Check out a program called EstimateS. You can find it by searching "estimates biodiversity" or here is the link.

http://viceroy.colorado.edu/estimates/EstimateSPages/AboutEstimateS.htm

There are many ways to do this. The problem is with replication. It will be easier in the analysis if you do quantitative sampling rather than presence-absence. The simple statistics is # of species A in habitat A relative to number of species A in habitat B. Each student samples all habitats once, and the sample from each student is one replicate. The class pools data (or teams pool data) to arrive at a finished experiment. Something like a t-test for the three most abundant species and the assignment is finished.

If a similar project is done next year, each class can compare their results to the results from previous classes.

Sampling will be a consideration: trap (baited or not), quadrat, or line-transect.

There could be a student effect. Some students are better at this than others. If this might be interesting then a student would have to take multiple samples from each habitat. There might even be a student by habitat interaction or an effect of order: I sample tallgrass prairie in the morning after breakfast and I pay attention to all the details. By late afternoon when I sample the disturbed area I am hot, tired, and thinking more about dinner.

If done carefully enough, the results could be publishable. While I am sure that is far from the goal, the effort expended by you and your students is locked away forever if it is not published in some form. Please consider this especially if there is some chance that this is a 5+ year activity.

Hi Filippo Maria Buzzetti you could iNEXT. There is some nice paper (http://chao.stat.nthu.edu.tw/wordpress/publication-list/) about it a online tool (https://chao.shinyapps.io/iNEXTOnline/) that you could use even if you don't use R.

Dear Timothy A Ebert and Moreno S. Rodrigues , thank you very much for your answers.

My idea is that the student use a plastic cylinder with basal surface of 1 square meter. This tube will be placed 10 times randomly in each habitat, all the Orthoptera will be collected and kept separated into their habitat of provenance.

The procedure will be repeated daily for 7 days in the same test area.

This is the first year I try this, but hopefully I will do it also in the coming years.

Thanks once more

Filippo

Thank you Ming Kai Tan , I will avoid putting the ecosampler in the same spot.

In each habitat I will count:

-number of species

-number of individuals of each sp

-number of individuals TOTAL

have you a reference paper to suggest?

Thanks

1) Grasshoppers are often very active. The act of placing the ecosampler will cause some grasshoppers to leave. To some extent this is a general problem in any sampling method. The following link is another option for sampling:

https://www-jstor-org.lp.hscl.ufl.edu/stable/3495558?origin=crossref&seq=3#metadata_info_tab_contents

I think this book also has a section on field collection, but I don't have the book with me so I am not 100% certain: http://www.cornellpress.cornell.edu/book/?GCOI=80140100971420

2) As pointed out in the first link, there can be problems in identifying immature grasshoppers. They reared some to adults for identification. This could take a few weeks. If this is not an option consider using species groups.

3) Sampling effort. One or two quadrat samples of one square meter each will probably not catch sufficient grasshoppers to make a working project. Maybe day 1 is a trial to figure out how to sample. It could be nothing more than telling students to go into an area and see how long it takes for them to catch 10 grasshoppers. In this method, it does not count if they see a grasshopper jump and then place the ecosampler on the grasshopper.

4) As an addition to #3 above, how long does it take for students to catch 5 species?

5) Try walking a fixed distance (say 10 meters) and see how many grasshoppers you find. Is your answer in #3 and #4 reasonable, or should you try a different sampling method?

6) Consider the same method but a different target (yes, grasshoppers are fun). The approach would work well for sampling plants in some areas. As I recall identifying grasses can be challenging. You could sample ground dwelling beetles using pitfall traps, or sample using bait: small piles of rolled oats attract some insects and one could look at the diversity of arthropods that arrive at 20 bait stations. This is more an evening (dusk, to early full dark) activity. Dark activities could include black light traps or fermented fruit placed on trees or other surfaces. One of my zoology classes used pitfall traps on the north versus south side of a hill.

7) You don't want the ecosampler in the same spot, You want the ecosampler in different spots far enough away from each other that it is unlikely that a grasshopper that you caught in the first sample is caught again in other samples. There is the possibility (given catch-and-release) that a grasshopper caught the first day is then caught again on the second, and fifth day. The alternative is a catch-and-kill approach, but then you are creating ecological holes that will be filled by something else moving into the ecological niche that you opened. Thus day 1 is a sample of the habitat, but day 7 is then a sample of that habitat as modified by the sampling protocol.

8) The analysis needs to be tailored to the skills of the students: grade school, high school, undergraduate or graduate. There could be a college class devoted to the design and analysis of experiments such as this one with implications in topics from biodiversity to habitat reclamation/restoration or designing urban parks to preserve biodiversity. While it would not be accepted for publication in a peer reviewed scientific journal, you could treat each day as a replicate and use a simple t-test to determine if habitat A was different than habitat B. For many grade (elementary) school students this would be a great achievement to get that far. You could also have students collect several samples and rather than using the number of species use mean number of species per trap.

Timothy A Ebert Moreno S. Rodrigues Ming Kai Tan

I use a plastic tube 1m high, so that grasshoppers can not jump out.

Immature Orthoptera cannot be identified (some at genus level yes).

Of course I will try to sample in different spots.

Students are from high school, so I want to show them: general features of systematic, general features of Orthoptera, sampling methods, identification, some statistics.

If the data from the day 1 are like in the attached file, what you suggest? How can I insert the data in INext, EstimateS or other?

Thanks

Filippo

More a problem of the grasshoppers leaving the area before you can place the tube. To a grasshopper you are a big scary monster and the best thing to do is run away. You will have to be close to them to place the tube. That said, that data that you included looks better than I had expected. I was expecting more zero counts.

EstimateS is fairly simple follow the directions. It is perhaps more complex than you want, giving you several different measures of diversity.

The problem with the existing data is replication. You might consider test1 and test2 replicates. You might consider males and females replicates (with a note that this is a class exercise and not a Ph.D. project).

Species 5 field: 12, 5, 12, 5

Species 5 ecotone: 5, 7, 5, 6

Hypothesized mean difference is 0.

t-test unequal variance 2-tailed p-value is 0.555

(Using the Excel data analysis add-in)

Might discuss why the abundance of males might not be independent of the abundance of females.

You could count the number of species recovered.

Test1) 5, 6, 4

Test2) 5, 5, 4

Is the number of species the same in ecotone versus wood? 2-tailed t-test p-value is 0.204833. With two replicates the answer is that they are the same.