Serviços customizados
texto em 
Inglês (pdf)
Artigo em XML
Referências do artigo
Enviar este artigo por email
Citado por SciELO 
Similares em
SciELO 
Permalink
Introduction
In the last decades, expansion and intensification of agriculture and cattle breeding have been recognized as main contributors to global environmental problems, including deforestation, climate change, and biodiversity loss (Gibbs et al. 2010). At the same time, cattle breeding is the means of subsistence for one million four hundred thousand people in the world (Herrero et al. 2013, 2016), and provides feed to eight hundred million people in food insecurity (Engler et al. 2018). Due to growing demand for animal protein (Valin et al. 2014), human population growth and per capita income raise (Alexandratos and Bruinsma 2012), a total of 517 million cattle and buffalo head is expected in Latin America and the Caribbean for 2050 (Alexandratos and Bruinsma 2012). Between 2010 and 2018, cattle breeding increase was related to loss of around 2,8 million ha/year in forests (FRA 2020 RSS). This pressure on ecosystems is exerted, partially, due to conventional cattle breeding practices are developed in extensive pasture areas without trees (Herrero et al. 2016), subject to continuous overgrazing of forages low in protein (7 %) and of low digestibility (55 %). This contributes to low cattle load (± 0.6 animal units ha−1), which makes tropical cattle breeding to record poor productivity and competitiveness (González et al. 2015). Extensive cattle breeding promotes unproductive pastures due to soil erosion as a result of excessive grazing, low humidity retention, and high temperature due to lack of shade, so larger number of forest areas are continuously turned into new grasslands. Another factor contributing to transform land in the tropics is the flourishing commerce of feedstuffs for meat production.
Meat, as compared to other European basic food products, contributes more than 50 % to land transformation for feedstuff production (Crenna et al. 2019 and EC-JRC 2022). In South America, half million forest hectares are replaced for soybean production (Goldman et al. 2020), out of which around 77 % are devoted to forages for intensified production of animals (fattening lots) (FAO 2020). Meat supply for a growing market will promote, even more, the conversion of forests into agricultural lands for feedstuff and pasture production devoted to conventional cattle breeding.
In such context, silvopastoral systems in Latin America represent a sustainable alternative to cattle breeding, as part of the solutions based on nature to catch carbon (López-Santiago et al. 2019), increment profitability in cattle production (Chará et al. 2019), release areas non-suitable for cattle breeding to be devoted to ecological recovery (Calle et al. 2012), and, thus, protect part of the great biodiversity accumulated in the tropics. Silvopastoral systems consist of a well-designed combination of different plant strata such as pastures, grasses, shrubs, trees and/or palms. Particularly, intensive silvopastoral systems (SSPi) provide innovative solutions since they are based on sowing protein-rich forage plants, due to their high content of N and P, serving to be directly browsed by cattle and being a key element in the system. Moreover, electric fences are used in cattle rotation, which permits controlling soil erosion by preventing constant animal tread, whereas fresh water is provided with permanent access. Therefore, cattle are prevented to move toward natural water sources. Some plant species currently used in shrub strata in silvopastoral systems are Leucaena leucocephala, Tithonia diversifolia, Guazuma ulmifoliay Sambucus peruviana (Murgueitio et al. 2016). They are sown at high densities (between 5,000 and 40,000 ha-1) in ranges. L. leucocephala promotes high nitrogen fixation and transference, whereas T. diversifolia favors phosphorus solubilization in acid soils. This benefits, in turn, associated pastures (Bacab 2013 and González 2013).
In Mexico, it is estimated that around ten thousand hectares of intensive silvopastoral systems of Leucaena leucocephala var. Cunningham have been sown in nine states of the Republic (Murgueitio et al. 2016 and Chará et al. 2019). Being native of Mexico, Central America and the Antilles (Grether et al. 2006), this species predominates in deciduous tropical forests with high demands of light (heliophile). It tolerates stony soils, with little resistance to flooded ones, and, above all, it grows suitably on neutral and alkaline soils (Murgueitio et al. 2016). That is the reason why Tithonia diversifolia is recommended to be sown on acid soil conditions. Also, as it is native to Mexico and Central America, this species is pioneer in different ecosystems, because it is commonly found in clearings, road verges, and anthropized places, surrounding from tropical forests up to pine-oak forests. It grows on different soils (clay, sand, lime), at 0 until 2500 m above sea level, and its use as forage species has been increased because it has a quality of up to 28.8 % protein. It is apt for different types of livestock (sheep, goat, swine, cattle) and it is widely adapted to different soils and climates (Calle and Murgueitio 2008 and Mauricio et al. 2017). Tithonia diversifolia has been used in intensive silvopastoral systems of South America and Cuba, in densities between 6 thousand and 40 thousand plants/ha. In Mexico, there are examples of its use as forage to be cut and carried in Veracruz, with promising results (Romero González 2018). Nevertheless, as part of intensive silvopastoral systems, it is still scarcely used. In the project Sustainable Cattle Breeding in the Mexican Tropics, in 2018-2020 collaborations were started with cattle farmers in the region of Los Tuxtlas, Veracruz. Currently, three ranches of dual-purpose cattle production have put in practice the vegetative reproduction of Tithonia diversifolia for the animals directly browse this species (figure 1). The purpose is to shift to seed sowing, so a research line has been developed to increase knowledge on sexual reproduction in this species.
Figure 1 Plantation of cuttings of Tithonia diversifolia for direct browsing in three ranches of Los Tuxtlas, Veracruz (Pictures: R. Santos-Gally).
The reproduction of T. diversifola in intensive silvopastoral systems has been developed mostly in a vegetative form (Ruiz et al. 2014), since seed sowing results in low germination (Zapata and Silva 2016). However, low germination has been proved in seeds sown fifteen days after collection, whereas those sown four months after being stored at room temperature (19 ºC) showed germination of 90 % (Santos-Gally et al. 2020) (figure 1). These outcomes corroborated latency of T. diversifolia in plant populations of Mexico and in plant populations introduced in Africa and Asia (Agboola et al. 2006, Wen 2015, Rodríguez et al. 2019 and Santos-Gally et al. 2020), which could account for the germination percentages reported (Ruiz et al. 2018). Likewise, there is discrete variation in the morphological expression of the fruit (achenes) between the heads (inflorescence) of T. diversifolia, which is related to the fact that in some populations there are germination differences (Santos-Gally et al. 2020); for example, seeds derived central achenes of a population from southern Veracruz had significant differences in respect to seeds derived from peripheral achenes (87.5 % and 68.7 %, respectively) (Santos-Gally et al. 2020). In agricultural contexts, fruit heteromorphism can be disadvantageous because it results in germination and terraced plantation of platelets reducing yield and making harder pest management and plant growth (Mitchell et al. 2017). Therefore, morphos rate analysis between populations is of great relevance, particularly if wild plant material collection is planned for seed sowing.
Germination success could be also associated with sexual reproduction; that is, viable fruit and seed production that can germinate to produce the next generation. In hermaphrodite plants (presence of both sexes in the same individual), sexual reproduction is conducted by self-fertilization or cross-fertilization (Barrett 2014). Two opposing forces condition evolution in the first: advantage of transmitting 50 % of self-compatible genes and inbreeding depression. Inbreeding depression refers to reduction of viability and fertility, or both, in offspring from self-fertilization, as compared to that from crossing of genetically different individuals. Self-fertilization provides reproductive certainty, especially in ecological situations where the number of possible partners is scarce (in a colonization process or bottleneck) or where there is scarcity of pollinators (pollen grains are poorly spread) (Jarne and Charlesworth 1993). Crossing between genetically different individuals (crossing-over) promotes genetic diversity, with new alleles in a lineage that can contribute to local adaptation (Linhart and Grant 1996).
In order to ensure crossing between individuals, there are several mechanisms. The most spread in angiosperms is the system of self-incompatibility, which prevents self-fertilization (Barrett 2014). A system of self-incompatibility combines physiological, genetic (diallelic), sporophytic and biochemical mechanisms to avoid self-fertilization, which promotes exclusive fertilizations if pollen is transferred successfully between individuals (Lloyd and Webb 1992). In allogamous species (crossed pollination), viable seed production would be determined by pollen transference between genetically different partners. Determining seed production by self-pollination or crossed pollination permits establishing the presence of an incompatibility system. Due to the importance of T. diversifolia for implementing silvopastoral systems by seeds, this research compiles data referring to the success of seed production after various treatments which served to determine whether T. diversifolia has a system of self-incompatibility or not.
Development
Fruit heteromorphism
During October and November in 2019, between three to ten flowerheads were collected, out of 3-10 individuals of 32 populations of T. diversifolia in Veracruz, Oaxaca and Chiapas. Achenes were classified as central when they had crown with subequal and peripheral edges and crown with absence of edges (after Blake 1921). Out of a total of 28,270 achenes from 182 individuals, 12,242 central and 16,028 peripheral achenes were characterized. In total, seven populations tended to central achenes, 13 populations tended to peripheral achenes and twelve had a similar ratio of central and peripheral achenes (figure 2).
Incompatibility system
During March in 2020, manual pollinations were performed to determine if T. diversifolia has a system of incompatibility. Two groups of manual pollinations were set out: self-pollination and crossed pollination in 4,994 and 1,525 flowers from 29 individuals, respectively. Ten thousand and twenty-eight (10,028), and two thousand two hundred and forty-six flowers (2,246) from 29 and 9 individuals were divided into control and autonomous self-pollination groups. Flowers in the crossed pollination group were emasculated prior to the dehiscence of the anthers. The four groups were randomly assigned to four lots in different plant positions. The number of repetitions per group was balanced among the individual.
Manual pollinations were performed daily for three days in lots that were marked and bagged with nets (pore size of 0.1 mm) to prevent any possible pollen contamination. Crossed pollinations involved only on donor of pollen, chosen randomly to transfer pollen to all the bloomed flowers in the lot. Self-pollination did not include emasculated flowers, which were pollinated by hand using pollen from the same flower. In both groups, flower buds surrounded flowers were removed to avoid confusion when collecting the achenes. The control for open pollination involved lots marked by buds that bloomed some days previous to or during the trial, and these buds remained available to visitors throughout the trial. With the achenes/seeds produced by these lots, it was possible to determine pollen limitation in the population or if manual pollinations were effective, which could account for the lack of significant differences between crossed pollination and the control. Finally, nine individuals from 25 lots were bagged to determine whether plants might bear fruit through non-manipulated self-pollination (spontaneous). Fruit was harvested three or four weeks after pollination, when the number of achenes with seed and aborted seed (empty achene) was counted. In order to test the effect of the different groups on fruit production (flower that turns into fruit), a beta regression model was used with beta distribution and logit binding functions. The group was included as categorical explanatory variable. For both analyses, betareg package (Simas and Rocha 2006) was used in R (Team 2018). The ratio of averages of self-pollination and crossed pollination groups was used to measure the index of self-compatibility proposed by Becerra and Lloyd (1992). Values equal or inferior to 0.75 show that species are self-compatible.
The outcomes showed that T. diversifolia had a system of incompatibility. The value of the self-compatibility index was lower than 0.75 (SCI = 0.55). Figure 3 shows spontaneous pollination and self-pollination had significantly lower success, as compared to crossed pollination and the control (P <0.0001). Average ratio of fruit produced per pollination was 65.31 %, whereas in spontaneous self-pollination it was 2 % (figure 3). The control did not differ significantly from the group of crossed pollination; therefore, it might be stated that there was no limitation due to pollen. The non-manipulated self-pollination group showed that Tithonia diversifolia demands biotic vectors to transfer male gametes and, thus, ovule fertilization.
Conclusions
In different countries, Tithonia diversifolia is of particular interest because of its high protein content (28 %) and due to the fact that it is able of fixating phosphorus through the association with mycorrhiza, which helps soil recovery and enhances productivity (Mahecha et al. 2007 and Rivera et al. 2011). In addition to these properties, its response capacity to herbivory turns it into an excellent candidate to be used as feed in intensive silvopastoral systems (Calle and Murgueitio 2008). However, in those countries where T. diversifolia has been used for such purposes, plant reproduction has been limited to vegetative reproduction due to the low success in germination. The latency found in different populations of Mexico may be broken if seeds are stored during four months before the plantation (Santos-Gally et al. 2020), as reported in previous studies on plants introduced out of their native distribution area (Wen 2015). Peripheral achenes had lower germination success without pre-germination treatment; therefore, they are attributed higher latency (Santos-Gally et al. 2020). In this research, 41 % of the populations showed deviation to peripheral achenes, whereas 22 % constitute populations with deviation to central achenes. The remaining 36 % of the populations are kept in a balance between the two types of achenes. The ratio of achenes in populations should be taken into account when planning seed collection to put into practice intensive silvopastoral systems.
The presence of a self-incompatible system in T. diversifolia highlights the importance of having genetically different individuals to obtain the largest number of viable seeds. If choosing to set a cultivated plot of plants to obtain seeds, it is advisable that plants in this plot come from a natural population as numerous as possible, as well as seeds be chosen from individuals between 10 and 15 meters of distance with the goal of favoring genetic differences between individuals and diminishing the risk of inbreeding depression. Although in this specific research, the presence of inbreeding depression was not analyzed in self-incompatible species (Raphanus sativus and Leontodon autumnalis). There is evidence of decline in seed production in plants produced from self-pollination (Nason and Ellstrand 1995 and Picó and Koubek 2003). Plants from crossing between relatives may present negative effects due to reduction in genetic variability, which could be expressed in production, germination, or plantelet growth decline (Cheptou et al. 2000).
Due to the interest in T. diversifolia for putting into practice intensive silvopastoral systems through seeds, its obtainment should be concentrated on populations of more than 200 individuals with the aim of avoiding the harvest of empty seeds. Moreover, it should be noted that if seeds are obtained from plants that were fertilized in a vegetative form, the resulting seeds may have a loss of genetic variability that could have non-desirable effects on the growth and production of fertile seeds. In the future, it is important to research if the decrease in genetic variability could also affect forage quality or hebivory resistance, or both. This research proves seed production through crossing-over or natural pollination is 50 % higher than through self-fertilization; therefore, seeds from natural populations in the region under study may be used to plant intensive silvopastoral systems in the future. These results also highlight the importance of natural pollinators for seed production and the successful reproduction of this species.
