The following details are explaining my experimental design and my doubts about the statistical analysis.
Plant is growing in Nutrient 1 and Nutrient 2 . The toxicants at difference concentrations were (0 to 5%) mixed with nutrient 1 and 2. The experiment is conducted at 5 different temperatures. (Nutrients Vs Concentration Vs Temperature)
In nutrient (1) - concentration of toxicant is 0 (control), 1%, 2%, 3%, 4% and 5% (Treatments). All plants were grown well in all temperature experiments therefore here, I used DMRT to compare the significant results between control (0%) and treatments (1% to 5%).
In Nutrient (2) - concentration of toxicant is 0 (control), 1%, 2%, 3%, 4% and 5% (Treatments). Here the plants fail to grow above 1% therefore, I have only 0% and 1% datas in all temperture experiments. Here, DMRT is not fit to compare the resutls since the datas are fewer than three groups. So, what type of statistics best fit to compared these results (compare 0% with 1%)?
For all temperature experiments, fail to run the two way-ANOVA since the un-even datas in Nutrient 1 (control and 5 concentrations, 1 to 5%) and nutrient 2 (control and 1% concentration). Here what type of statistical analysis is best fit to compare the results ?. The same issue is continuing for three-way ANOVA (nutriens vs concentration vs temperature). Is it possible to compare these three factors by three-way ANOVA or any other statistical methods fit for comparision?
Thanks for your help.
Assuming plants are replicates (each one or average from pot), you have 6 treatments with nutrient 1 and 1 or 2 (assuming 1% is acceptable). so you have 7 or 8 treatments at 5 temperatures (= 2-way anova). compare 2 'control' & 2 '1%' means. In plant growth, temperatures greater than around 35C growth may stop and plants may even die. Other option would be to fit some curve (e.g. diminishing returns Mitscherlich) to temperatures for each of 7 or 8 treatments in a grouped regression (parallel curve analysis)...good luck
Dear Mewa Singh Dhanoa
Thanks for your answer. All the experiments were triplicated. But still the answer is unclear to me. Could you please explain in detail?
I generated 2-way design on Genstat(vsni.co.uk) and got following ANOVA output (treatment means & SEDs for further protected t-tests to compare any pair of MEANs). See if you follow the output (layout bad, see attached instead):
>>>
56 "Two-way design" 57 DELETE [REDEFINE=yes] _ibalance 58 A2WAY [PRINT=aovtable,information,means; TREATMENTS=nutmix,Tempp; FACTORIAL=2; FPROB=yes;\ 59 PSE=diff; PLOT=*; COMBINATIONS=present; EXIT=_ibalance] zz; SAVE=_a2save Analysis of variance
Variate: zz
Source of variation d.f. s.s. m.s. v.r. F pr.
nutmix 6 2.06515 0.34419 6.99
- Badges
- Science topic
- Similar topics
- Agricultural Science
- Environmental Science