Step 1. Load the packages | Step 2. Look at a data set | Step 3. Add up over units | Step 4. Measure how "bursty" each series is | Step 5. Ask the key question | Step 6. Draw it | Step 7. See what a delay would actually predict | Step 8. The careful version of the test | Step 9. A check with a record where a delay is the truth | Step 10. The other way round: a seasonal plant | Step 11. What the four cases teach | Step 12. Your own data

The six records | The ceiling | Why the pooled comparison has no power | The host-level test | Is the host-level test calibrated? | What to take from the set

The one distinction that matters | The three routes | The cluster bootstrap: a confidence interval | The Bayesian bootstrap: a credible interval, same quantity | The negative-binomial model: a credible interval, different quantity | The four ways to cut an interval from a set of draws | The two probability statements, side by side | Choosing, and reporting | References

The two records | Concentration and the ceiling | What any lag profile predicts | How firm the exceedance is | Why the pooled comparison cannot decide | The host-level test | The verdict, keyed on labels

1. The hypothesis as a convolution | 2. The ceiling: averaging cannot concentrate | 3. The second diagnosis: the lag is unidentifiable | 4. Calibrating the pooled comparison | 5. The host-level null, stated as the hypothesis is asserted | 6. What the test does and does not say | Function map | References

1. The hypothesis as a convolution | 2. The ceiling: averaging cannot concentrate | 3. The second diagnosis: the lag is unidentifiable | 4. Calibrating the pooled comparison | 5. The host-level null, stated as the hypothesis is asserted | 6. What the test does and does not say | Function map | References

Introduction: What the Yonezawa (2000) model does | Purpose and scope | What the model does | Biological system | Notation and key formulas | Variables (for $s$ stages) | The one-year variance of allele-frequency change (Eq. 5) | General generation-time effective size $N_e$ (Eq. 6) | Clonal-dominant Poisson special case (Eq. 10) | Annual effective size $N_y$ (Eq. 11) | Glossary of R objects | Step 1 — Load packages and define pedagogical helpers | Step 2 — Enter Table 2 data (Miz and Nan populations) | Step 3 — Compute $\lambda$, stable stage distribution, and generation time $L$ | Step 4 — Manual computation of Table 4 quantities | Observed stage fractions | Expected (equilibrium) stage fractions | Step 5 — Replication check: computed vs. paper Table 4 | Step 6 — NeStage package: Ne_clonal_Y2000() | Inputs explained | Running Ne_clonal_Y2000() for both populations | Print output for Miz (observed D) | Comparing Ne_clonal_Y2000() output to paper Table 4 | Step 7 — General model: Ne_mixed_Y2000() | Internal consistency: general model must match clonal shortcut | Sensitivity to the clonal fraction $d_i$ | Final summary: full replication of Table 4 | Session information

Step 1. Load the packages | Step 2. Look at a data set | Step 3. Add up over units | Step 4. Measure how "bursty" each series is | Step 5. Ask the key question | Step 6. Draw it | Step 7. See what a delay would actually predict | Step 8. The careful version of the test | Step 9. A check with a record where a delay is the truth | Step 10. The other way round: a seasonal plant | Step 11. What the four cases teach | Step 12. Your own data

The six records | The ceiling | Why the pooled comparison has no power | The host-level test | Is the host-level test calibrated? | What to take from the set

The one distinction that matters | The three routes | The cluster bootstrap: a confidence interval | The Bayesian bootstrap: a credible interval, same quantity | The negative-binomial model: a credible interval, different quantity | The four ways to cut an interval from a set of draws | The two probability statements, side by side | Choosing, and reporting | References

The two records | Concentration and the ceiling | What any lag profile predicts | How firm the exceedance is | Why the pooled comparison cannot decide | The host-level test | The verdict, keyed on labels

Overview | What data do you need? | T_mat — Survival/transition matrix | F_vec — Fecundity vector | D — Stage frequency distribution | L — Generation time (optional but recommended) | Quick data checklist | Ne_clonal_Y2000() — Purely clonal populations | Key outputs | Ne_sexual_Y2000() — Purely sexual populations | Optional: adjust reproductive variance Vk_over_k | Ne_mixed_Y2000() — Mixed sexual + clonal populations | The d vector — most common source of confusion | Which function should I use? | Interpreting the results | Next steps | References

Introduction | From a ratio to a decision | The four sensitivity functions | Elasticity: a standardised sensitivity measure | Study populations | Sensitivity to sexual reproductive variance: Ne_sensitivity_Vk() | What is Vk/k̄ and why does it matter? | Running Ne_sensitivity_Vk() for the reproductive stage | Interpretation | Sensitivity to generation time: Ne_sensitivity_L() | Why L matters so much | Sweeping L across a biologically plausible range | Key insight: Ne/N is hyperbolic in L | Sensitivity to clonal fraction: Ne_sensitivity_d() | When does clonal fraction matter? | Cross-parameter comparison: which lever matters most? | Conservation prioritisation framework | Priority 1 — Reduce uncertainty in L | Priority 2 — Equalise reproductive output in high-fecundity stages | Priority 3 — Augment census size in the lowest-Ne/N populations | Session information | References

Introduction | What is effective population size and why does it matter? | Why stage-structured models? | The three NeStage functions | MatU and MatF: the standard input format | The study system: Lepanthes orchids in Puerto Rico | The six-stage life cycle of L. rupestris | Setting up the matrices | L. rupestris population 1 — monthly matrices | Stage frequency vector D | Monthly vs annual matrices: does the time step matter? | Converting monthly to annual matrices | Comparing Ne/N from monthly vs annual matrices | Interpreting the difference | Running Ne_sexual_Y2000: a complete worked example | All inputs explained | Full output for L. rupestris population 1 | Reading the output | All populations of all three species | Entering all matrices | Computing Ne/N for all 17 populations | Comparing to Orive (1993) estimates | Summary and conservation implications | Key conservation messages | Brief overview of the other two functions | Ne_clonal_Y2000 | Ne_mixed_Y2000 | Session information | References