Pediocactus in the landscape. Habitats, weather and population recruitment

Basalt sites in the northwest

Pediocactus nigrispinus occurs in Washington, Oregon and Idaho (also as P. simpsonii var. nigrispinus in older literature). Dillon-Zuppelli and co-authors compared seventeen known sites in the first two states. Substrates were predominantly volcanic, and the cacti grew with bunchgrasses. Poa secunda and Pseudoroegneria spicata occurred in all plots studied. Shrubs and trees varied across the landscape. Artemisia rigida was common in Washington, and Juniperus occidentalis in Oregon. [2-WWW][Dillon-Zuppelli et al. 2025, pp. 65–68]

Ground cover, vegetation, soil properties and climate were compared with cactus density, size and fruit numbers. Differences between states in vegetation height and slope did not provide a simple explanation of differences in cactus diameter or density. Fruiting was greater in the Oregon plots. The study covered occupied sites. The absence of plots without cacti limits the identification of conditions determining where the species can persist. [Dillon-Zuppelli et al. 2025, pp. 65–68, 72–74]

What successive censuses showed

Demographic monitoring was conducted in four Wild Horse plots in Washington. Censuses were carried out in 2016, 2017, 2018, 2019 and 2022, with a gap during two pandemic years. The results report a decline in the number of recorded plants from 415 to 344. No new seedlings were detected throughout this series. Stem clusters received a shared identification when it was impossible to determine whether they were offsets of one plant or neighbouring individuals. The recording unit therefore also had a practical field definition. [Dillon-Zuppelli et al. 2025, pp. 63–65, 68–71]

Stems were divided into diameter classes. The smallest, below 1 cm, corresponded to seedlings. The next class included non-fruiting plants of 1–4 cm, and the larger, potentially reproductive classes spanned 4–7 cm, 7–10 cm and above 10 cm. A matrix describes the probabilities of survival and transition between these classes. A living plant remaining in the same class was termed stasis. This includes a situation in which the cactus persists, although its growth does not move it into a higher interval. [Dillon-Zuppelli et al. 2025, pp. 64–65, 69–71]

In three successive one-year transitions, λ was 0,946, 0,935 and 0,974. Each value is below one and indicates a modelled decline if the corresponding combination of processes is maintained. The paper's final matrix covers the longer transition of 2019–2022. It is not another annual interval that can be compared without conversion. Adult stasis had high elasticity, especially for plants with a diameter of 7–10 cm. Changes in their survival strongly affected the model outcome. [Dillon-Zuppelli et al. 2025, pp. 69–75, Table 3–4]

From a viable seed to a new plant

Fruiting, germination and entry of a seedling into the population record describe successive stages. Between them, seeds may remain in the substrate, be moved or die. A seedling may appear between censuses and fail to survive until the next one. The absence of detected seedlings at Wild Horse is therefore an observation of absent recruitment in that series. It does not by itself determine the stage at which limitation occurs. [Dillon-Zuppelli et al. 2025, pp. 64–65, 73–75]

In a greenhouse trial, the authors obtained 79% for germination combined with survival to six months. Seeds were sown in a mixture of mineral components and horticultural substrate. Air temperatures ranged from 21–38°C. Watering was provided three times a week for the first two months and once a month thereafter. The result documents a specific procedure. The absence of new seedlings in the field meant that the recruitment contribution to the matrix was estimated using additional assumptions, without being determined from observed additions to the population. [Dillon-Zuppelli et al. 2025, pp. 64–65, 73–75]

Assumptions about recruitment matter for long-term projections of abundance. When the arrival of new plants remains low for a long time, the population depends on the survival of existing specimens. The model's sensitivity to adult classes identifies a particularly important process in this sample. A projection based on a short series nevertheless retains the estimated probabilities into the future. It does not determine a certain date for the disappearance of all sites of the species. [Dillon-Zuppelli et al. 2025, pp. 70–75]

Twenty years in one Utah plot

Pediocactus winkleri (formerly also P. bradyi subsp. winkleri) was observed in Capitol Reef during 1995–2014. The plot measured 10 × 10 m and lay at 1823 m above sea level. A total of 107 plants were marked throughout the series. The cacti emerged in spring, produced seeds by June and then retreated underground. Some remained invisible in subsequent years. [2-WWW][Clark et al. 2015, pp. 89–92]

In the first year, 44 living plants were counted. At the end of observation, there were 31 visible living plants and five of undetermined status, known to have been alive in the previous year. These categories retain information about seasonal burial. When a marked stem reappears, its return can be assigned to the history of an existing plant. New seedlings are recorded separately. This allows actual recruitment to be distinguished from changes in detectability. [Clark et al. 2015, pp. 92–96]

The comparison includes different species, sample sizes and observation periods. Three plots in Arizona, four in Washington and one in Utah are not a random sample of the entire genus. The shared importance of adult survival emerges from several analyses, while detailed parameters remain specific to each one. [Rojas-Pimentel et al. 2026, pp. 2–7][Dillon-Zuppelli et al. 2025, pp. 63–75][Clark et al. 2015, pp. 89–99]

Effects of repeated trampling

Evidence of disturbance by ungulates was recorded around 74 of the 107 P. winkleri plants. In the series of events analysed, 190 of 205 were attributed to livestock, ten to elk, and five remained unidentified. Disturbance included hoof contact with a plant or its immediate surroundings. Such an event may change the position of the stem and the state of the ground surface. Subsequent survival data were linked to the history of disturbance. [Clark et al. 2015, pp. 90–96]

The model associated disturbance with an odds ratio of 1,95 for death and 0,47 for flowering. An odds ratio compares the ratio of the probability of an event to the probability of its absence. It is not a direct multiplier of the probability itself. In another comparison, the modelled probability of death was approximately 3% without disturbance, 9% with two disturbances and 17% with three disturbances within three years. The other weather variables were held at their means. [Clark et al. 2015, pp. 94–95, Fig. 3]

A total of 71 plants died during the series, of which 43 had previously recorded ungulate disturbance. This links a history of damage with death. A direct cause was not established for every death. The strength of the observations comes from repeated censuses and the inclusion of weather, while their scope remains limited to one plot. [Clark et al. 2015, pp. 92–99]

Different parts of the season, different relationships

In P. winkleri, greater precipitation during the preceding twelve months favoured flowering. Higher minimum temperatures in February and March were associated with the appearance of new plants, and high April maxima with mortality. The stage of the year changes the significance of temperature conditions. Early warming may coincide with the beginning of growth, while later heat occurs under a different state of the substrate and plants. The paper described statistical relationships without experimentally separating their physiological causes. [Clark et al. 2015, pp. 94–99]

Uneven flowering between plants is equally significant. Ten specimens produced 31% of all flowers recorded. The most productive grew beneath shrubs or between rocks, where hoof contact with stems may have been limited. Position and outcome were observed together. The study did not experimentally isolate shading, moisture and mechanical protection, so attributing the whole difference to one of these factors would go beyond the data. [Clark et al. 2015, pp. 96–99]

Microorganisms and flowering plants nearby

The habitat also includes small organisms. Around P. nigrispinus, bacterial 16S rRNA gene sequences were compared from nine biological soil crust samples near cacti and seven near grasses. Most microorganism groups were shared, while some occurred more frequently near cacti. The analysis documents community composition. Its effect on nutrient availability and seedling development requires a separate functional study. [Dillon-Zuppelli et al. 2025, pp. 66–67, 71–74]

In P. paradinei, a comparison of sites in 2005 and 2022 showed a decline in vegetation, particularly herbaceous plants flowering at the same time as the cactus. The authors identify possible consequences for bee habitats. The abstract does not present a measurement of declining pollination effectiveness. The surrounding vegetation is an element of the habitat assessed here, and its effect on reproduction remains a research question. [Colombo et al. 2026, article abstract]

Adult persistence and entry of the next generation

In the observed populations, an adult cactus can remain at a similar size for a long time. In P. winkleri, 86% of tracked plants did not increase in diameter or increased by less than a centimetre during the observation period. In P. nigrispinus, the model was particularly sensitive to the maintenance of living plants in adult classes. Such persistence provides flowers and seeds in successive seasons, but the subsequent history of the site also includes the successful passage of seeds through the seedling stage. [Clark et al. 2015, pp. 92–98][Dillon-Zuppelli et al. 2025, pp. 69–75]

A precise population description therefore combines several records. It includes the fates of marked plants, their size and flowering, new seedlings and ground-surface conditions. In species that retract their stems into the soil, information on burial and reappearance is added. This allows the seasonal absence of a visible body, halted recruitment and the actual loss of adults to retain separate meanings in the assessment of the same site. [Rojas-Pimentel et al. 2026, pp. 2–4][Dillon-Zuppelli et al. 2025, pp. 63–65][Clark et al. 2015, pp. 90–92]