Successful cross of Arachis duranensis as female with A. ipaёnsis

Summary : Arachis hypogaea L. originated in South America and has been taken to most of the tropical and sub-tropical parts of the world as a valuable food crop with high protein content and a source of high energy unsaturated oil. The origin of the cultivated peanut, 2n = 4 x = 40, has been the subject of many discussions, but the primitive parents have been agreed on by most as A. duranensis being the A genome donor and A. ipaënsis the B donor; both diploids with 2 n = 20. Whether the chromosome doubling of this hybrid occurred in a natural setting or in the garden of a hunter-gatherer-cultivator is also a subject of debate, but most likely it occurred in nature. Molecular analyses have established that A. duranensis was the female of the cross. Until recently no one had been successful in making and establishing plants of the cross in that direction. However, the reciprocal cross is easily accomplished and has been reported several times. The primary objective of this paper is to report the successful cross and development of hybrid plants, amphidiploids and populations from the hybrid, A. duranensis × A. ipaënsis .


Introduction
The peanut (Arachis hypogaea L.) has been one of the most important food legumes for peoples living in the tropical and sub-tropical regions of the world for many centuries.It also has become an important food in developed areas of the world because of its high protein, unsaturated oil content, and its desirable flavor.A large percentage of the world's peanut production is utilized for its oil which is very popular for cooking because of its long-lasting quality and nutritional value.
The genus Arachis most likely originated on the old Brazilian Shield in west central Brazil or northeast Paraguay sometime around the mid-Tertiary uplift of the Shield (Gregory et al., 1980;Krapovickas & Gregory, 1994, 2007).The remnants of that eroded uplift exist even today in the area described above.Also, the three most morphologically primitive Arachis species described to date are still found growing in that region (Krapovickas & Gregory, 1994, 2007;Valls et al., 2013).Molecular estimates of that time of origin of Arachis have placed it around 3 million years ago (Bertioli et al., 2016a;Bertioli et al., 2016b).From the uplifted location, the genus has spread over much of Brazil, eastern Bolivia, most of Paraguay, northern Argentina, and western Uruguay.Some of this distribution has been through physical plant growth, placing seeds as much as 1 to 3 meters from the base of the mother plant.Additionally, with the geocarpic fruit of Arachis, running water has undoubtedly been involved in their dispersal.Documented movement by human hands has also played a significant part (Krapovickas & Gregory, 1994, 2007;Simpson et al., 2001;Custodio et al., 2005;J. F. M. Valls, personal communication).
The evolution of the genus has led to description of 83 species, to date, which have been classified into nine taxonomic sections (Krapovickas & Gregory, 1994, 2007;Valls & Simpson, 2005, 2017;Valls et al., 2013, Seijo et al., 2021).Based on the collection efforts, clearly the direction of evolution of Arachis has been from east to west on the South American continent.The most advanced section in the genus is Section Arachis which contains 33 species and includes the cultigen, A. hypogaea.Most of the wild Arachis species have 2n= 20 chromosomes.However, the cultivated peanut contains 2n= 4x= 40.Husted (1936) proposed that peanut is an allopolyploid based on the high formation of bivalents and the presence of only one pair of small chromosomes (A chromosomes), while Smartt et al. (1978) further named the two different chromosome sets as belonging to the A and B genomes.From the 1940's though the 1960's, several potential donors of the A-genome were proposed; most were undescribed and carried names listed as nom.nud.Until the mid-1970's, only one non-A group of plants in the Arachis section was known to exist, and it was later described and named A. batizocoi Krapov.& W.C. Greg.Smartt et al. (1978) proposed that this species might be the B-genome donor to the cultivated peanut.During the early 1970's, this author had initiated an introgression program with Arachis using this only known non-A as the B-genome parent to transfer leafspot resistance into the cultigen (Simpson, 1991).In 1976, extensive collection of Arachis germplasm was initiated and funded, in part by the International Board for Plant Genetic Resources (IBPGR) (Krapovickas & Gregory, 1994 page 7).Several new non-A genome materials were collected over the following years, providing more options to consider as the B-genome donor for A. hypogaea (Krapovickas & Gregory, 1994, 2007;Valls & Simpson, 2005;Robledo & Seijo, 2010;Seijo et al., 2021).Kochert et al. (1991) proposed that A. duranensis Krapov.& W.C. Greg.and A. ipaёnsis Krapov.& W.C. Greg.were the Aand B-genome donors to the cultivated peanut.Since then, numerous molecular studies have supported these two species as the likely genomic parents to the cultigen (Kochert et al., 1996;Burow et al., 2001;Seijo et al., 2007;Grabiele et al., 2012;Bertioli et al. 2016a, b).
During the development of data to identify and separate various groups of wild peanut accessions into species, plant morphology played a major part, but cross compatibility also proved to be a valuable tool in this difficult work (Gregory & Gregory, 1979;Singh, 1988;Simpson, 1991;Singh & Smartt, 1998).As time progressed, isozymes proved useful (Krapovickas, 1969;Lu & Pickersgill, 1993).Recently molecular analyses have become widely used.A major contribution of early molecular work was the capability to identify the female parent in a cross between two species (Kochert et al., 1996;Grabiele et al., 2012, and others).Molecular analyses determined that the female in the cross which formed cultivated peanut was A. duranensis (Kochert et al.,1996;Seijo et al., 2004;Robledo et al., 2009).Herein lies the problem, which is the focus of this paper, i.e., until recently no one had successfully made nor documented making the cross with A. duranensis as the female.Gregory & Gregory (1979) reported on diallel crosses of 91 wild and cultivated Arachis parents, and other researchers, including the author, have conducted several crossing programs involving materials collected since 1976 (many results unpublished).
The author's program has attempted the cross, A. duranensis × A. ipaёnsis, several times from 1980 to 2010 without success, using the accession K 7988 for A. duranensis and K 19455 and later the KGBPScS 30076 for A. ipaënsis.The author successfully made the reciprocal hybrid several times (unpublished).Fávero et al. (2006) reported the hybrid, A. ipaёnsis × A. duranensis, and produced several breeding lines after crossing the chromosome doubled progeny with each of four botanical varieties of the cultigen, A. hypogaea.Fávero did not make specific comparisons of the amphidiploid progeny to A. monticola or A. hypogaea; such activity was not a part of her objectives (personal communication, 2007).Grabiele et al. (2012) reported on molecular studies of the A. duranensis materials of northwest Argentina, and concluded that materials from the Rio Seco Valley were the most likely donors of the A-genome to the cultigen, A. hypogaea.More specifically, their accession Se 2741 (corresponding to our collection from the same population, KGBPScS 30067), was identified as the specific accession being the A-genome donor.Studying the publication of their work, I determined that in 1977 our team had collected from virtually the same sites as those reported by Grabiele et al. (2012).Thus, crosses were attempted with five collections as the female and A. ipaёnsis (KGBPScS 30076) as the male (Table 1).An oral presentation on the hybrids was made (Simpson, 2017) and a detailed paper was published (García et al., 2020) adding valuable information about this difficult cross.The hybrids are herein reported.

Materials and Methods
All plants in this study were grown in greenhouses to maintain purity of our germplasm.Outside grown wild Arachis are subject to numerous cross-pollination events virtually every summer day at our location (32°14'42"N, 98°11'49"W, 400 m).
The A. duranensis accessions used in these crosses are shown in Table 1.We included a check accession from outside the Rio Seco Valley, K 7988.The male of the crosses was one plant of A. ipaёnsis (KGBPScS 30076, PI 468322).
Female plants were grown in ½ bushel fruit baskets in a sandy loam soil mix that is ca.75% sand.Commercial inoculum was added to the mix.Seeds were pre-germinated and transplanted to baskets 3 to 4 D (days) after germination.
Plants initiated flowering ca. 25 DAP (days after pollination).All flowers that were not emasculated for crossing were picked as buds before daybreak (beginning of anthesis) each morning.Emasculations were made between 7:30 and 9:00 PM, CDT (central daylight time).
Pollinations were made between 7:00 and 8:00 AM the following morning.The emasculation/ pollination technique is one adapted from a former student from Senegal, whereby none of the petals are removed and the bud is closed promptly after emasculation.The treated buds are then either covered with a moist paper towel or misted with an atomizer.Pollination the following morning was accomplished by cutting the keel, including anthers and stigma, deep into the bud of the male flower.After opening the petals of the emasculated (female) flower, the male keel including anthers and stigma, was fitted over the stigma of the emasculated bud.The pegs emerged in ca.5-8 days, and a nylon string was tied around them and fixed to a wood stake with the date and cross ID (identification).Seeds were harvested at ca. 55 days after pollination.Breaking dormancy on the hybrid seeds was attempted with powdered ethylene compound and/or ethylene gas.Chromosome doubling of hybrids was accomplished with colchicine using a technique that was described by Banks (1977).A three-dayold seedling, ca. 5 cm long, was inverted in a vial and immersed to the base of the cotyledons in a 0.2% concentration of colchicine.The vial was stoppered and placed in a chamber maintained at 30 °C for 8 hr.The treated seedling was gently washed for ca.0.5 hr., then planted.
Pollen counts were made on the resulting amphidiploid by staining mature pollen with aceto-carmine mixed with glycerin, v/v.Full counts consisted of three flowers counted on different days with 500 random grains counted per flower, for a total of 1,500 (Gregory & Gregory, 1979).
A total of 17 plant traits were measured (Table 3) on the 21 plants studied (Table 2), the measurements were made with a hand-held digital caliper.

Results
Table 1 shows the numbers of pod segments harvested from the four crosses.All four of the accessions of A. duranensis collected from the Rio Seco Valley of NW Argentina produced seed.The total from all crosses was 27.Embryos of the attempted crosses with the K 7988 accession all aborted early in the developmental stages.It has been common practice in the author's breeding program to allow seed, especially interspecific Table 2. Identity of the Arachis plants in Fig. 2 and the pods in Figs. 3 and 4.

Figure Code Accessions
A -M Plants growing from the 13 seed harvested from the new amphidiploid from the A.  1975).When it was decided to germinate some of the seed for colchicine treatment, all attempts to break dormancy were unsuccessful for more than 25 months.The dormant seeds remained firm in the "rag-dolls" in the germinator, and seeds were re-treated numerous times with some form of ethylene (gas, liquid, powder), without success.However, after 26 months, two of the seeds germinated following an ethylene treatment.The two seedlings were treated with 0.2% colchicine and one amphidiploid was obtained (Fig. 1).The second plantlet did not double and was used for other studies (see Fig. 4).García et al., 2020, reported a similar dormancy issue with the hybrid seeds they made and studied.
Pollen counts of the amphidiploid were 98% stained, and pegs were produced on the two cotyledonary laterals and one secondary lateral (See Discussion below).
As a precaution, cuttings were made of the amphidiploid as soon as sufficient growth had occurred.During the second month of growth, a problem developed in the R/O (Reverse Osmosis) watering system that caused high concentrations of salt to enter the water.With the action of the salt, the amphidiploid started to deteriorate, which brought the problem to our attention.However, it was too late, and the plant died in spite of our efforts.The cuttings were also affected and eventually, died as well.Before death of the plant and cuttings, we were successful in obtaining 13 viable seeds.
The 13 seeds were planted for study and for comparison to three accessions of A. monticola, four primitive accessions of A. hypogaea, and the most primitive cultivar from Texas peanut production history.These 21 accessions are identified in Table 2 and are shown in Figures 2 and 3.
Measurements of the 17 plant traits (measurements not shown, only the means and their analysis, Table 4) taken at 75 DAP for comparison between the new amphidiploids and the more evolved tetraploids showed that a majority of the comparisons were not significantly different.For example, the main axis of the two groups were essentially equal; whereas, the cotyledonary laterals were significantly different in length, with the new amphidiploids having a much longer lateral branch.This is as expected in our experience with interspecific hybrids.Another interesting comparison is leaflet length and width.Leaflet length was not significant, but width was significant.Again, as one would expect, no great differences in leaflet length were observed, but greater variability has evolved in leaflet width, with the most primitive accessions containing narrower leaflets.
The fruit characteristics of the new amphidiploids closely resembled those of the A. monticola accessions (Fig. 3).However, significant evolution of pod traits has occurred in selections of the A. hypogaea-like derivatives made by early cultivators (See Fig. 3 Q, R, S,  T and U).
(2020), presented detailed analyses of this possible process in their A. duranensis × A. ipaënsis progeny.

Conclusion
The two tetraploid species in section Arachis of the genus Arachis, A. monticola and A. hypogaea, probably originated from the same cross between A. duranensis and A. ipaënsis.No doubt, the original cross occurred in nature because the resulting plant(s) would have been sterile and, thus, rogued out if it happened in the garden of a hunter/gatherer/cultivator.
The two-year dormant period of the hybrid seed in our study presents some amazement that the event occurred at all.After the chromosome doubling occurred, the amphidiploid seed germinated and grew into a primitive form of A. monticola.This statement is based on the doubled plant observed in this study.The original doubled hybrid(s) would likely have been recognized as something different from the A. duranensis and/or A. ipaënsis plants growing at the time, whether in nature or in the cultivator's plantings.Information indicates that human hands played a role in early development of both A. monticola and A. hypogaea (Krapovickas & Gregory, 1994, 2007).So, if the chromosome doubling occurred in nature, once discovered, the primitive A. monticola probably soon appeared in the cultivator's plantings.A widespread distribution, followed by 3500 to 3800 years of human selection and environmental pressures, led to the cultivated peanut that we know now.

Fig. 3 .
Fig. 3. Fruits (pods) from the 24 plants shown in Fig. 2. See Table 2 for identification of accessions.Note reticulation of 13 new amphidiploids (A-M) as compared to three accessions of Arachis monticola (N, O, P).

Table 1 .
Data from five attempted crosses of Arachis duranensis × A. ipaёnsis