Seed dormancy is an essential trait widely reported in ecological studies. Seed dormancy often influences the ecological dynamics of plant communities by determining germination timing (Baskin and Baskin, Reference Baskin and Baskin2014). Recent work by Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023) used light microscopy and scanning electron microscopy to describe the morphology and anatomy of seeds from 14 plant species in the Andean forests of Colombia, while also determining dormancy class with these techniques and germination trials. The authors found dormancy in 85% of their study species, three of which are in the genera Cecropia and were classified as having both physiological and physical dormancy (referred to as physiophysical dormancy). Seeds with physiophysical dormancy have impermeable coats, a fully developed embryo, and a mean germination time (MGT) of over 30 days (Rojas-Villa and Quijano-Abril, Reference Rojas-Villa and Quijano-Abril2023). Here, we present seed permeability data measured by fluorescent dye uptake into the seed endosperm and germination trails from a Panamanian lowland population of Cecropia peltata (one of the species studied by Rojas-Villa & Quijano-Abril, Reference Rojas-Villa and Quijano-Abril2023), as well as Ochroma pyramidale, Cecropia longipes, and Cecropia insignis. Our data do not support the presence of physical dormancy in seeds of Cecropia, while O. pyramidale does exhibit physical dormancy.

Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023) provide extensive anatomical descriptions of the Cecropia fruits and seeds included in their study. These descriptions, along with MGT of 30 days or more, were used to determine if seeds exhibited dormancy. Support for their dormancy classification of physiophysical dormancy in Cecropia was provided by the presence of a seed coat derived from two integuments, a ~0.14 mm thick membrane, and one to two layers of cuboidal cells with tannins present (Rojas-Villa and Quijano-Abril, Reference Rojas-Villa and Quijano-Abril2023). The three Cecropia species had MGT varying from 39.35 ± 7.36 days for C. angustifolia to 63.73 ± 6 for C. peltata, with C. telealba averaging 46.21 ± 9.59 days. Although the authors mentioned the inclusion of imbibition tests in their study, no seed permeability data were reported for any of the Cecropia species.

There are several ways to test the permeability of seeds. Given that physiophysical dormancy and physical dormancy, by definition, have impermeable seed coats, a test of permeability is needed to determine if a species exhibits these dormancy types. Permeability can be tested using imbibition or conductivity tests, and/or via microscopy by measuring fluorescent dye uptake by seeds (Tieu and Egerton-Warburton, Reference Tieu and Egerton-Warburton2000; Chen et al., Reference Chen, Zhang, Baskin and Hu2019). Imbibition tests quantify the increase in seed mass when seeds are exposed to water. For impermeable seeds, seed mass should not change after water exposure. However, the imbibition test may have limitations for seeds with small masses, as the change in mass may be difficult to detect for individual seeds. Furthermore, because a single population of seeds may contain intact impermeable seeds, and damaged permeable seeds, imbibition tests based on a pooled sample of multiple seeds may be difficult to interpret.

The electrical conductivity test measures the condition of the seed-protecting tissues, where low conductivity suggests there is little electrolyte leaching from the seed interior, and the seeds are likely impermeable or of high quality (Matthews and Powell, Reference Matthews and Powell2006). These tests may have similar limitations for small seeds, as the signal-to-noise ratio may be challenging to interpret when electrolyte leaching from small seeds may result in low conductivity rates.

Seed permeability can also be quantified by the extent to which a low molecular weight fluorescent dye penetrates inside of the seed coat (Fig. 1A). Fluorophore seed penetration is often quantified through fluorescence microscopy after 24 to 48 h of exposure. This dye can potentially permeate the seed anywhere, leading to a potential underestimation of permeability if the dye permeates into a location of the seed that is not visible from where the researcher has cut the seed before microscopy processing.

Figure 1. Example images (a) from C. insignis, a species with nondormant seeds, and O. pyramidale, a species with physically dormant seeds, with seed permeability scores of both 0 and 1. Bar plots (b) showing the percentage of permeable seeds quantified by Lucifer Yellow (LY) uptake from three tree species with nondormant seeds, C. insignis, C. longipes, C. peltata, and one tree species with physically dormant seeds, O. pyramidale. Seeds are scored 0 if the seed is classified as impermeable or 1 if the seed is classified as permeable. Numbers at the top of each bar in (b) indicate the sample size for each category.

Here, we used Lucifer Yellow (LY) dye to determine seed permeability because LY has a low molecular weight compared to other water-soluble fluorophores, making it especially useful for measuring seed permeability (Tieu and Egerton-Warburton, Reference Tieu and Egerton-Warburton2000). Combining data on seed permeability and germination trials on seeds collected on lowland tropical forests in Panama, we suggest that (1) permeability tests should be included when determining physical dormancy, and (2) a need for standardized and/or reported germination trail conditions (e.g., temperature, moisture, and light). In Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023), all the Cecropia species exhibit physiophysical dormancy and they suggest this may be a characteristic of the genus. We find high levels of seed permeability in all three of the Cecropia species in our study (Fig. 1B), one of which, C. peltata, is present in both studies. In comparison, Ochroma pyramidale, a species with known physical dormancy (Sautu et al., Reference Sautu, Baskin, Baskin, Deago and Condit2007; Zalamea et al., Reference Zalamea, Sarmiento, Arnold, Davis and Dalling2015), shows that only ~50% of seeds are permeable. Using four species with physically dormant seeds (including O. pyramidale), Zalamea et al. (Reference Zalamea, Sarmiento, Arnold, Davis and Dalling2015) showed that fresh seeds (i.e., fully mature seeds that have not been dispersed away from the maternal tree) often consist of two types: one representing a permeable (nondormant) fraction of seeds that is capable of germinating immediately, and a second fraction that is impermeable (dormant) and accounts for the seeds that persist in the soil seed bank.

Further, Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023) define species as dormant if the MGT is greater than 30 days. However, several environmental factors can influence MGT for a given species. These factors need to be reported in papers classifying dormancy types on seeds to avoid confounding results. Mean germination time may increase under colder conditions (Sampayo-Maldonado et al., Reference Sampayo-Maldonado, Ordoñez-Salanueva, Mattana, Ulian, Way, Castillo-Lorenzo, Dávila-Aranda, Lira-Saade, Téllez-Valdéz, Rodriguez-Arevalo and Flores-Ortíz2019), such as those in Andean montane forests in Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023). For example, the time required for germination of 50% of Cedrela odorata seeds was 13–42 days longer in temperature treatments of less than 20°C compared to treatments above 20°C (Sampayo-Maldonado et al., Reference Sampayo-Maldonado, Ordoñez-Salanueva, Mattana, Ulian, Way, Castillo-Lorenzo, Dávila-Aranda, Lira-Saade, Téllez-Valdéz, Rodriguez-Arevalo and Flores-Ortíz2019). Germination trials (see the methods at the end for details on the environmental conditions of these trials) with Cecropia species (and O. pyramidale as a physically dormant species as a comparison) in Panama showed the MGT is substantially less than 30 days (C. Insignis = 18.9 ± 0.6; C. longipes = 15.6 ± 0.8, C. peltata = 13.6 ± 0.2, O. pyramidale = 10.9 ± 0.7; Fig. 2). It is difficult to disentangle whether the differences in MGT between this study and Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023), especially for C. peltata used in both studies, are from differences in the physical environment of the germination trials or potential differences in sub-populations within species. This suggests caution should be taken using MGT as an indicator of physical dormancy without some standardization for comparison between studies and the additional use of permeability tests.

Figure 2. Germination time in days for three tree species with nondormant seeds, C. insignis, C. longipes, C. peltata, shown in gold and one tree species with physically dormant seeds, O. pyramidale shown in gray. Mean and standard error of C. peltata* from Rojas-Villa & Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023) are also shown for comparison between studies. Percent of seeds that germinated in germination trials for each species is shown in bold at the top of the figure. Below these percentages, we provide the percentage of seeds that did not germinate but were determined to be viable through tetrazolium staining for the four species in our study.

We also find that O. pyramidale, a physically dormant species, had the fastest germination time but the lowest percentage of seeds that germinated (27% compared to 44% or higher found in Cecropia species; Fig. 2). Further, 66.5% of the seeds in O. pyramidale that did not germinate after 6 weeks in the germination trials were determined to be viable using tetrazolium testing compared to the Cecropia species where only 1.5–2.5% of the non-germinated seeds were viable. This adds further evidence that there is a subset of seeds in physically dormant species, such as O. pyramidale, that do not exhibit dormancy.

Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023) comprehensively explain the functional role of physiophysical dormancy in Cecropia. They suggested that dormancy allows these seeds to persist in the soil for long periods of time until suitable (gap or high light) conditions arise. However in other work, we found that seeds of Trema micrantha and Zanthoxylum ekmanii, that have physiologically dormant seeds at the time of dispersal, exhibit seed dormancy alleviation in decades-old viable seeds extracted from the soil seed bank (Zalamea et al., Reference Zalamea, Sarmiento, Arnold, Kuo, Delevich, Davis, Brown and Dalling2023). Never-buried, fresh seeds had a lower proportion of germinated seeds in germination trails, compared to seeds that persisted in the soil for decades, despite both seed types having a similar proportion of viable seeds (determined by germination trails and tetrazolium testing). This suggests that seeds can persist in the soil seed bank even after dormancy begins to be alleviated. In a 30-month long seed burial experiment including the three species of Cecropia included in this study and Jacaranda copaia (also a nondormant species; Sautu et al., Reference Sautu, Baskin, Baskin, Deago and Condit2007), we found that the proportion of seeds that can germinate decreased gradually over time (Zalamea et al., Reference Zalamea, Sarmiento, Arnold, Davis, Ferrer and Dalling2021). Although the proportion of viable seeds after 30 months of soil incubation was consistently low and varied among localities and species, we were able to confirm that some seeds of three species of Cecropia and J. copaia persisted in a viable state in the soil without being species that exhibit seed dormancy.

Plant biologists have used morphological and anatomical studies for centuries to characterize and understand the diversity of plant form and function (Haberlandt, Reference Haberlandt1914). Rojas-Villa and Quijano-Abril's work extensively characterizes anatomical and morphological characteristics of seeds from 14 different species and links these traits to dormancy classes for these species. Contributions like those of Rojas-Villa and Quijano-Abril (Reference Rojas-Villa and Quijano-Abril2023) are extremely valuable to better understand how seeds of tropical species interact with the physical environment in which they germinate and recruit into as seedlings. In the case of Cecropia species, Rojas-Villa and Quijano-Abril's work built upon the pioneering and detailed morphological and anatomical study of Lobova et al. (Reference Lobova, Mori, Blanchard, Peckham and Charles-Dominique2003) that used fruits and seeds of Cecropia species from French Guiana. Lobova et al. (Reference Lobova, Mori, Blanchard, Peckham and Charles-Dominique2003) showed that in Cecropia fruits, the pericarp may play a key protective role that is functionally more important than the two thin layers of integuments that form the seed coat. Lobova et al. (Reference Lobova, Mori, Blanchard, Peckham and Charles-Dominique2003) hypothesized without any seed permeability measurements that Cecropia fruits may not be completely impermeable because the vascular bundle penetrates the pericarp at the pedicel scar; however, they concluded that seed impermeability may have evolved to allow seeds to remain dormant in the soil seed bank. As discussed in this study and provided empirical evidence in previous contributions (i.e., Zalamea et al., Reference Zalamea, Sarmiento, Arnold, Davis, Ferrer and Dalling2021), Cecropia seeds do not require dormancy to persist in the soil seed bank. In this contribution, we used seed permeability and germination data collected from four Neotropical pioneer tree species to demonstrate the need for formal testing of permeability when assigning physical dormancy to species. In addition, we also highlight the need to report the environmental conditions in which the germination trials are implemented to make studies comparable and replicable in the future.