Phytolith Assemblages in Grasses Native to Central Argentina

Abstract

• Background and Aims Phytolith reference collections are a prerequisite for accurate interpretation of soil phytolith assemblages aimed at reconstructing past vegetation. In this study a phytolith reference collection has been developed for several grasses native to central Argentina: Poa ligularis, Piptochaetium napostaense, Stipa clarazii, Stipa tenuis, Stipa tenuissima, Stipa eriostachya, Stipa ambigua, Stipa brachychaeta, Pappophorum subbulbosum, Digitaria californica, Bothriochloa edwardsiana and Aristida subulata.

• Methods For each species, phytoliths present in the leaf blades were classified into 47 morphotypes, and their relative frequency determined by observing 300–400 phytoliths per sample (n = 5). Data were analyzed by complete linkage cluster analysis, using the Morisita Index as measure of association.

• Key Results The results showed differentiation among phytolith assemblages at species level or at plant functional type level. Cluster analysis separated C₃ from C₄ species and palatable from non-palatable species.

• Conclusions This study highlights the possibility of reconstructing past vegetation in central Argentina grasslands through the analysis of soil phytolith assemblages.

Poaceae, native grasses, Argentina, phytolith assemblages, grass phytoliths

INTRODUCTION

Phytoliths are microscopic particles of hydrated silica deposited in intracellular and/or intercellular spaces of plant tissues, which can take on a considerable variety of forms (Rovner, 1983; Piperno, 1988). Because of their consistent shape within species, phytoliths provide significant taxonomic information (Jones and Handreck, 1965; Blackman, 1969; Raven, 1983; Piperno, 1988). Moreover, since these resistant siliceous structures can remain stable in the soil for hundreds and even millions of years after plant tissues decay (Rovner, 1983; Twiss, 1987; Piperno, 1988; Carter, 1999), they are important in palaeoecological studies.

Although the accumulation of silica occurs in various taxa of the plant kingdom, phytoliths are particularly abundant in the family Poaceae (Gramineae) (Rovner, 1971; Piperno, 1988; Mulholland, 1989; Twiss, 1992). In this family, individuals produce many types of phytolith forms (multiplicity), and a particular form may be produced by a number of different species (redundancy) (Rovner, 1983). For that reason, phytolith descriptions at species level include all morphotypes and their relative frequency, which together characterize specific phytolith assemblages. The diagnostic potential of grass phytoliths and their durability in the soil makes phytolith analysis an important microfossil technique for the reconstruction of past vegetation in grasslands (Sendulsky and Labouriau, 1966; Blackman, 1971; Bartolome et al., 1986; Twiss, 1987; Fisher et al., 1995; Alexandre et al., 1997; Barboni et al., 1999). To that end, phytolith reference collections are a prerequisite for accurate interpretation of soil phytolith assemblages.

In the grasslands of central Argentina a process of species replacement appears to have taken place since the introduction of domestic livestock at the beginning of the last century (Llorens, 1995; Distel and Bóo, 1996). However, this assumption is based on circumstantial evidence because of the lack of relict areas and historical data. The objective of this study was to examine phytolith assemblages in grasses native to central Argentina, in order to assist in the analysis of soil phytolith assemblages towards the reconstruction of past vegetation.

MATERIALS AND METHODS

Species were collected from a site located in the south-east of La Pampa Province, Argentina (38°45′S, 63°45′W, and 80 m a.s.l.). The climate of this region is temperate and semi-arid. Mean annual temperature is 15 °C, mean annual precipitation is 400 mm, and mean annual potential evapotranspiration is 800 mm (Instituto Nacional de Tecnología Agropecuaria, 1980). The dominant soils are coarse-textured Petrocalcic Calciustolls (Castelli et al., 1995). The site is located in the southern part of the Caldén Phytogeographical District (Cabrera, 1976). The physiognomy of the vegetation is grassland with isolated shrubs and trees. The more abundant herbaceous species are C₃, perennial, cool-season grasses; although some C₄, perennial, warm-season grasses are also present. Table 1 contains a list and description of the grass species examined in this study.

Table 1.

Grass species examined for phytoliths. Species descriptions conform to Cabrera (1970) and Lamberto et al. (1997)

Species	Tribe	Photosynthetic pathway	Animal grazing preference
Piptochaetium napostaense (Speg.) Hackel	Stipeae	C3	High
Stipa clarazii Ball.	Stipeae	C3	High
Stipa tenuis Phil.	Stipeae	C3	High
Poa ligularis Nees	Festuceae	C3	High
Stipa tenuissima Trin.	Stipeae	C3	Low
Stipa eriostachya H.B.K.	Stipeae	C3	Low
Stipa ambigua Speg.	Stipeae	C3	Low
Stipa brachychaeta Godr.	Stipeae	C3	Low
Aristida subulata Henrard	Aristideae	C4	Low
Pappophorum subbulbosum Arech.	Pappophoreae	C4	High
Digitaria californica (Benth.) Henr.	Paniceae	C4	High
Bothriochloa edwardsiana (Gould) Parodi	Andropogoneae	C4	High

Leaf blade samples (n = 5 individuals) from each of the examined species were collected at the end of their annual growing cycle, to better represent phytolith composition at the moment of incorporation into the soil (Parry and Smithson, 1964; Geis, 1978; Rovner, 1983). Cool and warm-season grasses were collected in December of 1998 and April of 1999, respectively.

Phytolith extraction from the leaf blades was accomplished through the calcination technique (Labouriau, 1983). The material was first washed with distilled water and dried at 100 °C. Immediately after this it was charred at 200 °C for 2 h, boiled in 5 N HCl for 30 min, filtered through ashless filter paper and washed with distilled water until no more chloride ions were detected. Finally, the material was dried and charred at 200 °C for 20 min and subsequently ignited at 800 °C for 2 h. The ashes were mounted on microscope slides in Canada balsam and observed in a Wild M11 microscope.

We observed 300–400 phytoliths per sample and calculated the percentage of each type present, following the morphological classification proposed by Twiss (1992) and Zucol (2001) (Table 2). Phytoliths were grouped into 47 morphotypes. Ten were present in associated forms (phytolith skeletons formed of more than one anatomic element), and 37 in isolated forms (phytolith formed of only one anatomic element) (Table 2 and Figs 1–4). Phytolith assemblages were subjected to complete linkage cluster analysis. The Morisita Index was used to measure the strength of the association among specific phytolith assemblages (Horn, 1966).

Fig. 1.

Light micrographs of isolated-form phytoliths. (A) Elliptical. (B) Hat. (C) Biconvex. (D) Crescent moon. (E) Oblong. (F) Round. (G) Rectangular. (H) Normal saddle. (I) Thin saddle. (J) Cross, thick shank. (K) Cross, thin shank. (L) Dumb-bell, short shank, convex end. (M) Dumb-bell, short shank, straight end. (N) Dumb-bell, long shank, convex end. (O) Dumb-bell, long shank, straight end. a, Poa ligularis; b, Stipa clarazii; c, Piptochaetium napostaense; d, Stipa tenuis; e, Stipa gynerioides; f, Stipa tenuissima; g, Stipa ambigua; h, Stipa brachychaeta; i, Aristida subulata; j, Bothriochloa edwardsiana; k, Digitaria californica; l, Pappophorum subbulbosum. Scale bar = 10 μm.

Fig. 2.

Light micrographs of isolated-form phytoliths. (A) Dumb-bell, nodular shank, convex end. (B) Dumb-bell, nodular shank, straight end. (C) Dumb-bell, spiny central portion, convex end. (D) Dumb-bell, spiny shank, straight end. (E) Regular, complex dumb-bell, convex end. (F) Regular, complex dumb-bell, straight end. (G) Irregular, complex dumb-bell, convex end. (H) Irregular, complex dumb-bell, straight end. (I) Crenate dumb-bell, convex end. (J) Crenate dumb-bell, straight end. b, Stipa clarazii; c, Piptochaetium napostaense; d, Stipa tenuis; e, Stipa gynerioides; f, Stipa tenuissima; g, Stipa ambigua; h, Stipa brachychaeta; i, Aristida subulata; j, Bothriochloa edwardsiana. Scale bar = 10 μm.

Fig. 3.

Light micrographs of isolated-form phytoliths. (A) Elongate, smooth. (B) Elongate, sinuous. (C) Elongate, spiny. (D) Elongate, concave ends. (E) Fan. (F) Polyhedrical. (G) Apex of sharp-pointed shapes (hairs, prickles or hooks). (H) Macro-hairs. (I) Micro-hairs. (J) Prickles hairs. (K) Hooks hairs. (L) Triangular shapes. a, Poa ligularis; b, Stipa clarazii; c, Piptochaetium napostaense; d, Stipa tenuis; e, Stipa gynerioides; f, Stipa tenuissima; g, Stipa ambigua; h, Stipa brachychaeta; i, Aristida subulata; j, Bothriochloa edwardsiana; k, Digitaria californica; l, Pappophorum subbulbosum. Scale bar = 20 μm.

Fig. 4.

Light micrographs of associated-form phytoliths. (A) Subepidermal elements. (B) Bulliform elements. (C) Stomatal complex. (D) Large cells. (E) Short cells. (F) Papillae. (G) Macro-hairs. (H) Micro-hairs. (I) Prickles hairs. (J) Hooks hairs. a, Poa ligularis; b, Stipa clarazii; c, Piptochaetium napostaense; d, Stipa tenuis; e, Stipa gynerioides; f, Stipa tenuissima; g, Stipa ambigua; h, Stipa brachychaeta; i, Aristida subulata; j, Bothriochloa edwardsiana; k, Digitaria californica; l, Pappophorum subbulbosum. Scale bar = 20 μm.

Table 2.

Morphological classification of phytoliths (according to Twiss, 1992, and as modified by Zucol, 2001)

Isolated forms
Pooide Class
1. Elliptical
2. Hat
3. Biconvex
4. Crescent moon
5. Oblong
6. Round
7. Rectangular
Chloridoide Class
8. Normal saddle
9. Thin saddle
Panicoide Class
10. Cross, thick shank
11. Cross, thin shank
12. Dumb-bell with short central portion and convex end
13. Dumb-bell with short central portion and straight end
14. Dumb-bell with long central portion and convex end
15. Dumb-bell with long central portion and straight end
16. Dumb-bell with nodular central portion and convex end
17. Dumb-bell with nodular central portion and straight end
18. Dumb-bell, spiny central portion, convex end
19. Dumb-bell, spiny central portion, straight end
20. Regular, complex dumb-bell, convex end
21. Regular, complex dumb-bell, straight end
22. Irregular, complex dumb-bell, convex end
23. Irregular, complex dumb-bell, straight end
24. Crenate dumb-bell, convex end
25. Crenate dumb-bell, straight end
Elongate Class
26. Elongate, smooth
27. Elongate, sinuous
28. Elongate, spiny
29. Elongate, concave ends
Fan-shaped and Polyhedrical Class
30. Fan
31. Polyhedrical
Point-shaped Class
32. Apex of sharp-pointed shapes (hairs, prickles or hooks)
33. Macro-hairs
34. Micro-hairs
35. Prickles hairs
36. Hooks hairs
37. Triangular shapes
Associated forms
38. Subepidermal elements
39. Bulliform elements
40. Stomatal complex
41. Large cells
42. Short cells
43. Papillae
44. Macro-hairs
45. Micro-hairs
46. Prickles hairs
47. Hooks hairs

RESULTS

Table 3 contains the phytolith assemblages of the examined species. The dendrogram from the cluster analysis showed distinctive phytolith assemblages for Poa ligularis, Aristida subulata and Pappophorum subbulbosum (Fig. 5). Phytolith assemblages in these species were characterized by a high frequency of round, elliptical and oblong shapes in P. ligularis, dumb-bell with a short central portion and straight end in A. subulata, and saddle and cross in P. subbulbosum. The rest of the species were arranged in the following subgroups:

• Stipa tenuis and Piptochaetium napostaense

• Digitaria californica and Bothriochloa springfieldii

• Stipa eriostachya and Stipa tenuissima associate with Stipa clarazii, Stipa ambigua and Stipa brachychaeta.

Fig. 5.

Dendrogram showing the relationships among phytolith assemblages of the examined species. The Morisita Index was used to measure the strength of the association among specific phytolith assemblages (Horn, 1966).

Table 3.

Relative frequency of phytolith morphotypes. Morphotypes are as in Table 2

Morphotypes	Poa ligularisa	Stipa claraziia	Piptochaetium napostaensea	Stipa tenuisa	Stipa eriostachyaa	Stipa tenuissimaa	Stipa ambigua	Stipa brachychaeta	Bothriochloa edwardsiana	Digitaria californica	Pappophorum subbulbosum	Aristida subulata
1	18·55 ± 3·03	4·84 ± 0·67	0·87 ± 0·21	12·27 ± 1·00	7·19 ± 2·16	3·65 ± 0·80	0·99 ± 0·09	2·30 ± 0·54	–	0·23 ± 0·10	0·05 ± 0·05	–
2	–	1·19 ± 0·28	–	0·66 ± 0·19	5·82 ± 1·98	1·70 ± 0·16	0·42 ± 0·15	0·39 ± 0·25	–	–	–	–
3	–	0·17 ± 0·06	–	0·04 ± 0·04	1·24 ± 0·37	–	0·44 ± 0·17	0·10 ± 0·10	–	–	–	–
4	1·70 ± 0·39	2·08 ± 0·35	0·43 ± 0·17	4·16 ± 0·53	2·43 ± 0·64	1·23 ± 0·29	0·29 ± 0·19	0·05 ± 0·05	–	–	0·10 ± 0·10	0·05 ± 0·05
5	6·32 ± 2·09	4·55 ± 0·37	–	2·67 ± 0·56	0·05 ± 0·05	1·02 ± 0·46	–	–	–	–	0·34 ± 0·21	0·05 ± 0·05
6	34·40 ± 2·93	3·27 ± 1·34	0·34 ± 0·17	4·63 ± 0·96	0·96 ± 0·30	0·27 ± 0·12	–	0·20 ± 0·10	–	0·30 ± 0·18	0·14 ± 0·10	0·10 ± 0·06
7	4·09 ± 0·63	8·19 ± 1·07	0·82 ± 0·31	1·41 ± 0·25	1·31 ± 0·29	1·58 ± 0·22	5·64 ± 0·97	6·79 ± 1·66	0·29 ± 0·20	3·08 ± 0·69	1·46 ± 0·47	0·31 ± 0·15
8							–	–	–	–	30·59 ± 1·79	–
9							–	–	–	–	0·88 ± 0·59	–
10							–	–	1·37 ± 0·45	–	11·92 ± 1·77	–
11							–	–	0·03 ± 0·03	–	0·04 ± 0·04	–
12	1·14 ± 0·92	6·03 ± 1·14	15·13 ± 1·46	15·68 ± 2·01	1·76 ± 0·40	4·97 ± 0·84	3·87 ± 0·80	5·20 ± 1·33	0·17 ± 0·08	6·25 ± 0·50	9·54 ± 2·34	1·54 ± 0·48
13	1·10 ± 0·76	24·08 ± 1·48	11·68 ± 1·31	9·91 ± 1·32	35·59 ± 3·45	32·12 ± 1·98	29·77 ± 1·75	33·51 ± 3·42	32·19 ± 2·22	43·43 ± 1·28	10·90 ± 3·31	5·49 ± 1·72
14	–	0·06 ± 0·06	6·16 ± 1·10	0·22 ± 0·16	0·56 ± 0·26	0·21 ± 0·07	1·28 ± 0·61	–	0·10 ± 0·10	3·00 ± 0·72	1·48 ± 0·48	5·75 ± 1·67
15	–	0·50 ± 0·19	1·30 ± 0·24	–	0·96 ± 0·16	0·40 ± 0·17	0·30 ± 0·09	3·64 ± 0·93	10·22 ± 0·43	12·04 ± 1·26	–	37·96 ± 3·22
16	0·05 ± 0·05	0·09 ± 0·06	8·13 ± 0·91	2·51 ± 0·55	0·24 ± 0·07	0·13 ± 0·09	1·22 ± 0·33	0·45 ± 0·19	0·04 ± 0·04	0·32 ± 0·19	–	0·36 ± 0·10
17	0·14 ± 0·14	1·96 ± 0·41	1·15 ± 0·13	1·02 ± 0·17	1·09 ± 0·26	1·87 ± 0·49	4·48 ± 0·91	3·29 ± 0·93	3·89 ± 0·84	0·11 ± 0·07	0·20 ± 0·12	0·91 ± 0·37
18	–	–	1·07 ± 0·33	0·13 ± 0·05	0·09 ± 0·06	0·25 ± 0·15	0·07 ± 0·07	0·10 ± 0·06	–	–		–
19	–	0·92 ± 0·30	0·26 ± 0·14	0·27 ± 0·09	1·08 ± 0·53	2·95 ± 1·32	0·93 ± 0·24	0·98 ± 0·17	0·20 ± 0·20	0·06 ± 0·06	–	–
20	–	0·03 ± 0·03	4·84 ± 0·40	0·21 ± 0·11	0·10 ± 0·06	0·17 ± 0·08	0·25 ± 0·11	0·10 ± 0·10	–	–	0·05 ± 0·05	0·20 ± 0·09
21	–	0·37 ± 0·10	0·91 ± 0·33	0·04 ± 0·04	0·09 ± 0·06	0·63 ± 0·32	0·98 ± 0·52	0·83 ± 0·41	3·07 ± 0·70	–	–	0·20 ± 0·14
22	–	–	1·97 ± 0·21	0·09 ± 0·05	–	0·13 ± 0·09	0·07 ± 0·07	0·15 ± 0·10	–	–	–	0·05 ± 0·05
23	–	0·23 ± 0·12	0·10 ± 0·10	0·08 ± 0·08	0·28 ± 0·09	0·26 ± 0·08	0·63 ± 0·15	0·79 ± 0·23	0·27 ± 0·13	–	–	0·10 ± 0·06
24	–	–	2·11 ± 0·47	–	–	–	2·51 ± 0·88	0·64 ± 0·10	0·15 ± 0·07	0·11 ± 0·11	–	0·16 ± 0·10
25	–	–	–	–	0·05 ± 0·05	–	–	–	0·16 ± 0·12	–	–	–
26	5·71 ± 1·88	5·68 ± 1·35	4·25 ± 1·78	16·51 ± 2·57	0·64 ± 0·28	3·65 ± 1·04	10·48 ± 0·65	14·06 ± 2·90	1·76 ± 0·48	1·94 ± 0·25	0·59 ± 0·23	1·70 ± 0·12
27	0·55 ± 0·15	0·46 ± 0·13	1·43 ± 0·52	1·34 ± 0·33	0·33 ± 0·17	0·74 ± 0·16	1·88 ± 0·58	0·73 ± 0·22	0·46 ± 0·18	0·75 ± 0·15	0·69 ± 0·14	1·39 ± 0·12
28	1·49 ± 0·58	0·61 ± 0·22	3·36 ± 0·72	1·54 ± 0·31	0·77 ± 0·33	0·37 ± 0·12	2·29 ± 0·45	1·72 ± 0·63	0·16 ± 0·07	1·10 ± 0·18	2·12 ± 0·32	3·87 ± 0·72
29	0·09 ± 0·05	0·03 ± 0·03	0·05 ± 0·05	0·19 ± 0·09	0·09 ± 0·06	0·28 ± 0·14	0·69 ± 0·21	0·15 ± 0·10	0·26 ± 0·13	0·17 ± 0·11	0·29 ± 0·06	0·05 ± 0·05
30	0·04 ± 0·04	–	–	–	–	–	–	–	0·48 ± 0·19	–	0·91 ± 0·57	–
31	0·18 ± 0·13	0·06 ± 0·06	0·05 ± 0·05	0·09 ± 0·09	–	–	0·12 ± 0·07	0·30 ± 0·18	1·03 ± 0·31	0·67 ± 0·22	1·01 ± 0·21	0·31 ± 0·15
32	6·74 ± 1·17	2·82 ± 0·66	1·75 ± 0·53	0·94 ± 0·30	9·41 ± 1·26	9·47 ± 2·21	1·73 ± 0·54	4·85 ± 0·98	0·85 ± 0·33	3·35 ± 0·26	6·18 ± 0·40	6·35 ± 1·00
33	3·66 ± 0·67	0·37 ± 0·12	2·25 ± 0·26	0·44 ± 0·20	2·23 ± 0·45	16·13 ± 1·60	1·52 ± 0·55	–	–	0·05 ± 0·05	0·17 ± 0·11	–
34							–	–	0·69 ± 0·20	–	2·20 ± 1·47	–
35	1·50 ± 0·31	1·99 ± 0·07	4·87 ± 0·75	1·81 ± 0·18	1·99 ± 0·39	1·05 ± 0·07	8·41 ± 1·17	7·09 ± 0·81	0·66 ± 0·27	5·67 ± 1·06	1·12 ± 0·40	3·93 ± 0·80
36	–	2·88 ± 0·78	10·45 ± 1·32	7·93 ± 1·37	–	–	6·37 ± 1·98	3·70 ± 1·12	0·85 ± 0·21	4·71 ± 0·42	2·05 ± 0·34	4·49 ± 0·81
37	–	0·24 ± 0·14	0·10 ± 0·10	–	–	0·76 ± 0·28	0·07 ± 0·07	–	0·08 ± 0·05	–	0·04 ± 0·04	–
38	0·13 ± 0·08	0·65 ± 0·15	0·22 ± 0·10	0·37 ± 0·12	0·57 ± 0·12	0·10 ± 0·10	0·92 ± 0·08	0·10 ± 0·10	0·85 ± 0·30	0·06 ± 0·06	0·73 ± 0·11	2·03 ± 0·69
39							1·11 ± 0·25	1·18 ± 0·36	0·65 ± 0·14	0·20 ± 0·09	1·11 ± 0·12	0·10 ± 0·06
40	4·22 ± 0·94	5·89 ± 0·65	0·82 ± 0·13	1·02 ± 0·19	2·13 ± 0·51	0·66 ± 0·26	3·14 ± 0·46	2·16 ± 0·75	0·42 ± 0·14	0·32 ± 0·10	3·16 ± 0·41	5·81 ± 0·73
41	3·27 ± 0·32	4·61 ± 0·48	3·22 ± 0·16	4·87 ± 0·41	3·05 ± 0·30	2·97 ± 0·77	3·07 ± 0·65	1·98 ± 0·60	4·52 ± 0·35	5·85 ± 0·60	2·12 ± 0·43	7·39 ± 1·23
42	2·46 ± 0·62	7·49 ± 0·73	2·85 ± 0·61	1·96 ± 0·46	6·17 ± 0·56	4·93 ± 1·27	1·28 ± 0·33	0·74 ± 0·30	10·36 ± 0·99	3·47 ± 0·71	4·53 ± 0·40	4·66 ± 0·99
43							–	–	19·00 ± 1·58	–	1·51 ± 0·35	–
44	1·11 ± 0·27	0·48 ± 0·26	0·05 ± 0·05	0·22 ± 0·07	4·00 ± 1·08	2·92 ± 0·54	–	–	–	–	0·05 ± 0·05	–
45							–	–	0·86 ± 0·37	0·05 ± 0·05	1·07 ± 0·31	0·81 ± 0·28
46	1·37 ± 0·44	2·78 ± 0·29	3·63 ± 0·52	1·74 ± 0·25	7·74 ± 0·79	2·45 ± 0·62	1·67 ± 0·65	1·18 ± 0·35	2·85 ± 0·64	1·95 ± 0·19	0·46 ± 0·09	2·14 ± 0·41
47	–	4·40 ± 1·08	3·42 ± 0·70	3·03 ± 0·41	–	–	1·11 ± 0·25	0·54 ± 0·30	1·06 ± 0·31	0·73 ± 0·21	0·21 ± 0·13	1·71 ± 0·46

Values are mean ± s.e. (n = 5).

a

Morphotypes 8–11, 34, 39, 43 and 45 were absent for these species.

Phytolith assemblages in these subgroups were characterized by a high frequency of dumb-bells with a short central portion and convex ends, and isolated hooks in Subgroup I, dumb-bells with large central portion and straight end in Subgroup II, and dumb-bells with a short central portion and straight ends, rectangular, smooth elongate, hooks and apex of sharp-pointed shapes in Subgroup III.

A detailed comparison of morphotype frequency and presence showed species differences within subgroups. In Subgroup I, S. tenuis was characterized by a high frequency of smooth elongate and elliptical, round and crescent moon morphotypes, whereas P. napostaense was characterized by a high frequency of dumb-bells with a large and nodular central portion (both with convex ends), regular-complex dumb-bells with a convex end, and prickles. In Subgroup II, D. californica was distinguished by a high frequency of dumb-bells with a short central portion, isolated hooks and isolated prickles, whereas a high frequency of short cells and the presence of regular-complex dumb-bells with straight end, papillae and cross shapes characterized B. edwarsiana. In Subgroup III, a high frequency ofsharp-pointed apex shapes characterized S. eriostachya and S. tenuissima, whereas a high frequency of rectangular, smooth elongate, and the presence of hooks characterized S. clarazii, S. ambigua and S. brachychaeta. Furthermore, a high frequency of elliptical, hat and associated prickles distinguished S. eriostachya, whereas a high frequency of isolated macro-hairs distinguished S. tenuissima. On the other hand, a high frequency of elliptical, short cells, stomata complex and associated hooks, and the presence of round and oblong phytoliths characterized S. clarazii, whereas the presence of nodular dumb-bells with a straight end, smooth elongate, isolated hooks, and isolated prickles characterized S. ambigua and S. brachychaeta. The latter phytolith assemblages differed from each other by a high frequency of dumb-bells with a large central portion and straight end, and apex sharp-pointed shapes in S. brachychaeta, whereas isolated hooks and large cells characterized S. ambigua.

DISCUSSION

The phytolith morphotypes present in the species studied are characteristic of the Poaceae, as defined by Twiss (1992). Although phytolith multiplicity and redundancy can limit the differentiation between species (Rovner, 1971), our cluster analysis results show the possibility of distinguishing at the species level or between plant functional types (i.e. C₃ from C₄ grasses, preferred from avoided grasses) in the native grasslands of central Argentina.

The dominant phytolith forms described for P. ligularis, P. subbulbosum and A. subulata were consistent with previous descriptions for the same species obtained through microhistological analysis (Lindström et al., 1998). The phytolith assemblages of these species corresponded to the subfamilies Pooideae, Chloridoideae and Arundinoideae, respectively, as defined by Twiss (1992). The subfamily Pooideae is characterized by the presence of round, rectangular, oblong, elliptical and crescent-moon shapes, and by the absence of saddle, dumb-bells and cross shapes. The subfamily Chloridoideae is characterized by the presence of cross, dumb-bell and, particularly, saddle shapes. The subfamily Arundinoideae represents a highly heterogeneous group, in which the dominant forms are rectangular, oblong, crenate and dumb-bell.

In A. subulata the dominant form was the dumb-bell type, with a large and narrow central portion, rounded end and lateral plane of symmetry well developed (Fig. 6A). This form corresponds to the simple bilobate type described by Mulholland (1989), Fredlund and Tieszen (1994) and Kerns (2001). These authors distinguished this type from other types of dumb-bell, denominating them Stipa and Panicoid types. The Stipa-type dumb-bell (observed in Subgroups I and III) has two opposite broad faces of different size (top or planar view showing the outline, with two lobes connected by a shaft), and a trapezoidal side view (lateral faces connecting the two opposite broad faces) (Fig. 6C). The Panicoid-type dumb-bell (observed in P. subbulbosum) has indented ends, a double outline in planar view, and a tabular side view (Fig. 6B).

Fig. 6.

Light micrographs of phytoliths forms in three-dimensional view. (A) Simple-type dumb-bell, with large and narrow central portion, rounded end in planar view (pv), and lateral plane of symmetry well developed (sv). (B) Panicoid type dumb-bell, with indented end, a double outline in planar view (pv), and a tabular side view (sv). (C) Stipa type dumb-bell, with two opposite broad faces of different size (ad: adaxial face, ab: abaxial face) in planar view (pv), and a trapezoidal side view (sv). c, Piptochaetium napostaense; d, Stipa tenuis; f, Stipa tenuissima; g, Stipa ambigua; h, Stipa brachychaeta; i, Aristida subulata; j, Bothriochloa edwardsiana; k, Digitaria californica; l, Pappophorum subbulbosum. Scale bar = 10 μm.

The phytolith characterization of D. californica and B. edwardsiana corresponded to the subfamily Panicoideae, as defined by Twiss (1992). This subfamily is characterized by the presence of smooth and sinuous elongate, dumb-bell (Panicoid-type) and cross shapes. Forms representative of the subfamily Arundinoideae, with a high frequency of Stipa-type dumb-bells forms (Fig. 6C), characterized the rest of the species. Zucol (1996) reported similar results for species of the genus Stipa inhabiting the Entre Ríos province, Argentina.

Our examination of phytolith assemblages in several grasses native to the central Argentina also highlights the possibility to differentiate between C₃ and C₄ species (Table 1 and Fig. 5). The C₄ grasses were dominated by the dumb-bell (simple bilobate and Panicoid-types), cross and saddle forms, whereas the C₃ grasses were dominated by the Pooide Class forms (mainly in P. ligularis) and Stipa-type dumb-bells. The distinction between C₃ / C₄ grasses based upon phytolith forms was first suggested by Twiss et al. (1969). Later studies have improved the differentiation between C₃ / C₄ grasses by adding a three-dimensional examination of phytoliths, to minimize problems associated with multiplicity and redundancy of phytolith forms (Brown, 1984; Fredlund and Tiezen, 1994). Our results also show that differences in phytolith assemblages allow differentiation between preferred (palatable, high forage value) and avoided (non-palatable, low forage value) grasses within photosynthetic pathway groups (Table 1 and Fig. 5). However, the distinction between preferred and avoided grasses based on phytolith analysis need to be examined further to know its generality. In summary, we conclude that the present phytolith reference collection can assist in the interpretation of soil phytolith assemblages aimed at reconstructing past vegetation in the central Argentina grasslands.

This research was supported by Universidad Nacional del Sur, Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) de la República Argentina, and Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT) de la República Argentina. A fellowship from CONICET to L. Gallego is acknowledged.