In ecology, we generally could observe the positive relationship between plant diversity (A) and plant biomass (B). Also, we could observe the positive relationship between plant biomass (B) and soil carbon storage (C). But here, I got a negative relationship between plant diversity (A) and soil carbon storage (C). The data is collected from field investigation of grassland.
I'm afraid it appears there really are no relationships since the r2 values are all very low and the data points are very scattered with not many near the extrapolated lines. You have to have low p and high r2 values for there to be a significant correlation i.e. you need both :)
Hi Xiangdeng, I can see why it feels paradoxical when A&B are positively correlated, and B&C are positively correlated, but A&C are negatively correlated. But it happens all the time! Like more sun means plants grow taller, more rain means plants grow taller, but more sun means less rain. I'm in Psychology, so I'm not sure how your science explains your relationships, but I feel it's definitely not paradoxical just from the math. I agree with Andrew you have really small effect sizes, but you also have statistical significance. I feel like it means you probably have some powerful mechanisms you haven't measured. Hope you figure it out! ~ Katie
The problem seems to be that data is too widely spread. This can create "false" correlations. High diversity and low productivity can often be linked. For example, unfertilized natural meadow, towards fully fertilized fields.
- Badges
- Science method