You would use the Chi square test to compare diversities :)

I the response variable is discrete integer (number of species) a "count model" or "GLM" could provide an answer. Depending on how the variance is related to the mean either poisson or negative binomial error distribution could be selected. However, you did not clearly specify your independent variable (https://westcoastuniversitylibrary.libanswers.com/research/faq/295836). As such I do not know how the independent variable can be linked to the dependent variable. Are there two groups or do you have a continues gradient?

For example, two groups:

number of species = group1 + group2 + error

Wim Kaijser In my case the plants are the independent variable and the insects are the dependent variable.

You can read about paired samples t-test.

Do you want to know the species number of insects per plant species and consequently, the difference number of insect species between the i.e., plant species_1 or plant species_3?

number of insect species = plant species_i + ...+ plant species_n + error

Wim Kaijser Yes, I want to compare the diversity of insect species between different field margins (area in the border of field) types like irrigated/tree rich/ herb rich/ shrub rich/maintained. The diversity measure I am using is Effective Number of Species. I have Effective Number of Species values for both insect and plants in field margins.

Senthilraja N. If you work in R my first impression would be to go for the poisson model. This can be to restrictive if overdisperions becomes an issue switching to a negative binomial might relieve some of the restrictions that apply to the poisson model. I would suggest to read up on count-models and the link function (e.g., https://environmentalcomputing.net/statistics/glms/glm-2/) and the emmeans function (https://cran.r-project.org/web/packages/emmeans/vignettes/comparisons.html) . In the code below a very rudimentary code.

library(parameters)

library(see)

library(tidyr)

library(lsmeans)

library(DHARMa)

#Create example data.frame

set.seed(123)

df

Wim Kaijser thank you very much for your efforts in explaining. I will read the links provided by you to make the decision.. thanks once again