Two seed pathways from the same plant
Mammillaria pectinifera forms a small globose stem covered with appressed white spines. Flowers develop in whorls near the apex. Their whitish fruits occupy spaces between tubercles, and successive years of fruit production can be identified by their position on the stem. The adult cactus thus retains a record of several reproductive seasons. Peters, Martorell and Ezcurra used this structure to investigate serotiny, the delayed dispersal of mature seeds. [Peters et al. 2009, pp. 537–538, Fig. 1]
A young fruit can be pushed out of the stem with its entire contents. A retained fruit remains in the plant, gradually loses seeds through an opening in its upper part and, once emptied, persists as a dry remnant. These pathways separate seed access to the substrate over time. One releases the contents of a new fruit in a single event, while the other spreads release over successive seasons. In the observed population, actively ejected fruits always belonged to the youngest year's cohort. [Peters et al. 2009, pp. 537–539, Fig. 1]
Two seasons with different water availability
The study was conducted near Texcala in the Tehuacán Valley in Mexico. In an area of 238 m², 81 adult plants of different sizes were marked. Newly produced and ejected fruits were counted monthly from June 1997 to May 1999. Seedlings were sought after the summer rainy season in 55 randomly selected quadrats of 1 m². Their survival was then tracked until May, covering the dry season following the rains. [Peters et al. 2009, p. 538]
Rainfall was 267 mm in the first hydrological year and 817 mm in the second. The long-term mean reported in the paper was 424 mm. The comparison therefore included a distinctly dry and a distinctly wet season. Mean production was 6,1 fruits per plant per year, and the difference in fruit numbers produced between years was not statistically significant. The main change concerned how fruits were released. No ejection of whole fruits was recorded in the 1997 season. In 1998, 21,5% of new fruits were ejected with the arrival of heavy rains. [Peters et al. 2009, pp. 538–539]
A seed bank retained on the cactus
A total of 615 fruits were collected from 20 adult plants, with their age determined from annual whorls. Older fruits contained progressively fewer seeds. In the results, the authors report that a mean of 42,6% of seeds remained in unejected fruits in the first year. Seeds were already rare in fruits older than four years. Retaining fruits therefore maintains a seed supply on the plant while allowing it to empty gradually. [Peters et al. 2009, pp. 538–539, Fig. 3]
Seeds from individual yearly cohorts were placed on agar at 25°C with a twelve-hour photoperiod. Germination was assessed after 30 days. A mean of approximately 46% of seeds germinated, and no significant decline in this value with age was demonstrated in the available sample. Seed retention and the ability to germinate are separate characteristics. The first describes the size of the reserve remaining in fruits, and the second indicates what proportion of recovered seeds germinates under the test conditions used. [Peters et al. 2009, pp. 538–539]
Watering as a controlled cue
30 adult plants with recently formed fruits were transferred to a greenhouse. Two plants were randomly assigned to each watering treatment. Doses ranged from 0 to 30 mm, supplied three times a week for 21 weeks. Increasing the amount of water supplied increased the probability of fruit ejection. This experiment tested the response to water availability independently of the comparison of two years in nature. [Peters et al. 2009, pp. 538–540, Fig. 4]
The greenhouse figure caption indicates that, with cumulative watering of at least 600 mm over 21 weeks, the model predicted the release of almost all fruits. The results text describes more than 1000 mm of simulated annual rainfall. The two values have different time frames, and evaporation in the greenhouse was lower than in the field. The results describe the response of plants in the experiment. An appropriate conversion to natural rainfall requires consideration of its distribution, infiltration into the soil and loss between rains. [Peters et al. 2009, pp. 539–540, Fig. 4]
From a released seed to a seedling after the dry season
The numbers of seeds released in the field were estimated from fruiting and retention data, assuming approximately constant production and negligible fruit ejection in the two preceding dry years. The number of seedlings surviving the dry season was extrapolated from quadrats to the entire 238 m² patch. The ratio of these quantities describes seedling establishment after seed release. It includes germination and subsequent survival together. [Peters et al. 2009, pp. 539–540, Table 1]
| Quantity | 1997–1998 | 1998–1999 |
|---|---|---|
| Seeds released from retained fruits | 35 508 | 32 259 |
| Seeds released from ejected fruits | 0 | 7634 |
| All released seeds, estimate | 35 508 | 39 893 |
| Seedlings surviving the following dry season, estimate | 17 | 95 |
| Seedling establishment success, ‰ | 0,47 | 2,38 |
The source of the table data is shared by both seasons. In the wet year, seedling establishment success was approximately five times higher. The values 0,47 and 2,38 are per mille, so they concern approximately 0,047% and 0,238% of released seeds, respectively. Seedling numbers remained small even in the more favourable season. High seed production and gradual delivery to the substrate may therefore work together to maintain successive generations in a habitat with high mortality of young plants. [Peters et al. 2009, p. 540, Table 1]
Benefits of two dispersal rhythms
The authors interpret gradual seed release as spreading risk between years. A distinctly wet period additionally triggers the ejection of new fruits, delivering more seeds when the chance of seedling establishment is greater. Field observations and the response to watering support the adaptive significance of this combination. The description itself does not determine which cellular signals and tissue changes cause the mechanical extrusion of a fruit. The physical effect is documented, while its detailed regulation remains a further level of explanation. [Peters et al. 2009, pp. 539–540]
Collection records benefit from distinguishing empty remnants, retained fruits and newly released fruits. Their presence on the stem represents different developmental stages, and seeds may come from different seasons. The experiment described also supports recording the collection year and method of obtaining seeds. This information allows subsequent germination results to be compared while retaining the material's history. [Peters et al. 2009, pp. 538–539]
The adaptive value of cued seed dispersal in desert plants: Seed retention and release in Mammillaria pectinifera (Cactaceae), a small globose cactus
Peters E. M., Martorell C., Ezcurra E., The adaptive value of cued seed dispersal in desert plants: Seed retention and release in Mammillaria pectinifera (Cactaceae), a small globose cactus, American Journal of Botany, 2009, 96(2):537-541, doi: 10.3732/ajb.0800157
The study of serotiny in Mammillaria pectinifera combines two-year observations of a population near Texcala with a greenhouse watering experiment. The authors determine seed content in fruits of different ages, seed germination, release of new fruits and seedling establishment after the dry season. The paper explains the interaction of gradual seed dispersal and expulsion of entire fruits in a wet year.
Plants of the World Online — Royal Botanic Gardens, Kew
Accepted names, synonyms and distribution. A citation beside a taxon leads to its record.
Accessed 2026-10-08.
