In our current study, we found that both GEP and ER had positive correlation with importance value of grass+forbs but showed negative correlation with the importance value of sedge. What more interesting is that the slopes of the linear regression was higher in GEP than in ER no matter the direction of the correlation. Can we conclude that under a warming climate species composition change favors net carbon sink in the alpine meadow based on the regressions slopes?
Thanks Nina. The importance value (IV) was calculated according to the relative height and frequency of each species. IV of plant group is the sum of species that belong to sedge, or grasses and forbs. Yes, we measured GEP and ER, please see the figure that I posted with the ask. Have you submitted or published papers about your findings?
I agree with Nina Sajna. The pH change has a direct effect on the fluctuation of C-balance of an ecosystem.
Fei Peng -
Hello. - Looking at your two scatterplots, I have a few comments. First, it is great that you are looking at scatterplots. I think that when people skip that step they can miss a good deal of useful information, and insight. As for the scatterplots themselves, I think the first one may make you want to reconsider whether or not it is linear. I cannot make out the information written there, and am not a subject matter expert on this, but could this be a nonlinear curve approaching the y-axis as shown, exponentially? What does the theory of your subject matter say about that?
Also, these data appear strongly heteroscedastic. The second graph appears, though I cannot read it well, as a typical linear regression through (actually to) the origin, with increasing variance in y as x (the size measure) increases. This is heteroscedasticity, and oddly in statistical applications, it is often ignored, which is a mistake, and can be a big one. It will often impact the estimate of the slope, and generally has a bigger impact on the standard error of the slope. Further, confidence intervals about the predicted y-values are greatly affected. (See link below.) This may be a factor in decision making, and especially in determining if you have adequate data.
Best wishes - Jim
Data CRE Prediction 'Bounds' and Graphs Example for Section 4 of ...
Dear Prof. Jim
Thanks so much for your advice on the data processing.
I redid the linear regression with the link your providedk, but it is diffcult in finding an appropriate expression to describe the statistical proecessing the Material and Methedology section in a paper. Could you help me?
Will you like to be listed as the co-author? I really appreciate your advice and suggestion in the data processing.
Best Regards!
Fei
- Badges
- Science topic