Very good question , unfortunately the plant health traits which you have choosen , may not be express any significant difference . it would be mor eprudent to identify traits at molecular level , physiological parameters , biochemical parametres and finally the phenotypical characters. We also need to re-assess changes taking place in rhizosphere ecology facilitating better or poor plant health traits. for example carbon accumulating ability has direct consequences on accumulation of othe rnutrients like nitrogen and phosphorous , aprt form carbon fertilizing capacity translating into improved soil fertility , microbial turnover and mcirobial quotient are another good parameters. Its such an open arena for you , depending upon the facilities that you have at your working place plus your own expertise , all these matter a lot , what should be the excat premise of such work....good luck

Hi Nicholas,

Measuring plant health is unfortunately highly ambiguous what specifically are you investigating with regards to "healthy" plants. For example from a demographic perspective reproductive output would be considered an indicator of long-term health while physiological parameters such as native embolism or C:N ratios of leaves would indicate immediate health characterization of plants.

I would caution using "generic" (broad applicable) traits such as SLA (conversely LMA) as these traits are often abstract in nature and have no direct relation to a specific function rather they are applied as generic explanatory variables for plant strategies. For example SLA is related to drought, nutrient, resource capture, etc. this is why such a trait has very little explanatory or correlatory power when discussing functional plant responses. Might I suggest exploring what Brodribb (2017) references as "mechanistic traits" with direct physiological implications.

One added thing you mentioned you were investigating "within species", in such instances of intraspecific trait contrasts I would recommended having a extremely large sample size from multiple populations to capture intraspecific variance.

If you consider a community, there are so many indices, survival rate, AGB - which are good indicators. You can see this small clip

Thank you all for your insightful answers. It appears that immediate health characteristics would be best. Anoop Kumar Srivastava, your suggestion of exploring soil traits and microbial health is ideal. I will be based at a soils science department for the duration of the study and will be able to upskill in that field.

Daniel Buttner, I should have just given you a call about this, Al said you're doing some similar work. I agree that large sample sizes are needed, but probably unlikely considering the state of the thicket-wide plots.

Are there physical/physiological measurments that can be paired with soil moisture, carbon, and microbial diversity?

Yes , quite possible. Please try to see , how carbon sequestering ability of treated soil triggers carbon foot print of plant or does it create any additional nutrient sink in plant vis-a-vis crop physiological parameters in order to achieve better physiological preparedness of treated plant, lets explore these areas ..thank you

There are many potential measurements to be paired with soil moisture, however the traits will be superficial not directly linking soil moisture to some functional attribute of the plant, but will be descriptive of resource allocation/plant economic spectrum discussions. Additionally, if considering soil moisture and plant physiological parameters a range of values will need to be collected not only the extrema, for example Fm/Fv at 20% and 80% soil moisture content as this will do little to demonstrate the physiological thresholds and range tolerable by plants rather provide a snapshot of values at a specific timepoint this would be appropriate for highly conserved attributes such as in the case of drought embolism. Again there are many 'physical/physiological' parameters that can be considered but depends on aim or main question being addressed if soil moisture is being consider would explore more "generic" (broad inference) traits such as SLA as you will get a correlation regardless, but if mechanistic traits are to be considered would explore root attributes, localized N2 conditions and possible N2 and H2O isotopes for stronger suppositions to plant strategies. Given all this, and the little data available from TWPs, I would suggest the use of remote sensing platforms as well where extracting productivity indicators (NDVI, VCI) both historical (MODIS - 2000, LANDSAT - 1980s) would provide value data to be coupled with short-term on the ground data.

One addition considering that you are researching the TWPs might be interesting as the dominant restoration species is Portulacaria afra to see the return at least proportionally of the canopy woody species as while many propose such few have document said return and only speculate. A ratio contrast and possibly if differing restoration years of plots might infer age required if at all necessary for biodiversity return as opposed to success of only the introduced dominant.

Daniel Buttner the timelines are too short. I walked through a TWP last week and while there were some additional species in there it was just small karoo shrubs. The woody elements need decades and rain to return, with more than half the time since establishing the TWPs being a drought it's unlikely that tree seedlings will be surviving. Hench the interest in seeing how the existing mature woody species are doing. But I agree that looking at species return across a restoration time gradient will be very interesting in the next 10 years.

Thought they might be, interestingly the karroid shrublands appearing, however anticipated as they are pioneering species. The issue might come with woody components as derived from subtropical origins and given specific circumstances for germination and growth capturing health indices at a specific time frame might be limited as many trees that might appear healthy may already exist within a carbon deficit thus risky flushing events and desynchronized flowering might be apparent providing a false trail of productivity. Given aspects of drought as focus might suggest that cognizance of the snapshot detection of health be consider. Consider looking into the carbon starvation theory under broader why trees die from drought. Simply capturing health condition is not an adequate representation of population survival probably this is particularly important in arid woodlands. If interest is in canopy tree conditions, taking cognizance of carbon deficit skewing condition, canopy vitality indicators and potentially leaf "hard' traits might be ideal other traits if you are considering to monitor soil moisture might be relative water content and photosynthetic performance parameters.