position
  • 5 Environmental data
  • 6 Practical Report Instructions
    • 6.1 Introduction
    • 6.2 Methods
    • 6.3 Results
    • 6.4 Discussion
  • 7 Species abbreviations

1 Aims & Objectives

There are four aims for the Dune Transect Practical on the Nelson Mandela University Nature Reserve, each with their own set of objectives:

  1. To expose students to coding and thinking in R.
    • Objective 1.1: To demonstrate how to import and manage data in R using scripts, addressing common issues that arise during the process.
    • Objective 1.2: To explain key coding concepts such as objects, functions and parameters.
    • Objective 1.3: To equip students with problem-solving skills by guiding them through troubleshooting techniques and debugging in R.
  2. To illustrate the way community continua are depicted using multivariate analysis techniques.
    • Objective 2.1: To use this year’s data to generate ordinations using a range of algorithms (i.e. PCA, DCA, and NMDS) in R and the vegan library.
    • Objective 1.2: To draw these ordinations by hand and link the quadrats along each transect.
  3. To compare the various multivariate analysis techniques.
    • Objective 3.1: Using the hand-drawn figures from Objective 1.2, describe the appropriateness of each ordination algorithm to the data (see Section 2 below).
  4. To illustrate successional changes in vegetation after fire.
    • Objective 4.1: To interpret figures provided to you to include in your write up.
    • Objective 4.2: To assess the impact of fire on Dune Fynbos vegetation.

2 Which Ordination is Best?

I suggest a very practical approach—apply the various ordination techniques and then select the simplest one that effectively spreads the points (the order of complexity, from low to high, is: CA, PCA, DCA, NMDS). This allows you to visually compare how well each method captures the variation in your data and ensures that you choose a method suited to both the structure and complexity of your dataset.

Key things to consider while following this strategy:

  • NMDS often provides good separation even for complex, non-linear datasets but can be more challenging to interpret compared to PCA or RDA.
  • PCA could be preferred if you find that linear methods offer a sufficiently clear spread of points with simpler interpretation, especially when there’s less species-environment interaction complexity.
  • DCA is useful when there are clear gradients in the data that simpler methods might not detect as well.

This approach gives flexibility while allowing you to balance interpretability and statistical power based on the dataset’s underlying structure.

3 Comparing Ordination Algorithms

3.1 Base plot

The following ordination plots use three different techniques, with each transect represented by a unique color and quadrat numbers labeled. These plots reflect what you performed in the practical session.

3.2 Base plot with transect lines

3.3 Base plot with zones

The same ordinations as above are now shown with the quadrats combined across the three transects but split into different zones:

  • Quadrats 1-3: Dune crest
  • Quadrats 4-7: Dune slope
  • Quadrats 12-13: Dune slack

The grouping of these zones are shown using convex hulls. Convex hulls are used to visualize and delineate the boundaries of groups of points (e.g., species, samples, or quadrats) in ordination space. Convex hulls are a helpful tool for exploring patterns and relationships between groups within the multivariate space of an ordination plot.

There are two important things to note here:

  1. Quadrats 8 to 11 varied in their positioning between the dune slope and slacks amongst the quadrats, so I decided to exclude them in the convex hulls.
  2. Transect C encountered a bushclump between the crest and the slack. Bushclumps often break the pattern of vegetation change down a slope. Thus, I excluded the dune slope points of Transect C from the dune slope convex hull.
  3. It is crucial to keep track of what happened on the ground to understand what is happening in the ordination. The occurrence of the bushclump generated substational “noise” in the ordination (see quadrats 4-11 in Transect C). I suggest that these quadrats be excluded from the subsequent discussion about the vegetation change across the dune face.

3.3.1 Convex Hull Key Points

  • A convex hull is the smallest polygon that can enclose a set of points in an ordination plot. It connects the outermost points in a group, creating a boundary that contains all other points in that group.
  • In ordination analyses (e.g., PCA, NMDS), convex hulls are often used to highlight and compare the spatial arrangement of different groups, such as treatment groups, habitats, or species clusters.
  • Convex hulls help visualize the spread and separation of groups, making it easier to interpret similarities or differences between them. If the convex hulls of two groups overlap, it may suggest that the groups are more similar in terms of species composition or other variables. Conversely, separate hulls indicate more distinct groups.

3.4 Base plots with species

Below are the base plots with the descriptors added to the ordination, i.e. the species.

3.4.1 Interpreting species and quadrat relationships

In vegetation ordination, interpreting the positioning of species (descriptors) relative to quadrats (objects) helps reveal patterns in species distribution and community composition. Here are the key points for interpretation:

  • Proximity of Species to Quadrats:
    • Species that are positioned closer to a quadrat indicate that those species are more likely to be strongly associated or abundant in that particular quadrat.
    • Conversely, species positioned farther from a quadrat are more likely to be less abundant or absent in that quadrat.(Remember that ordinations only capture general trends).
  • Distance between Quadrats:
    • As covered in the lectures, quadrats that are close to each other in ordination space share similar species compositions. They likely have a similar set of species or are influenced by similar environmental factors.
    • Quadrats that are far apart in the plot are more dissimilar in species composition or are subject to different environmental conditions.
  • Species in the Center:
    • Species positioned near the center of the ordination plot tend either be rare (restricted to only one or two quadrats) OR are more generalist or widespread, occurring across multiple quadrats with no strong preference for any specific group of quadrats.
    • Species positioned near the edges of the plot are usually more specialized, showing a stronger preference for certain quadrats or environmental conditions.
  • Clusters or Groups:
    • If species cluster around specific groups of quadrats, it suggests a species assemblage unique to those quadrats, reflecting similar ecological conditions or interactions.
  • Take away points
    • Species near quadrats = Higher abundance or presence in those quadrats.
    • Quadrats close together = Similar species composition.
    • Species near the center = Rare (with low cover) OR generalist, widespread.
    • Species on the edges = Specialist, specific habitat preference.

4 Let’s do the time warp again…

This same practical has been running every since 2016 (except 2020 and 2022). In 2017, this section of the reserve burnt. Thus, not only can we use this data to see how the vegetation shifts across the dune, but also see how the vegetation has recovered after the fire. The 2017 data was collected 5 months after the fire.

4.1 All quadrats across years

4.2 Convex hulls across years

4.3 Convex hulls across years and position

Below I show a subset of points, splitting the up across the different zones. These aren’t new analyses! They are the same analyses as those shown directly above, but show the points in the different zone (by year).

HINT: In the figures above, one zone is remarkably stable, while the other two seem to have shifts in their community. Why could this be? Think of resprouters and reseeders, and the topics covered in the BOTV211 ecology module. Look at the species dominant in these various zones.

5 Environmental data

While environmental data can be incorporated directly into ordinations to explore how environmental gradients influence species distributions, for this practical report, we will take a different approach. You are encouraged to explore your environmental data across the three dune zones (crest, slope, and slack) using standard figure comparisons, such as box and whisker plots. These plots will allow you to visualize how the environmental variables differ across the dune positions, helping you to understand the distinct conditions in each zone.

Focus on creating clear comparisons that highlight how factors like soil moisture, pH, or other measured variables vary between the dune crest, slope, and slack. Use these comparisons to help interpret how environmental differences might drive species patterns across the landscape.

Assign the following quadrats to the various zones:

  • Quadrats 1-3: Dune crest
  • Quadrats 4-7: Dune slope
  • Quadrats 12-13: Dune slack

6 Practical Report Instructions

For this assignment, you are required to write a practical report based on the vegetation quadrat surveys you conducted in the field. Your report should explore ordination techniques while also integrating your understanding of dune ecology, using both the data you collected and the provided environmental variables.

Report Structure: The report must follow the IMRAD format (Introduction, Methods, Results, and Discussion):

6.1 Introduction

  • Objective: Provide a concise introduction to the report, outlining the ecological context of the study. Discuss dune ecology, touching on the importance of vegetation dynamics in coastal dune systems, species composition, and environmental factors (such as soil composition, wind, moisture gradients, and fire) that influence plant communities.
  • Background on Ordination Techniques: Briefly introduce ordination as a method for analyzing ecological data. Highlight why ordination is important in understanding patterns of species distribution and community composition in relation to environmental gradients.
  • Aim: Clearly state the aims of the practical (relevant to the practical write-up!).

6.2 Methods

  • Fieldwork Summary: Describe the quadrat survey methods used during the fieldwork. Include the size and number of quadrats sampled, the species recorded, and any relevant environmental data collected (e.g., soil pH, moisture, and light availability).
  • Ordination Analysis: Outline the ordination techniques applied (PCA, NMDS, CCA, etc.) to the vegetation data. Explain why you applied each method and describe how the ordination plots were generated.
  • Environmental Data: Briefly describe the environmental data provided and how it was incorporated into the analysis.

6.3 Results

  • Ordination Plots: Present your ordination results (figures), ensuring that axes are properly labeled, and legends are clear. For each plot, mention the ordination technique used (PCA, NMDS, CCA, etc.) and describe the spread of points (species and quadrats) along the ordination axes. You are encouraged to use the provided figures in your report, but ensure you select only those relevant to your analysis. Use PowerPoint to cut out the necessary figures, keeping their legends intact, and resize them as needed without altering the content. When including figures, make sure they are clearly labeled, accurately referenced in your report, and accompanied by the correct legends. I suggest that for the time-series analysis, you select only one ordination algorithm to report on.

  • Environmental Data: Provide your environmental data analyses comparing and contrasting the values across the three dune zones.

  • Interpretation of Results:

    • Describe any visible patterns or trends in species composition or community structure. Are there distinct clusters of quadrats?
    • How are environmental variables (if included) influencing the species distribution?
    • Compare the results from different ordination techniques—do some techniques give a clearer or simpler spread than others?

6.4 Discussion

  • Comparison of Ordination Techniques: Reflect on the performance of each ordination method. Which method provided the best spread of points and the clearest ecological insight? Discuss the strengths and limitations of each technique in the context of your data.
  • Ecological Interpretation: Link the ordination results to your understanding of dune ecology. Discuss how the observed patterns of species distribution relate to ecological processes in the dune environment, such as succession, disturbance by wind or sea spray, nutrient availability, or fire.
  • Field Experience: Incorporate insights from your fieldwork. Did your observations in the field match the patterns revealed by the ordination? Were there any species or environmental trends that stood out during data collection and were confirmed by the analysis?
  • Conclusion: Summarize your key findings, emphasizing the value of ordination in exploring species-environment relationships and how this contributes to understanding dune ecology. General Guidelines: Ensure the report is concise and follows a clear logical flow from fieldwork through analysis to interpretation.

Use headings and subheadings to organize sections clearly.

Include all necessary figures and tables, ensuring they are well-labeled.

Cite relevant literature on ordination techniques and dune ecology where applicable.

By following these guidelines, your report should provide a thorough exploration of both the ordination techniques and the ecological dynamics of dune vegetation, incorporating both your analytical results and the experience gained during fieldwork.

7 Species abbreviations

Spp_abbrev Species
Aca_sali Acacia saligna
Ant_aeth Anthospermum aethiopicum
Asp_aeth Asparagus aethiopicus
Azi_tetr Azima tetracantha
Car_bisp Carissa bispinosa
Car_deli Carpobrotus deliciosus
Chi_bacc Chironia baccifera
Col_pulc Coleonema pulchellum
Cra_eric Crassula ericoides subsp. ericoides
Cra_expa Crassula expansa subsp. filicaulis
Cyr_lodd Cyrtanthus loddigesianus
Ehr_vill Ehrharta villosa
Ele_limo Eleocharis limosa
Eri_glum Erica glumiflora
Euc_race Euclea racemosa subsp. racemosa
Fel_echi Felicia echinata
Fic_ramo Ficinia ramosissima
Gal_secu Galenia secunda
Hel_tere Helichrysum teretifolium
Ind_glau Indigofera glaucescens
Ind_verr Indigofera verrucosa
Jam_micr Jamesbrittenia microphylla
Ked_nana Kedrostis nana var. nana
Lau_tetr Lauridia tetragona
Lys_arve Lysimachia arvensis var. caerulea
May_proc Maytenus procumbens
Met_muri Metalasia muricata
Mor_quer Morella quercifolia
Mys_aeth Mystroxylon aethiopicum subsp. aethiopicum
Ole_exas Olea exasperata
Pan_deus Panicum deustum
Pas_cory Passerina corymbosa
Phy_eric Phylica ericoides
Rap_gill Rapanea gilliana
Res_eleo Restio eleocharis
Rom_rose Romulea rosea
Sea_glau Searsia glauca
Sea_laev Searsia laevigata var. laevigata