# Inter-stride variability triggers gait transitions in mammals and birds

> Why does a walking animal suddenly break into a trot? Comparing nine species, from opossums and dogs to otters and ostriches, researchers found that saving energy does not explain the timing. Instead, animals seem to switch gait when their strides grow uneven, moving into a steadier rhythm that may help them avoid tripping or falling.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2018-granatosky-bryce-inter-stride-variability-triggers/

## Citation and access

- **Authors:** Michael C. Granatosky, Caleb M. Bryce, Jandy Hanna, Aidan Fitzsimons, Myra F. Laird, Kelsey Stilson, Christine E. Wall and Callum F. Ross
- **Year published:** 2018
- **Journal or source:** Proceedings of the Royal Society B: Biological Sciences
- **Full citation:** Granatosky, M. C., Bryce, C. M., Hanna, J., Fitzsimons, A., Laird, M. F., Stilson, K., Wall, C. E. and Ross, C. F. 2018. Inter-stride variability triggers gait transitions in mammals and birds. Proceedings of the Royal Society B: Biological Sciences 285(1893): 20181766.
- **DOI:** 10.1098/rspb.2018.1766
- **Access status:** Hosted (freely available elsewhere)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2018_Granatosky_and_Bryce_etal_inter-stride-variability-triggers-gait-transitions-in-mammals-and-birds.pdf

## Tags

- **Species:** Mammals > American mink; Mammals > Australian water rat; Mammals > Brush-tailed bettong; Mammals > Domestic dog; Mammals > North American river otter; Birds > Ostrich; Birds > Svalbard rock ptarmigan; Mammals > Tufted capuchin; Mammals > Virginia opossum
- **Region:** North America > United States
- **Tracking:** Gaits
- **Behaviour and ecology:** Behaviour; Movement
- **Methods and technology:** Chi-square test; Mann–Whitney U test; Simulation
- **Data type:** Existing data; New field data
- **Observer qualification:** Not needed
- **Discipline:** Behaviour and ecology
- **Year published:** 2018

## Detailed keywords

- **Species:** Studied with new treadmill data: Virginia opossum (Didelphis virginiana) and tufted capuchin (Sapajus apella). Studied with data from the authors’ earlier study: domestic dog (Canis lupus familiaris), in three breed groups (northern breeds, hounds and retrievers). Studied with data taken from other published studies: brush-tailed bettong (Bettongia penicillata), Australian water rat (Hydromys chrysogaster), American mink (Neovison vison), North American river otter (Lontra canadensis), Svalbard rock ptarmigan (Lagopus muta hyperborea; spelled “hyperborean” in the figure caption) and common ostrich (Struthio camelus). Humans are discussed from earlier studies only; horses are mentioned from the veterinary literature on limbs striking each other.
- **Location and habitat:** Laboratory treadmill inside a clear Plexiglas metabolic chamber, in temperature-controlled rooms (22 to 24 °C). The study place is not stated; animal protocols were approved by the University of Chicago and the University of California, Santa Cruz (United States). Wild ostrich data from another study are used as a check on natural settings.
- **Topics:** gaits; gait transitions; walk to trot, walk to bound and walk to grounded run transitions; preferred transition speed (PTS); energetically optimal transition speed (EOTS); cost of transport (COT); stride duration variability (inter-stride variability); locomotor rhythmicity; dynamic stability; energy minimization hypothesis; dynamic stability hypothesis; interlimb interference, tripping and falling; grounded running in birds; treadmill training.
- **Statistics or models:** Coefficient of variation of stride cycle duration corrected for sample size (CV*); linear and quadratic curve fitting of cost of transport against speed (coefficient of determination, R²); Mann–Whitney U test; chi-square test; Kruskal–Wallis test (on dog limb length, reported with an F value, p = 0.052); jack-knife resampling; simulated data points from published means and standard deviations (normal random number generator); figure data extracted with DataThief III; analysis in MATLAB. Inconsistency: Table 1 lists the brush-tailed bettong’s transition as “quadrupedal bound to bipedal hop” (an en dash in the table), while the text and Figure 1 treat it as a walk to bound transition.
- **Data type:** New experimental data (treadmill and open flow respirometry) for opossums and capuchins; existing data for domestic dogs (from the authors’ earlier study) and for six species from eight published studies. No dates of data collection given. Animals were trained for about six weeks before testing.

## Synopsis

### 1. What they did

As animals speed up, they switch gaits, for example from a walk to a trot. The most common explanation is that they switch to save energy. But it is hard to see how an animal could sense, from moment to moment, that a new gait would be cheaper. The authors tested a second idea: that animals change gait when their steps become irregular and unsteady, and switch to regain a steady rhythm.

They set out two competing hypotheses. If gaits change to reduce variability, then stride-to-stride variation should be highest just before the switch and drop after it. If gaits change to save energy, then the speed at which animals choose to switch should match the speed at which switching becomes cheaper, and energy use should always drop after the switch.

- New experiments. Virginia opossums and tufted capuchins were trained for about six weeks to walk and trot on a treadmill inside a clear chamber that measured their oxygen use. The speed was raised every 15 seconds to find where each animal switched from a walk to a trot, filmed from the side. For each speed, 30 strides were chosen at random and the variation in stride duration was calculated.

- Earlier data. Data for domestic dogs (northern breeds, hounds and retrievers) came from an earlier study by one of the authors.

- Published data. They added six more species from eight published studies: brush-tailed bettong, Australian water rat, American mink, North American river otter, Svalbard rock ptarmigan and common ostrich. Because raw data are rarely published, they rebuilt data points from reported averages and figures and resampled them.

- Comparisons. For each species they compared the energy cost of moving a kilogram one metre (the cost of transport) and the variation in stride duration, before and after the gait change.

In birds, which do not trot, they looked at the switch from walking to “grounded running”, a running style with no moment when both feet are off the ground.

### 2. What they found

Energy

- In opossums, capuchins, bettongs and water rats, switching from a walk to a trot or bound significantly lowered the energy cost.

- In domestic dogs, ptarmigan, ostriches and river otters, the energy cost did not change significantly after the switch.

- In capuchins, opossums and dogs, the speed at which switching would save energy was significantly lower than the speed at which the animals actually switched. Mink and ostriches also kept walking past the energy-saving point, while ptarmigan switched before it.

- For river otters, bettongs and water rats, the energy-saving speed did fall within the range of speeds where they switched.

- Of the nine species, only the bettong and the water rat met both energy predictions.

Stride variability

- In all species, variation in stride duration changed with speed, and in almost all species it peaked near the speed where the animals switched gaits.

- In every species except the Australian water rat, variation dropped significantly after the switch.

- In capuchins, opossums and dogs, the speed with the most variation did not differ significantly from the switching speed. In mink, ptarmigan, ostriches, otters and bettongs, the peak in variation fell within the switching range far more often than expected by chance.

- The water rat was the exception: its peak in variation fell within the switching range less often than expected.

- The pattern held across birds and mammals, across body sizes from 0.67 to 114 kg, and across walk to trot and walk to bound transitions.

- Data from wild ostriches, from another study, also showed a large spike in stride variation just before grounded running began.

### 3. What they concluded

- Energy savings do not reliably predict when mammals and birds change gait.

- Instead, animals seem to switch gait when their strides become too variable, moving into a steadier, more rhythmic and more predictable way of moving. This may lower the risk of limbs striking each other, tripping or falling.

- Saving energy is an important benefit of changing gait, but it is probably secondary. The trigger is likely the loss of steadiness. Irregular strides may also cost more energy, which links the two ideas.

- Birds and mammals have probably each evolved ways of sensing stride steadiness, and this is worth further study.

- Limitations given by the authors: rebuilding published data rests on assumptions that are hard to check; almost all data come from treadmills, not natural settings; and only the switch from walking to a faster gait was studied, not the trot to gallop or the trot to walk change.

### 6. Other technologies, skills or innovations

- Treadmill inside a clear Plexiglas metabolic chamber, with open flow respirometry (Field Metabolic System, Sable Systems International) measuring oxygen use and carbon dioxide output.

- Side-view video to measure the share of each stride a foot is on the ground (duty factor) and to confirm the gait near the switching speed.

- A method for rebuilding and resampling data from published studies (DataThief III to read values from figures, simulated data points and jack-knife resampling in MATLAB).

## Cite this summary

Lawrence, K. 2026. Summary of: Inter-stride variability triggers gait transitions in mammals and birds (Granatosky and Bryce et al., 2018). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2018-granatosky-bryce-inter-stride-variability-triggers/