Industrial manufactures have produced a range of engineered nanoparticles (ENPs) which have resulted in a higher level of nanosized particles that had always been present in nature. This increase is raising serious concerns over their potential impact on the environment and potential adverse effects on ecosystems, as well as on human health. Dissimilar to bulk material, nanoparticles have individual physical and chemical properties derived from their morphology and composition. Size, shape, purity and catalytic activity of nanoparticles determine their interaction with the environment and living organisms (Darlington et al., 2009). There are very few reports on the effect of nanoceria on terrestrial plants (i.e. Lopez-Moreno et al., 2010a; Lopez-Moreno et al., 2010b). Furthermore, interactions of nanoparticles with other organisms that share similarities with higher plant cells, such as algae, have been poorly studied, remaining unclear the general consequences of nanoparticle exposure for plant cells (Zhang et al., 2012). These studies are mainly based on seed germination tests, seedling growth, nanoparticle uptake and alterations in chlorophyll content or photosynthetic process (Remedios et al., 2012).

Cerium oxide nanoparticles (nanoceria) is a nanosized material which includes particles of 100 nm or less. Nanoceria has a wide range of applications being its main use the formulation of slurries for the chemomechanical planarization of silicon wafers in the production of integrated circuits. Diverse results have been achieved; positive effects on plant growth were reached only at low concentrations (Diatloff et al., 2008) while toxicity was shown at higher concentrations (Lopez-Moreno et al., 2010b). Furthermore, several studies demonstrated that nanoparticles can also have no significant or even positive effects on plants (Remedios et al., 2012). However, all these studies analyze the effects of nanoceria in seed plants and fewer have focused on algae (Rodea-Palomares et al., 2011; Rodea-Palomares et al., 2012).

Ferns, with more than 9,000 species, are the second most successful lineage of vascular plants, after angiosperms. They have a very long evolutionary history related to so important events as the origin of land plants and emergence of the seed. Also, ferns are relevant ecological elements in many ecosystems, especially in the tropics, where they can be dominant (Prada, 2004). Germination is a biological process of capital relevance for seed and spore plants and fungi. It is defined as the set of mechanisms occurring in the dormant germ (seed or spore) that culminates with the growth of the embryo or cell to form a seedling or sporeling able to establish in the substrate (Gabriel y Galán, 2010).

Fern spores are unicellular haploid structures of specific variable size, produced via meiosis, with the faculty to create a gametophyte. Despite of a certain variation in the developmental pattern, the process is as follows: a) the spore germinates, with appearance of the first rhizoidal and prothallial cells; b) a filamentous prothallus is developed, first as uniseriate, then biseriate and planar; c) a meristem is organized, which produces an adult, pre-sexual gametophyte (Gabriel y Galán, 2010). The rhizoid is a single, elongated, nonphotosynthetic cell that is thought to function in anchoring and absorption of nutrients. The protonemal initial eventually gives rise to the photosynthetic prothallus of the fern gametophyte (Banks, 1999). For many ferns, the whole process is very quick, lasting from some days to several weeks. Many physiological and ecological aspects of ferns germination have been studied (Weinberg, 1969; Lloyd, 1970; Raghavan, 1989; Sheffield, 1996; Gabriel y Galán, 2010).

Asplenium adiantum-nigrum is a frequent understory species of Mediterranean evergreen oak forests (Rodà et al. 1999) The haploid phase of the fern Asplenium adiantum-nigrum L. has been previously studied (Prada et al., 1995). Germination and development processes fit the typical leptosporangiate fern model. Ferns, terrestrial and aquatic, have the capacity to take up large amounts of trace elements (Ozaki et al., 2000) and light rare elements, such as cerium (Shan et al., 2003), through their shoots. This ability makes them ideal environmental indicators of contamination, especially those ferns which can tolerate a wide range of environmental extremes (Chang et al., 2009). In this sense, the species Asplenium adiantum-nigrum exhibits resistance to metals and potential to clean up toxic metals growing on mine refuse (Prasad, 2003).

The rationale of this study is based on two facts: first, there is no previous study on the influence of CeO₂ on ferns; second, due to is relative developmental simplicity and speed, fern spore germination and early developmental stages and the tolerance of ferns to some pollutants make ferns an interesting model to study possible toxic effects of bulk-CeO₂ and nanoceria.

Spores of Asplenium adiantum-nigrum examined in the present study began to germinate between seven and nine days after sowing, reaching maximum percentages that ranged between 18 and 100%, depending on the treatment. In the control, seven days after sowing, a 67% spores germinated, and by day nine, germination was complete (100%) (Fig. 1). Seven days after sowing (Fig. 1A), a significant effect of nanoceria and also of most of the bulk-CeO₂ concentrations tested was observed on spore germination as compared to the control. The germination rate of spores treated with nanoceria 100 to 2000 mg L^–1 was significantly higher, showing a stimulating, speeding effect of this product. On the other hand, the germination at 3000 mg L^–1 was significantly lower, with a 14.3% rate of germinated spores. Bulk-CeO₂ also showed a speeding up effect on germination at 500 to 2000 mg L^–1, with the maximum percentage of spore germination (90.7%) observed at 500 mg L^–1 bulk-CeO₂, while no significant effects were observed at other concentrations (Fig. 1A). The lowest-observed-adverse-effect-concentration (LOEC) concentration for spore germination rate (Table 1) was determined as the lowest tested nanoceria treatment (100 mg L^–1), while for bulk-CeO₂ LOEC was 500 mg L^–1 and the no-observed-adverse-effect concentration (NOAEC) was ≤100 mg L^–1.

Fig. 1B shows the results on spore germination eight days after sowing. At this point, only the highest concentration tested (3000 mg L^–1) showed a toxic effect for both nanoceria and bulk materials (Table 1).

The end of the experiment was established nine days after sowing, when spore germination was complete (100%) in the control. In most treatments, germination also reached a 100%. Fig. 1C shows the final results. There were no differences between the unexposed plants and most nanoceria treatments. Only for the 3000 mg L^–1 nanoceria treatment an important toxic effect was observed. Diverse results were obtained for bulk material. A lower number of germinated spores was observed above 1000 mg L^–1 concentration and this effect was significant at higher concentrations (Table 1). Thus, the observed effect on spore germination was dependent on CeO₂ concentration and particle size (nano- or bulk material).

The rhizoid growth was estimated as relative length value in comparison to the control (Fig. 2). Significant decreases in rhizoid growth were observed for 2000 and 3000 mg L^–1 nano-CeO₂ and for 3000 mg L^–1 bulk-CeO₂ while lower concentrations had no significant effect.

In order to further explore the cellular mechanisms of the observed toxicity, we took bright-field micrographs of Asplenium adiantum-nigrum germinated spores exposed to the treatments. In the control (Fig. 3A), as well as in 100 (Fig. 3B) and 500 mg L^–1 nanoceria (Fig. 3C), well formed protonema cells with bright green chloroplasts were clearly visible. In contrast, nanoceria at concentration above 1000 mg L^–1 and bulk material above100 mg L^–1 generally resulted in germinated spores, many of which manifested abnormalities, such as damages to their internal structure, collapsed 2-3 celled filaments (Fig. 3E); filaments of 2-3 cells aborted and collapsed (Fig. 3I), chloroplast damaged cells (Fig. 3D-K) and multiple rhizoid formation or aborted protonema (Fig. 4), that sharply contrast with the healthy controls.

The germination of the spores of Asplenium adiantum-nigrum is affected by the addition of CeO₂ to the culture medium. Materials surface chemistry is vital in biological interaction (Karakoti et al., 2006). In general, for a fixed mass of particles, surface area increases as particle size becomes smaller. Thus, a dose-dependence on particle surface area may explain the greater toxicity of nanoparticles compared to an equal mass of fine particles of the same material (Donaldson et al., 2002; Monteiller et al., 2007; Sager, 2009). However, other reports (Warheit et al., 2006; Warheit et al., 2007) have questioned this hypothesis and, in agreement with other results (Rodea-Palomares et al., 2011), our data do not support the view that surface area might work better than mass concentration as a dose variable when photosynthetic organisms are tested.

Differences in the germination rate have been observed between spores subjected either to nanoceria or bulk-CeO₂ even for the lowest concentration tested (100 mg L^–1, Fig. 1 A and B). At this concentration, germination was speeded up by nanoceria, while spore germination in bulk-CeO₂ was retarded (Fig. 1 B). On the other hand, only 500 to 2000 mg L^–1 concentrations of bulk-CeO₂ showed a similar pattern. In fact, bulk material responds to a dose/response curve (Fig. 1 B). Differences between the control and the treatments decreased for the eighth and ninth day after sowing, when only the highest concentration (3000 mg L^–1) treatment showed toxicity for both chemicals, although it was stronger for nanoceria (Fig. 1 B and C). The parameters in the spore germination test, including NOAEC and LOEC, were lower for bulk-CeO₂ endpoint (nine days) than for nanoceria (Table 1). This supports the requirement of the use of both, nano and bulk material for toxicity tests.

Taking into account the differences among angiosperms and ferns, some authors (Lopez-Moreno et al., 2010b) showed, at concentrations of nanoceria from 500 to 2000 mg L^–1, relatively low to moderate toxicity on seed germination of alfalfa (Medicago sativa), cucumber (Cucumis sativa), tomato (Solanum lycopersicum) and corn (Zea mays). Nanoceria did not affect germination in lettuce (Lactuca sativa), tomato (Solanum lycopersicum), cabbage (Brassica oleracea), soybean (Glycine max), carrot (Daucus carota), perennial ryegrass (Lolium perenne), corn (Zea mays), cucumber (Cucumis sativus), oat (Avena sativa), and onion (Allium cepa) at concentrations between 250 and 1000 mg L^–1 (Andersen et al., 2016). The toxic effect at the highest concentration tested (3000 mg L^–1) was more marked for nanoceria than for bulk-CeO₂. At this highest nanoparticle concentration, cell toxicity could be related to concentration and to the presence of nanoparticle aggregates (Rodea-Palomares et al., 2011).

No significant differences were observed for rhizoid growth among the control treatments (Fig. 2). The rhizoid is the cell responsible for the absorption of nutrients and, keeping again the differences, acts as the root. It has been reported previously (Lopez-Moreno et al., 2010b) that root growth was reduced in alfalfa and tomato but was significantly promoted in cucumber and corn. Andersen et al. (2016) found that nanoceria alter average root length, and hence root growth was decreased in cabbage and corn, but was promoted in cucumber and onion. Ma et al. (2010) only detected a reduction on the root elongation in lettuce but no effect was detected for a suspension of 2000 mg L^–1 nanoceria for rape, radish, wheat, cabbage, tomato, and cucumber. All these results suggest that the effects produced on early plant growth of nanoceria is species dependent. Nevertheless, rhizoid is a unicellular structure and, in this sense, differences with an organ (root) are evident. Furthermore, alterations in normal development of gametophyte were observed. For example, as shown in Fig. 4, more than one rhizoid was emitted by single spores. However, we did not observe cell damage in rhizoid structures.

Two explanations have been proposed for CeO₂ toxicity effects: mechanical damage to the cell membranes due to the numerous edges, corners, and reactive sites present in the crystal structure of the nanoparticles (Rogers et al., 2010); or the generation of ROS (reactive oxygen species), thus inducing oxidative stress and cell toxicity leading to lipid/protein oxidation (Thill et al., 2006; Park et al., 2008; Zeyons et al., 2009). Although we did not examine the presence of particles anchored to the cell surface, we did observe membrane rupture, cytoplasm leakage, and intracellular damage including chloroplast lysis for most treatments (Fig. 3 D-K). Similar results were obtained in the green alga Pseudokirchneriella subcapitata (Rodea-Palomares et al., 2011). These authors indicate that nanoceria cytotoxicity could be mediated by the effect of nanoceria on photosynthesis, which could be involved in the generation of ROS when photosynthetic reactions are not well balanced (Rico et al., 2015). Nanoceria can increase the production of hydrogen peroxide through an oxidative reaction (Rodea-Palomares et al., 2012), which can cause lipid peroxidation compromising membrane integrity. We have observed critical chloroplast damage, which consequently resulted in cell and gametophyte death for all the bulk-CeO₂ treatments (Fig. 3G, H, I, J, and K), and for the high concentrations of nanoceria (1000, 2000 and 3000 mg L^–1, Fig. 3D, E, and F), while in the lower concentrations of nanoceria (100 and 500 mg L^–1, Fig. 3B and C) well defined chloroplasts were observed.

Present results showed cellular damage in the protonema, while rhizoid cells seemed not to be affected (growth and membrane integrity remained) either by bulk-CeO₂ or nanoceria treatments. As chloroplasts are present in the protonema and absent in the rhizoid, the question remains open to a possible role of CeO₂ toxicity through oxidative stress. In such hypothesis, photosynthetic cells would be more affected by the reduction of cerium (IV) to cerium (III) (Rodea-Palomares et al., 2012). In agreement with this, the main driver of toxicity was found to be the percentage of surface content of Ce³⁺ sites (Pulido-Reyes et al., 2015).

Cerium oxide is toxic for Asplenium adiantum-nigrum. The sensibility of this species to CeO₂ makes it especially interesting as indicator of its presence. Diverse toxic effects have been observed: i) chloroplast membrane damage and lysis, ii) cell wall and membrane disruption which leads to cell lysis; and iii) alterations in morphology and development. Nevertheless, this toxicity is not apparent when low/moderate concentrations of nanoceria are added to the culture medium.