**1. Introduction**

Hexaploid triticale (×*Triticosecale* Wittmack, 2*n* = 6*x* = 42, AABBRR) is one of the few artificial crops cultivated at a large scale. Triticale combines the quality traits of wheat (*Triticum aestivum* L.) and the adaptation abilities of rye (*Secale cereale* L). Over the last three decades the global harvested area of triticale has constantly increased (2,101,405 ha in 1996; 3,662,363 ha in 2006, and 4,157,018 ha in 2016) [1], and the range of uses has also grown. Forage production is the principal end use for this crop, but there are new niches proposed, such as: biofuel production [2,3], baking [4], brewing [5] and food production [6]. As a recent man-made crop, triticale suffers narrow genetic variability for breeders to select upon.

In triticale there has been a breakdown of resistance against leaf rust (caused by *Puccinia triticiana* Eriks.) and stripe (yellow) rust (caused by *P. striiformis* f. sp. *tritici* Westend.). As the harvested area of triticale has increased, new pathotypes of *Puccinia* have evolved, moving from wheat and rye into triticale [7]. Both pathogens can reduce the grain yield by 40% [8,9]. There is an urgen<sup>t</sup> need to improve the genepool of triticale, and introduce genetic resistance to *Puccinia* infections. There are approximately eighty leaf rust resistance genes identified in Triticeae, and nearly the same amount of stripe rust resistance genes have also been identified. Some of these genes have already been transferred

from wild relatives into the wheat genetic background [10]. Recently, several attempts were made to transfer rust resistance genes from *Aegilops*, *Agropyron* and *Triticum* species into triticale [11–15].

*Aegilops* species are closely related to wheat (and triticale, *per se*) and carry a number of valuable traits, which have been e ffectively incorporated into wheat by developing wheat–*Aegilops* hybrids and deriving addition, substitution and translocation lines [16]. *Aegilops kotschyi* Boiss. (2*n* = 4*x* = 28 chromosomes, U- and S-genomes) is a wild tetraploid goatgrass native to Northern Africa, the Mid-East, and Western Asia. *Ae. kotschyi* germplasm is exploited in wheat breeding [17] as a source of high grain protein, iron and zinc [18]. Moreover, Antonov and Marais [19] observed leaf rust resistance that was e ffective against the infection of *Puccinia triticina* in *Ae. kotschyi*. Marais et al. [20] identified the *Lr54* and *Yr37* leaf rust and stripe rust resistance genes, and developed aT2DS.2SkL wheat-*Ae. kotschyi* translocation line. The first *Lr54* + *Yr37* marker was developed by Heyns et al. [21]. Moreover, translocation gene sequences were cloned and specific SSR markers were developed [22].

Homoeologous recombination based engineering is the most common way for e fficiently utilizing the wild relative gene pool for crop improvement [23]. The generation of translocation lines is the most promising pathway for the exploitation of alien germplasm in crop breeding [23]. In distant hybrids, unpaired chromosomes are present as univalents during meiosis. Monosomic chromosomes are prone to centric breaks at anaphase I of meiosis, which misdivide and the broken ends fuse during the interkinesis of meiosis II [24–26]. Fusion of the misdivided products may result in the formation of a Robertsonian translocation (RobT) [27].

Several steps are required to generate RobTs (Figure 1a), with self-pollination of double-monosomic plants being the most common method used in the induction of RobTs [26]. Wheat breeders can use a large collection of aneuploid stocks for the induction of wheat-wheat and wheat-alien RobTs. This can be performed in a directed manner by producing the appropriate wheat or alien monosomic lines [28–31]. In wheat breeding the most common RobTs are the T1BL.1RS and T1AL.1RS translocations [25]. Lukaszewski and Curtis developed 1RS.1DL chromosome translocation in triticale "Rhino" and transferred this to cv. "Presto" by backcrossing for several generations [32]. The 1RS.1DL translocation was later used for the induction of multi-breakpoint translocation lines [29,33,34]. A key issue for the centric breakage-fusion mechanism is the frequency at which the centric breakage events involving the univalent occurs. Friebe et al. [26] reported that this can range from 1% to 11%. Another issue is the resultant telocentric chromosomes have a tendency of rejoin as RobTs. It was reported that the frequency of wheat–alien RobTs recovered ranges from 4% to 20%, depending on the genetic background, the chromosomes involved, and environmental conditions used [26,30,31].

To increase the rate of RobT formation, it is therefore necessary to reduce the random factors connected with appearance of centric breaks. In this study, we postulated that the use of telosomic plants for cross-hybridization with plants carrying an alien-substitution will overcome the random process of centric break formation in the univalent (Figure 1b). This approach was used to transfer the *Lr54* + *Yr37* resistance genes into triticale cv. "Sekundo" through a T2RS.2SkL RobT. It was hypothesized that the presence of the 2R donor chromosomes in a telosomic condition would increase the frequency of RobTs recovery (Figure 1b). To test this we used a monosomic substitution triticale line carrying a single 2S<sup>k</sup> chromosome, instead of a 2R triticale chromosome (40 + M2R + M2Sk), and crossed this with a triticale line carrying a single copy of 2RS (short arm of 2R chromosome) and 2RL (long arm of 2R chromosome) telosomic chromosomes (40T + D2RS + D2RL). As a control experiment, the classical approach, based on the self-pollination of double monosomic triticale-*Ae. kotschyi* plants (40T + M2R + M2Sk), was also undertaken (Figure 1a). The goal was to compare the frequencies of chromosomal breaks and recovery of RobTs that arose throughout the two independent approaches tested (Figure 1a,b).

**Figure 1.** Ideogram of the two contrasting approaches for the recovery of Robertsonian translocations.
