Understand the relationship
The reasoning behind the result
Metamorphosis is represented by declared endpoints
Ri = Yexit,i/Yentry,i
The ratio compares the same entered trait at two operationally staged endpoints. It does not reconstruct the path between them or decide when metamorphosis began or ended.
Some traits can fall during tissue resorption or energetic reorganization while others rise. The workbench therefore retains signed change instead of requiring growth.
Log change makes multiplicative responses comparable
ΔlnYi = ln(Yexit,i/Yentry,i)
A log response of zero means equal positive endpoint values; positive and negative values preserve reciprocal gains and losses symmetrically. The corresponding percentage change is exp(ΔlnY) − 1.
The interval log rate divides this response by the entered transition duration. It is a descriptive endpoint rate, not evidence that change was exponential within the interval.
Absolute and proportional changes answer different questions
ΔYi = Yexit,i − Yentry,i
Absolute change retains the trait unit while proportional and log changes normalize by the entry value. A larger absolute change need not be a larger proportional response.
The ledger keeps all three views so an endpoint ratio is not substituted for a measured amount.
Individual pairs and group summaries are not interchangeable
A paired row follows one individual across endpoints; a group-summary row compares externally produced estimates. The same arithmetic does not give them the same sampling interpretation.
Equal-weight aggregate values do not model nesting, survival selection, missing individuals, stage-dependent sampling or uncertainty in group estimates.
Follow the numbers
Retain an amphibian mass decrease
- Use 1.25 g at the declared entry and 0.98 g at completion after 3 d.
- Signed absolute change is 0.98 − 1.25 = −0.27 g.
- The endpoint ratio is 0.98/1.25 = 0.784.
- Percentage change is (0.784 − 1) × 100 = −21.6%.
- Log change is ln(0.784) = −0.243346, and the interval log rate is −0.081115 d⁻¹.
The negative record is retained as an endpoint trait change and does not identify a cause or fitness consequence.
Quick guide
How to use this calculator
- Declare the entry and exit stage criteria, trait definition, time unit and paired-individual or summary basis.
- Use the same positive trait quantity and measurement protocol at both endpoints and across all rows.
- Inspect signed absolute, percentage and log changes for every record; decreases are not converted into gains.
- Use the aggregate rows only as equal-weight descriptions of the entered endpoints, not a continuous metamorphic growth model.
Calculation method
Calculation and interpretation
Describe how one consistently defined positive size trait differs across declared metamorphic endpoints without treating metamorphosis as continuous exponential growth or assigning a fitness meaning.
Endpoint ratio Ri = Yexit/Yentry; log change ΔlnYi = ln(Ri); interval log rate gi = ΔlnYi/Δti
Worked example
Retain an amphibian mass decrease
The negative record is retained as an endpoint trait change and does not identify a cause or fitness consequence.
Endpoint ratio Ri = Yexit/Yentry; log change ΔlnYi = ln(Ri); interval log rate gi = ΔlnYi/Δti
Supported inputs
Precision and limits
Endpoint description
The calculation does not reconstruct within-transition trajectories, stage timing, tissue allocation, feeding, water balance or mechanism.
One consistently defined trait
Every row must use the same positive trait and unit convention at entry and exit. Length, mass, area and volume are never mixed.
Pairing remains declared
Group summaries do not become paired individuals, and surviving or observed endpoints may not represent the entry population.
No performance interpretation
Trait gain, loss or rate does not establish successful metamorphosis, fitness, survival probability, health or developmental normality.
Numerical support
Positive endpoint traits support 10⁻¹² through 10¹², elapsed times support 10⁻¹² through 10¹² in the selected unit, and a calculation retains 1–20 records. A 32-machine-epsilon log-response rule resolves computational zero without defining biological equivalence.
Continue calculating
Related calculators