Plant traits can be used to understand how plants respond to changes in the environment, including increasingly severe drought. Leaf mass area (LMA, dry mass of leaf per unit ofarea) is a key trait that provides information about how plants invest their resources, and the position of a species along the Leaf Economics Spectrum. This trait featured heavily in Lectures 11-14 (those on functional traits, strategies, and adaptations to various environmental stresses). During today’s prac, we will investigate variation in this widely used trait – variation in relation to site climate, and variation in relation to other key traits.
The second functional trait we will focus on today is wood density (WD), which featured in lectures on traits/strategies (L11) and adaptations to aridity (L12). WD is the ratio of wood dry mass to fresh volume. Higher WD indicates higher construction costs per unit wood volume. Higher density wood is more mechanically robust than low density wood and is resistant to wood-boring insects and pathogens. But, species with high WD tend to only ever achieve slow growth rates. That is, there is a trade-off between wood construction costs, mortality risk and growth rate. But wood density is also related to plant hydraulics (i.e. water relations); high WD species typically have more narrow xylem vessels and thus slower stem hydraulic conductance (Lecture 12).
The third trait we will consider is adult plant height (also featured in several lectures). Height is important because it strongly influences how much light plants receive, however, it comes with significant costs. Plants that grow very tall have higher stem construction costs, higher risks of mortality (due to biomechanical failure or exposure to strong winds), and it can also be difficult for them to pull water up from soil and into their leaf canopy.
In lecture 11 you were introduced to the concept of ecological strategies, which is essentially how a species “makes its living” to ensure genetic continuity into the future. A key underlying concept with regards to strategies are that there are tradeoffs between different processes or functions. For example, there might be tradeoffs between seed mass and the number of seeds. In this prac, we might hypothesize that there is trade-off between building dense (high LMA) leaves and/or branches with dense wood versus increasing plant height. Or between having high LMA leaves and being very tall. The strategy that a plant “chooses” might influence its growth and survival, especially in water-limited or otherwise stressful environments. Today we will explore these tradeoffs using a trait dataset for Australian trees.
To investigate trait variation across species in response to aridity, we will conduct a traitbased data analysis exercise using Minitab 18. During the prac, we will answer some specific ecological questions using statistical tests covered in the Stats Refresher Prac. We will use a dataset with three key plant traits (LMA, WD, and Height) collected for eight Australian tree species measured along an aridity gradient (Anderegg et al. 2020). Site aridity was determined on the basis of a moisture index (MI, where high MI means more moisture) calculated as the ratio of total precipitation to potential evapotranspiration (see Lecture 18). We can use MI to classify sites by the degree of aridity. Here we’ll adopt the convention that low MI values (MI < 0.5) correspond to arid sites, intermediate values between 0.5 and 0.65 correspond to dry sub-humid sites, and values > 0.65 correspond to humid sites.
1. First thing’s first, create a folder in your local directory (e.g. C:/prac7), download “DataAnalysisPrac7.csv” from iLearn and save it within the folder you just created. Open the file in Minitab 18 (go to File/Open then browse for the file in your computer and select it using the default options. You should see a dataset with 6 columns and 99 rows of categorical and numerical data. Columns 1-2 containcategorical data including Species ID (note that we use a short name for each tree species) and Climate (drid, Dry sub-humid and humid), and 4 numerical columns include moisture index (unitless), LMA (units g/cm2), WD (units g/cm3) and plant Height (units m).
2. Now let’s visually inspect the distributions of the data, which is standard practice before doing a statistical analysis. Go ahead and create histograms for each trait. Do the data look bell-shaped or not?
3. Note: In the Stats Refresher Prac, we made histograms and then tested whether the raw data met the assumptions of normality, but this isn’t the best approach. When doing a parametric test like ANOVA, it’s more appropriate to test whether the residuals of the model meet assumptions of normality. In most cases, when the raw data are normally distributed so are the model residuals, but this isn’t always the case.
1. Do the various species within each climate zone (arid, dry sub-humid, and humid) differ significantly in their LMA?
2. Considering all species together, does LMA vary in relation to site moisture? Do we see the same trends across species?
3. Considering all species together, does wood density (WD) vary systematically with LMA? Does this relationship vary among species?
4. Considering all species together, does wood density (WD) vary systematically with plant height? Does this relationship vary among species?