Understand the relationship
The reasoning behind the result
Match taxon identity before comparing communities
Observed union = a+b+c
The same taxon label represents the same taxonomic entity in both observations. A shared taxon contributes to a; one recorded only in A contributes to b; one recorded only in B contributes to c. A listed taxon with two zeros is outside the observed union.
The calculator cannot reconcile synonyms, different identification resolutions or changes in classification. Missing values must be resolved before entry. Treating an unknown record as zero would change the comparison without evidence.
Jaccard and Sørensen use different shared-taxon weights
J=a/(a+b+c); S=2a/(2a+b+c)
Jaccard divides the shared taxa by all distinct taxa recorded in either community. Sørensen counts the shared component twice, corresponding to its presence in each community's richness total. Both range from zero to one when the observed union is nonempty.
They are related by S=2J/(1+J), but their numerical values are not interchangeable. Dissimilarity is one minus the corresponding similarity. No significance level or ecological quality threshold is attached to either result.
Incidence intentionally discards abundance magnitude
In abundance-input mode, each positive value is converted to presence before calculating either index. This lets a visitor use a matched abundance ledger while making the conversion explicit. A taxon with one recorded individual and one with a thousand each contribute one presence.
Two communities can therefore have incidence similarity one while their abundance distributions differ greatly. Quantitative overlap measures such as Bray–Curtis answer a different question and should not be substituted silently. Use the community diversity workbench to inspect abundance balance.
Empty records and double zeros require explicit handling
Adding taxa recorded in neither community does not change these incidence similarities. This differs from coefficients that count shared absences, whose result depends on the supplied taxon universe.
When both communities are empty, there is no observed union and both formulas have zero denominators. The workbench reports them as undefined. When only one is empty, the nonempty union makes similarity zero. Neither result establishes true absence, equal sampling effort or adequate detection.
Follow the numbers
Two shared taxa in a union of five
- Community A records taxa 1, 2 and 3. Community B records taxa 1, 2, 4 and 5. Thus a=2, b=1 and c=2.
- Jaccard=2/(2+1+2)=0.4. Sørensen=4/(4+1+2)=4/7≈0.571429.
- A further listed taxon with zero in both observations is excluded from both denominators and leaves these results unchanged.
The two values describe the same incidence records with different weights; the classification ledger explains their denominators.
Quick guide
How to use this calculator
- Choose explicit presence values or abundance-derived presence.
- Match taxon identities across the two observations and enter both values for every row.
- Read the shared and unique counts before comparing the two similarity conventions.
- Use the classification ledger to locate mismatched or missing records. The indices describe recorded incidence, not detection-adjusted ecological similarity.
Calculation method
Calculation and interpretation
Trace community similarity back to shared taxa and taxa unique to each observation.
Jaccard=a/(a+b+c); Sørensen=2a/(2a+b+c), where a is shared, b only A and c only B
Worked example
Two shared taxa in a union of five
The two values describe the same incidence records with different weights; the classification ledger explains their denominators.
Jaccard=a/(a+b+c); Sørensen=2a/(2a+b+c), where a is shared, b only A and c only B
Supported inputs
Precision and limits
Recorded incidence only
No correction for missed detection, taxonomic mismatch, unequal effort or sampling uncertainty is inferred. A recorded zero does not establish ecological absence.
No abundance-sensitive score
Positive abundances are explicitly reduced to presence in these incidence formulas. The output does not measure quantitative abundance overlap or establish community quality.
Continue calculating
Related calculators
