Evaluating riparian vegetation condition

After a brief interlude to write posts about some of the things I just find interesting, I am returning to some work from my thesis.

In post #1 I described the effect of livestock grazing on the cover of different vegetation life forms. Here I focus on the amount of cover at these sites, and if it is a good amount or not.

Vegetation change in itself is certainly not a bad thing at all. Vegetation communities are constantly changing and it would be ridiculous to think that we could ever manage a completely static environment. Change becomes a negative when the vegetation is no longer able to provide the functions that are expected of it, e.g. habitat or food for fauna. Negative changes such as this occur naturally in the landscape from events such as fires, floods, droughts, etc. and there is little we can do to influence these naturally occurring events. Our primary concern, as researchers and managers are the negative effects that caused by humans.

We can only tell if vegetation has been negatively impacted if we know what relative condition it is in. Previously I wrote a post about what is vegetation condition, which described the use of benchmarks against structured sampling methods to evaluate how different the vegetation of a given site is from what it ‘should’ be in the absence of human disturbance. I am referring to the Victorian government’s standardised process known as Vegetation Quality Assessment. Accompanying the standard assessment method and the calculation process of condition scores, is the set of benchmark attributes for each vegetation type (or specifically, Ecological Vegetation Class, EVC) in Victoria.

So if these benchmarks of vegetation condition are freely available, can I use them to evaluate the condition of my study sites, even though I used a different survey method? Well, yes, and I did. Using a different survey method makes it impossible to generate the condition score based on the standardised methods, but I can still use the benchmark attributes to asses specific native vegetation attributes of my sites. For this post I will focus on vegetation cover.

This site looks ok but does it have the amount and diversity of vegetation that would be expected of a high quality site?

This site looks ok but does it have the amount and diversity of vegetation that would be expected of a high quality site? Does it have nearly enough of some things and far too few of other things?

To determine the condition of the native vegetation attributes at my sites I needed to do 3 things:

1. Determine the cover of each attribute for each site

2. Determine the vegetation type (EVC) for each site

3. Compare observed cover values against benchmark values

COVER: So firstly I needed to find out the cover of different attributes within my sites. I surveyed 180 sites along creeks in northern Victoria (for some more background on my study look here, and more on survey methods look here). I separated each site into an inner (near the creek) and outer (away from the creek) zone to explore differential effects  with distance from the water. The results for the cover of different life forms, bare ground, litter and tree cover are shown below. The spread of data is similar in the inner and outer zones, but there were commonly large differences between zones for a specific site (as indicated by lines). This information is interesting in itself and tells us that average cover for many of these life forms was low – but is that ok? How much cover should there be? Comparing these values against the benchmarks can help us evaluate this.

Percentage cover of ground layer, native understorey and midstorey life forms, and tree canopy across all sites. Each point indicates cover within a site, where closed circles (•) are cover within the inner zone and open circles (ο) within the outer zone. Black plus signs indicate the median values. Grey lines connect cover from the inner and outer zone within the same site. Points without connecting lines occur for sites with insufficient data in either zone to generate a cover estimate. Data are presented on a logit scale. I added 0.0025 to scores of zero, equating to one hit every 400 points, to enable plotting on the logit scale. Vertical lines are medians. Note change of scale for litter and tree cover.

Percentage cover of ground layer, native understorey and midstorey life forms, and tree canopy across all sites. Each point indicates cover within a site, where closed circles (•) are cover within the inner zone and open circles (ο) within the outer zone. Black plus signs indicate the median values. Grey lines connect cover from the inner and outer zone within the same site. Points without connecting lines occur for sites with insufficient data in either zone to generate a cover estimate. Data are presented on a logit scale. I added 0.0025 to scores of zero, equating to one hit every 400 points, to enable plotting on the logit scale. Vertical lines are medians. Note change of scale for litter and tree cover.

BENCHMARKS: The next step in comparing vegetation to the benchmarks is to determine vegetation type or EVC for each site. Normally this would be done on delimited patches (‘habitat zones’) of similar vegetation but in this study I determined an EVC for each zone (inner and outer) within each site, based on statewide EVC mapping. If you want to know more about how I did this just let me know.

COMPARISON: Benchmark Comparisons were made by subtracting the benchmark cover value from the observed cover from my surveys. So anything below ‘0’ meant that the cover was below the benchmark level, i.e. below what it ‘should’ be in a good quality site. Small shrubs, litter, and tree cover were the only attributes that were generally at or above the benchmark levels, while all others were below. Medium tufted graminoids (such as Austrostipa scabra) and medium herbs (such as Calocephalus citreus) were by far the worst performing. Interestingly, despite the differences in cover between inner and outer zones at many sites, the difference to benchmark levels was roughly equivalent across attributes. This suggests that the different benchmark levels for each zone were appropriately accounting for these differences.

Difference in percentage point cover between observed and EVC benchmark cover of ground layer, native understorey and midstorey life forms, and tree canopy cover. Box plots show the range of cover differences across all sites. White boxes indicate inner zone covers; grey boxes indicate outer zone covers. Boxes indicate the median and inter-quartile range and whiskers extend to the most extreme data point which is no more than 1.5 times the inter-quartile range from the box.

Difference in percentage point cover between observed and EVC benchmark cover of ground layer, native understorey and midstorey life forms, and tree canopy cover. Box plots show the range of cover differences across all sites. White boxes indicate inner zone covers; grey boxes indicate outer zone covers. Boxes indicate the median and inter-quartile range and whiskers extend to the most extreme data point which is no more than 1.5 times the inter-quartile range from the box.

It is clear that some of these vegetation attributes are well below the desired levels within my study area. This information can help inform management of the current status of these sites, suggest which life forms should be targeted, and the amount of change that might be required to restore the sites.

This is just a small part of the picture and it needs to be evaluated in the context of land management at each site, but I think it is a neat use of existing resources to help evaluate vegetation condition at a site to inform management.

Cheers.

Advertisements
This entry was posted in PhD Research and tagged , , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s