Seed Bank and Diaspore Morphology
Total Page:16
File Type:pdf, Size:1020Kb
ECOGRAPHY 25: 336–344, 2002 Gap colonization in the Patagonian semidesert: seed bank and diaspore morphology Roberto J. Ferna´ndez, Rodolfo A. Golluscio, Alejandro J. Bisigato and Alberto Sorianoc Ferna´ndez, R. J., Golluscio, R. A., Bisigato, A. J. and Soriano, A. 2002. Gap colonization in the Patagonian semidesert: seed bank and diaspore morphology. – Ecography 25: 336–344. We report work on a cold, windy South American steppe dominated by tussock grasses and shrubs of small stature that, together, cover only half of the soil surface. Our objective was to find out why seedlings and juveniles of these dominant species are generally absent from the bare or poorly-populated spots (gaps) that exist between established individuals. We hypothesized that matrix-forming species fail to colonize gaps because of a lack of properly-placed seeds, contained in diaspores which because of their morphology are blown away from gaps that otherwise would constitute safe sites for recruitment. We evaluated diaspore size and wing loading (weight:area ratio) for all common species in the community, collected seed-bank samples in different occasions and microsites, and performed detailed field observa- tions for one gap-dwelling species during several years. We found that: 1) Seeds of the dominant, matrix-forming species are uncommon in the soil bank of the center of the gaps between established vegetation. 2) Seeds of the dominant species are more abundant towards the edges of the gaps than at their center. 3) Diaspores of those species present in the seed bank of the gaps are smaller than diaspores of absent species; contrary to expectations, not all gap-dwelling species had larger diaspore wing loading than non-gap species. 4) Seeds and adult densities of the most common gap species (the annual Camissonia dentata), were correlated between them and across subsequent years. We conclude that it is not an overall shortage of seeds that precludes the dominant species from becoming established in the gaps, but rather the seeds’ uneven spatial distribution. We further argue that gaps would be suitable sites for recruitment, but large diaspore size makes seeds of the dominant species to be blown away by the strong westerly winds. R. J. Ferna´ndez ( [email protected]), R. A. Golluscio, A. J. Bisigato, IFEVA- Ecologı´a, Uni6. de Buenos Aires – CONICET, A6. San Martı´n 4453, Buenos Aires C1417DSQ, Argentina (present address of A.J.B.: Centro Nacional Patago´nico, CON- ICET, Bl6d. Brown s/n, Puerto Madryn, Chubut U9120ABT, Argentina). Vegetation in arid and semiarid zones typically displays tence of low-cover or bare patches. This work examines some sort of ‘‘mosaic structure’’ of high- and low-cover the role of seed availability in the persistence of these patches (Aguiar and Sala 1997). Substantial progress spots for a particularly windy semiarid environment. has been made in understanding the processes deter- In a cold Patagonian semidesert of southern South mining the dynamics, and especially the persistence, of America, adult shrubs and perennial tussock grasses high-cover patches (e.g. Garner and Steinberger 1989, form a high-cover vegetation matrix that covers ca 50% Schlesinger et al. 1996, Klausmeier 1999). Much less is of the soil surface. The rest is occupied by bare or understood, however, about the causes of the persis- poorly-populated spots (gaps) of an average diameter of 35 cm (Soriano et al. 1994). The dominant matrix- forming species are three small-statured shrubs c deceased. Accepted 26 September 2001 Copyright © ECOGRAPHY 2002 ISSN 0906-7590 336 ECOGRAPHY 25:3 (2002) (Mulinum spinosum (Cav.) Pers., Senecio filaginoides Desert (Reichman 1984), and the Negev (Boeken et al. DC and Adesmia campestris (Rendle) Skottsb.) and 1998). three tussock grasses (Stipa speciosa Trin. et Rupr., S. Our main hypothesis was that Patagonian matrix- humilis Cav., Poa ligularis Nees ap Steud.). Gaps host a forming species fail to colonize gaps because of a lack number of annual and small perennial forbs but are of properly-placed seeds, which in turn is caused by generally unoccupied by seedlings and juveniles of the their diaspore morphology. In this windy environment, matrix-forming species (Table 1). diaspores of certain shape and size would be unable to For the non-dominant grass Bromus pictus (Hook), remain over the relatively smooth surface of the gaps Aguiar and Sala (1997) showed that, despite an abun- that otherwise would constitute safe sites for recruit- dant seed rain, establishment in gaps is infrequent ment. Diaspore morphology determines not only pri- because the strong prevalent west winds blow seeds mary dispersal by wind (Greene and Johnson 1986), but away from these bare-soil areas (see also Alippe and also retention at the soil surface (Chambers et al. 1991). Soriano 1978). Soil-water availability during the sum- A parameter that has been successfully used to predict mer drought is higher in the center of the gaps than diaspore movement during free fall is ‘‘wing loading’’, closer to the plants encircling them (Soriano and Sala defined as the ratio between the weight and the area of 1986). The establishment of Bromus seedlings in gaps is the diaspore (Matlack 1987). For a given wind intensity possible as long as its relatively large diaspores (ca and initial height, the distance traveled is inversely 15×3 mm) are experimentally attached to the surface related to the square root of the wing loading (Aguiar et al. 1992, 1994). Theories of population dy- (Augspurger and Franson 1987, Greene and Johnson namics would classify this case as one of recruitment 1993). We reasoned that, under strong winds like the uncoupled from disturbance (Grubb 1986), in which Patagonian ones, this weight:area ratio would not only affect primary dispersal but also later movements on populations are neither strictly seed- nor microsite-lim- the soil surface. Another parameter we chose to study ited (Crawley 1990, Eriksson and Ehrle´n 1992). was diaspore size because, regardless of the rugosity of Vegetation patterning may in many cases be ac- the soil surface, the smaller the diaspore, the better its counted for by after-landing movements of seeds anchorage capability (Chambers et al. 1991). (Chambers and MacMahon 1994), also known as sec- We evaluated diaspore size and wing loading for all ondary dispersal. In Patagonia, dominant grasses and common species in the community, collected seed-bank shrubs have an abundant seed production in most years samples in different microsites, and performed detailed (Soriano 1983, Ferna´ndez et al. 1992). We wondered field observations in one of the gap-dwelling species whether seeds of these species – having diaspores of during several years in order to evaluate four predic- different size and shape than those of Bromus – are tions that follow from our hypothesis: 1) Seeds of most able to persist in the gaps or also accumulate against matrix species are absent from, or very poorly repre- established individuals. This kind of wind piling-up has sented in, the seed bank of the gaps. Our definition of been reported for other environments, such as western seed bank is broad, and includes all viable seeds in the Australia (Mott and McComb 1977), the Sonoran surface soil. 2) The abundance of seeds of the matrix species will be much higher close to the established Table 1. Floristic composition of the gaps between Patago- vegetation (the gap edge) than towards the center of the nian bunchgrasses. Data obtained at the end of the growing season within circular 30-cm diameter plots (January 1992; gaps. 3) Diaspore morphology of species found in the n=200). Asterisks indicate those species also found in the seed bank of the gaps will be different from that of seed bank of the gaps (cf. Table 2). species absent from it. We expected gap species to have SpeciesGrowth form1 Frequency (%) smaller diaspores, with a larger wing loading, than non-gap species. A corollary from this – taking into Camissonia dentata* Annual dicot 57 account that most gap species are of a very short Microsteris gracilis* Annual dicot 39 stature – is that gap species will be poor colonizers of Poa lanuginosa Rhizomatous grass 25 Poa ligularis* Bunchgrass 13 distant gaps, thus tending to re-colonize the same gaps (seedlings) year after year. We chose to work with the most Calceolaria Rhizomatous 11 common gap colonizer, Camissonia dentata Cav. Reiche polyrhiza* deciduous dicot (synonyms given below), to test our last prediction: 4) Polygala darwiniana Stoloniferous 10 evergreen dicot The number of both extant plants and seeds of each Adesmia lotoides Rhizomatous 8 species in a given gap will be positively correlated evergreen dicot across subsequent years. Gilia laciniata* Annual dicot 7 Cerastium ar6ense Rhizomatous 4 deciduous dicot Oxalis Rhizomatous 3 squamoso-radicosa deciduous dicot Site description Other species Perennials B2 This work was performed in central-western Patagonia, 1 See Golluscio and Sala (1993) for details. in an experimental field of the Instituto Nacional de ECOGRAPHY 25:3 (2002) 337 Tecnolog´ıa Agropecuaria (INTA) located near R´ıo fully inspected in order to return to the fine-soil frac- Mayo (45°41%S, 70°16%W). From the phytogeographic tion any material that might contain seeds. The fine soil point of view, the vegetation is typical of the Occidental was spread, as a 1-cm thick layer, on paper towels District of the Patagonian Province (Paruelo et al. placed on plastic flat containers with drainage holes. 1991). All our observations took place in a representa- Soil in the containers was spray-watered regularly while tive 2-ha plot that had been closed to large herbivores maintained in a growth chamber at 15°C with 12 h of since 1983. According to a survey performed on rela- light per day.