For organisms feeding on plant materials, the proportion of non-C elements in their food may be more limiting than energy (Pokarzhevskii et al., 2003; Moe et al., 2005) and diet supplementation with high-quality resources may promote the development of animals feeding on dead plant matter (Filipiak and Weiner, 2014, 2016; Filipiak et al., 2016; Horvathova et al., 2016). Hence, it might be advantageous to include in the diet plant matter that is nutritive, easily available in considerable mass and relatively easily digestible, i.e., pollen. Indeed, it was suggested that, in detrital food webs, litter-decomposing fungi are colimited by the scarcity of N, P, and S in litter, which they mitigate by foraging on pollen and which allows them to complete litter decomposition (Stark, 1972; Staaf and Berg, 1982; Hutchison and Barron, 1997). Forests may produce pollen in masses reaching 100 to 1000 kg/ha (Greenfield, 1996), and the bulk of this mass is deposited onto forest floors and in lakes (Richerson et al., 1970; Proctor et al., 1996; Shumilovskikh et al., 2015), thus increasing the productivity of the ecosystem (Graham et al., 2006; Masclaux et al., 2013). The present study considers the ecological stoichiometry framework to relate data on pollen nutritional quality to the role of pollen in nutrient cycling within and between terrestrial and aquatic ecosystems.

Pine pollen and pine litter have been studied in sufficient depth to provide data that are appropriate for a comparative analysis of their elemental compositions in various pine species distributed worldwide. To that end, to comprehensively determine how pollen stoichiometry influences the nutritional mismatches experienced by detritivores, the pine forest ecosystem was considered as an exemplary food web in the present study. To determine whether pollen flux may promote the development of detritivores, the elemental compositions and stoichiometry of pollen, litter and detritivores were compared. The available data on the elemental composition of pine litter and pollen of various pine species inhabiting different forests worldwide were collected (details in Supplementary Table 2).

A comparison of the elemental composition of pine pollen with that of litter (Table 1) demonstrated that the C concentration in pollen is similar to that in litter; the concentrations of N, P, S, K, Mg, and Cu are approximately 2- to 12-fold higher in pollen; and concentrations of Ca, Mn, and Fe are approximately 3- to 19-fold higher in litter. As a detritivore food source, pollen stoichiometry is advantageous compared with that of litter, particularly the C:P ratio (12-fold lower in pollen). P-rich pine pollen is deposited in the spring, when young detritivores develop and have the greatest need for P and the highest vulnerability to suboptimal diets during their ontogeny (Bullejos et al., 2014). Thus, feeding on pine pollen may benefit detritivore development. Other elements supplied to a high degree by pine pollen include K and S. Together, these three elements may be limiting for fungi exploiting pine litter (Staaf and Berg, 1982), so pollen might be a particularly important resource for these organisms. Indeed, the utilization of pollen by forest fungi has been reported previously (Stark, 1972; Hutchison and Barron, 1997; Perez-Moreno and Read, 2001).

Nutritionally scarce food is stoichiometrically unbalanced, which is reflected in a stoichiometric mismatch (differences in the concentrations of elements in the food and in the body of the consumer) that limits consumer growth and development (Sterner and Elser, 2002; Denno and Fagan, 2003; Fagan and Denno, 2004; Hessen et al., 2013). To detect potential stoichiometric mismatches and their negative consequences (development limitation), simple comparisons of element ratios are sufficient (Filipiak and Weiner, 2016), so the stoichiometric ratios in food and in the bodies of consumers (C:X_food/C:X_consumer, where C is the carbon concentration and X is the concentration of the other element) were calculated to comprehensively detect and compare the stoichiometric mismatches that reflect the variety of detritivores consuming either litter or pollen. Hereafter, these values will be referred to as trophic stoichiometric ratios (TSRs, cf. Filipiak and Weiner, 2016). To this end, data from the literature (means) and newly collected data on the elemental composition of detritivores inhabiting forest litters and soils (means) were used (details in Supplementary Tables 2–4).

Five elements (P, K, N, S, and Cu; Figure 1) were found to be the most limiting to detritivore development. The TSRs calculated for feeding on pine pollen were approximately 10-fold lower than those for feeding on pine litter for P, K, N, and S and 2-times lower for Cu; the mitigating effect was strongest for P and K. The pollen diet might be particularly advantageous for isopods, millipedes and fungi because these organisms have the greatest need for K, P, and S supplementation (Figure 1). Pine pollen might be an excellent source of nutrients for terrestrial detritivores, mitigating stoichiometric mismatches, thereby it might promote detritivore growth and development.

Pollen has been suggested to be an important source of nutrients for aquatic organisms and a factor that influences nutrient cycling in aquatic ecosystems (Masclaux et al., 2011; Rösel et al., 2012), and the stoichiometric characteristics of pollen are consistent with this suggestion. Pine pollen C:N:P stoichiometry is more similar to that of freshwater than terrestrial autotrophs; i.e., pollen is relatively more stoichiometrically balanced for aquatic consumers compared with other types of plant matter (Elser et al., 2000). Pine pollen has markedly lower C:P and N:P ratios compared with the other matter that flows in large masses from terrestrial to aquatic ecosystems, e.g., foliage (Elser et al., 2000). This may promote the growth of P-limited aquatic detritivores (Sterner and Hessen, 1994; Sterner and Elser, 2002; Sardans et al., 2012). It is possible that pollen does not act as a direct source of nutrients for aquatic herbivores but is first utilized by aquatic fungi, consequently increasing the productivity of the entire system (Masclaux et al., 2011, 2013).

In aquatic food webs, fungal action may be more important to the introduction of pollen-derived nutrients than in terrestrial ecosystems. Although pollen is nutritive and digestible, recent studies have shown that pollen alone is insufficient to promote zooplankton growth (Masclaux et al., 2011, 2013), suggesting that pollen walls are highly resistant to the action of zooplankton and that saprotrophic fungi are responsible for weakening and degrading these walls, thereby introducing pollen-derived nutrients into the food web and boosting ecosystem productivity (Goldstein, 1960; Masclaux et al., 2011, 2013; Rösel et al., 2012; Wurzbacher et al., 2014). Pine pollen is an attractive source of nutrients for fungi due to its C:N:P ratio, but it may also provide other limiting elements, especially K, S, and Cu (Figure 1).

Knowledge of the ecological stoichiometry of aquatic fungi is lacking, and specific predictions or comparisons with food stoichiometry are currently impossible (Danger et al., 2016). The limited data on elemental compositions are either for terrestrial fungi consumed by humans and consider only a finite number of elements, excluding C (Rudawska and Leski, 2005; Dursun et al., 2006), or are related to fungal strains cultured in laboratories on artificial media, which are not relevant to natural situations (Mouginot et al., 2014). A study on the mineral requirements of aquatic fungi, which covered other elements besides C, N, and P (Schoenlein-Crusius et al., 1999), showed that aquatic hyphomycetes may be sensitive to the stoichiometry of nutrients, specifically the contents of Ca, S, K, Mg, Mn, Na, and Zn (positively correlated) and Fe and Al (negatively correlated). Among these elements, S and K are highly concentrated in pine pollen (Table 1), and the positive effect of Zn and the negative effect of Fe on fungi might be associated with competition among these elements for absorption sites, as the excess of one could induce a deficiency of the other. This reported effect occurred with food with a low Zn:Fe atomic ratio (approximately 0.01–0.007); the mean Zn:Fe atomic ratio in pine pollen is 1.2, thus neutral and 24-fold higher than that in pine litter and thus relatively favorable. Pine pollen is also relatively high in Cu, another element that is rich in fungal tissues.

Nutrients that are incorporated into the ecosystem with pollen rain might either be directly utilized by invertebrates, thus mitigating their stoichiometric mismatches (Figure 1), or introduced into the food web via microorganisms (mainly fungi). Pollen on the forest floor decomposes rapidly, losing approximately 20–70% of its initial mass in a month solely as a result of microbial action (Greenfield, 1999; Webster et al., 2008). The majority of the water-extractable macronutrients in pollen (more than 80%) can leach within a few hours in both land (Lee et al., 1996a) and water (Rösel et al., 2012) ecosystems. These nutrients might subsequently be incorporated, redistributed and recycled by microorganisms (Stark, 1972; Hutchison and Barron, 1997; Davidson et al., 1999; Perez-Moreno and Read, 2001; Van Mourik, 2003).

The contribution of pollen to litter decomposition is underrated. It was reported that the N, P, and S supplied by pine pollen to the forest floor enabled fungi to decompose nutritionally scarce pine litter (Stark, 1972; Staaf and Berg, 1982; Hutchison and Barron, 1997), but pine pollen may be even more important, supplying decomposers with sufficient amounts of K and Cu and having a desirable Zn:Fe atomic ratio.

Data on the elemental content of pollen are scarce. Surprisingly, data concerning the concentration of C in litter, pollen and detritivores are also scarce, so studies of particular food webs that utilize a set of elements and allow C:X ratios to be calculated are needed to understand the role of pollen in nutrient cycling. These studies should move beyond traditional C:N:P stoichiometry and incorporate other physiologically important elements, of which K, S, Zn, Fe, and Cu might be the most important.

Even if the annual contribution of pollen-derived nutrients is not significant in terms of the mass of the elements input to the ecosystem, pollen rains may act as a temporal pulse of nutrients. Pollen is cycled through ecosystems during short periods from early spring (primarily from trees) to late summer (primarily from herbs; Lee et al., 1996a; Proctor et al., 1996; Cho et al., 2003). Cole et al. (1990) claimed that summer pollen rains do not contribute to enriching aquatic ecosystems, but data show that spring pollen rains dominated by anemophilous trees relocate considerable amounts of nutrients from terrestrial to aquatic ecosystems (Graham et al., 2006; Rösel et al., 2012). Rösel et al. (2012) suggested that algal blooms may be connected with short-duration but massive pollen deposition from trees. The present study indicates that pine pollen stoichiometry and the ability of pollen rains to rapidly introduce nutrients into food webs makes pollen an ideal agent for triggering algal-blooms.

Future studies could (1) track the pathways of pollen-derived nutrients within and between ecosystems, (2) undertake feeding experiments that supplement the diets of detritivores with various amounts of pollen and evaluate their physiological responses and life history traits (e.g., growth rate, larval development time, size at maturity, assimilation rates of limiting elements, etc.), (3) experimentally manipulate natural food webs by preventing pollen deposition during pollen rain periods and observing the impact on ecosystem productivity. Also needed are studies of the seasonal variations in pollen stoichiometry and deposition that consider various pollen species in different ecosystems. In the case of aquatic ecosystems, the factors affecting the quantity of deposited pollen (e.g., surrounding flora, length of shoreline/lake area ratio, distance from the shore, etc.) should be acknowledged as well as the potential power of a pollen enrichment effect (e.g., trophic state of the pollen-receiving ecosystem). Further suggestions for merging ecological stoichiometry and cross-ecosystem material flows in a spatial context were presented by Sitters et al. (2015).

The author confirms being the sole contributor of this work and approved it for publication.

This study was supported by grants from the Polish Ministry of Science and Higher Education (Grant No. DS/WBiNoZ/INoŚ/DS 756) and the National Science Centre (Grant No. DEC-2013/11/N/NZ8/00929).