Morphological characterization of trichomes shows enormous variation in shape, density and dimensions across the leaves of 14 Solanum species

Abstract Trichomes are the epidermal appendages commonly observed on plant surfaces including leaves, stem and fruits. Plant trichomes have been well studied as a structural plant defence designed to protect plants against abiotic and biotic stressors such as UV rays, temperature extremities and herbivores. Trichomes are primarily classified into glandular and non-glandular trichomes, based on the presence or absence of a glandular head. The plant genus Solanum is the largest genus of family Solanaceae that houses ~3500 species of ecological and economic importance have a diverse set of trichomes that vary in density and morphology. However, due to the incomplete and contradictory classification system, trichomes have subjective names and have been largely limited to be grouped into glandular or non-glandular types. Through this study, we did a complete workup to classify and characterize trichomes on both adaxial and abaxial leaf surface of 14 wild and domesticated species of the genus Solanum. Using electron microscopy, statistical analyses and artistic rendition, we examined finer details of trichomes and measured their density and dimensions to compile a detailed data set which can be of use for estimating the variation in trichome types, and their density, with consequences for understanding their functional roles. Our study is the first of its kind that provides us with a better and well-defined classification, density and dimension analysis to complete the morphological classification of trichomes on both leaf surfaces of a diverse range of members in Solanum genus.


Introduction
Plant surfaces show spectacular variation in the shape, size, location, function and origin of epidermal projections (Werker 2000). The most important and well-studied among these are trichomes: unicellular or multicellular appendages (hair-like structures) originating from epidermal cells of various plant parts including leaves, stems and flowers (Oksanen 2018), and developing outwards (Werker 2000). Trichomes are distributed almost universally in the plant kingdom and exhibit dramatic variation in their morphology (Seithe and Sullivan 1990;van Dam et al. 1999;Adedeji et al. 2007;Kang et al. 2010;Nurit-Silva and Fatima Agra 2011;Munien et al. 2015;Mehdi Talebi et al. 2018) and density (Mehdi Talebi et al. 2018), both intra- (van Dam et al. 1999;Kang et al. 2010;Munien et al. 2015) and interspecifically (Seithe and Sullivan 1990;Navarro and Oualidi 1999;Mannethody and Purayidathkandy 2018;Mehdi Talebi et al. 2018;Yu et al. 2018), and also among and between related and distant plant families (Kariyat et al. 2018;Deore 2020;. For example, Munien et al. (2015) found four types of trichomes (glandular, non-glandular dendritic, non-glandular bicellular and non-glandular multicellular) in Withania somnifera (intraspecific); Yu et al. (2018) found a great variation in trichome morphology, dimensions, distribution and density among seven Mentha species (interspecific); Deore (2020) identified variations in trichomes among 20 species belonging to 12 different plant families and most of the species were reported to have trichomes ranging from unicellular to multicellular, conical to elongated, smooth to grooved, thin to thick walled and with or without a flat disc at the base. Further, Tsujii et al. (2016) found tremendous variation in trichome leaf dry mass per area in the plant tissue of Metrosideros polymorpha at different elevations (Tsujii et al. 2016). Clearly, within flowering plants, trichomes are both ubiquitous and morphologically diverse.
Trichomes, in general, are considered as one of the first line of defences possessed by plants to protect against abiotic stresses such as UV rays, water loss, temperature extremities (Ehleringer 1982;Li et al. 2018;Oksanen 2018) and herbivore damage (Kaur and Kariyat 2020a, b;. Moreover, the leaf trichomes can also act as mechanoreceptors for detection of insects on leaf surface as observed in case of Arabidopsis (Zhou et al. 2017). In addition to defence-related functions, trichomes play a role in water usage strategies through maintenance of leaf water content and stomatal traits to name a few. For example, Pan et al. (2021) showed higher percentage increase of leaf water content in epiphytic plants with trichomes (Pan et al. 2021). Additionally, trichomes can play role in translocation and homeostasis of minerals in plants (Blamey et al. 1986(Blamey et al. , 2015Li et al. 2021). But the relationships of such traits with trichome types and measurements such as density and dimensions are poorly understood.
Broadly, trichomes can be classified into glandular (presence of glandular head) and non-glandular (absence of glandular head) (Werker 2000). Both glandular (Tang et al. 2020) and nonglandular (Karabourniotis et al. 2019) trichomes have been well documented to protect plants either by production of chemicals in their glandular heads or by their sturdy structure that assist plants to adapt and/or protect from environmental conditions such as UV radiations and cold stresses. Glandular trichomes deter herbivory by physically entrapping herbivore into sticky exudates (Tingey and Laubengayer 1981;Neal et al. 1990;Elle et al. 1999;Zalucki et al. 2002), secreting defensive chemical compounds such as proteinase inhibitors (Peiffer et al. 2009), production of volatile organic compounds (Avé et al. 1987;Murungi et al. 2016) or by altering herbivore body odour after providing a sugar-rich first meal (Weinhold and Baldwin 2011). On the other hand, non-glandular trichomes in the Solanum species are mostly spike-like structures which deter herbivory primarily by deterring herbivore movement, feeding and oviposition (Corsi and Bottega 1999;Kennedy 2003;Løe et al. 2007;Dalin et al. 2008;Sletvold et al. 2010;Tian et al. 2012;Weigend et al. 2018;Kariyat et al. 2019). Additionally, non-glandular trichomes can cause post-feeding damage to caterpillars by rupturing of caterpillars' peritrophic membrane (gut lining; Kariyat et al. 2017;Andama et al. 2020). Moreover, some plant species also possess stinging hairs (trichomes with stinging cells which contain irritant fluids) which act as hypodermal syringes and can cause various allergic reactions such as pain, itching, oedema and visible dermal reactions to mammalian herbivores (Ensikat et al. 2021).
Glandular trichomes are more pliable, so may not cause physical damage (Kariyat 2017(Kariyat , 2019 but can be toxic and can release chemicals to intoxicate herbivores (Hare 2005) and attract predators of herbivores in association with carcasses of herbivores (LoPresti et al. 2015). Contrary to this, non-glandular trichomes are usually devoid of toxins but their sharp and edgy structure can cause physical damage to herbivores. Thus, while the size and structure of the apical gland can inform about the amount of toxins and their content, knowledge about various types (e.g. unbranched vs. stellate) and sizes of non-glandular trichomes can appraise us about their functional significanceeither in deterring herbivory or protecting against abiotic stressors. More specifically, in case of non-glandular trichomes, we now have multiple lines of evidence on how variation in trichome type and density differentially defend them against insect herbivores (Cho et al. 2017;Kariyat et al. 2017Kariyat et al. , 2019. For instance,  found significant variation in trichome density on abaxial and adaxial leaf surface of 11 Solanaceae species and demonstrated that this variation has functional consequences for caterpillar growth and feeding. Clearly, a detailed examination and characterization of trichome morphology (even on abaxial and adaxial surfaces) can have multiple benefits, including a reliable and noncontradictory nomenclature of trichomes, understanding the prominent trichome types found in nature and their diversitywhich later can be explored for defensive functions against different herbivores, and abiotic stressors, with possible implications for our efforts to produce better defended plants for sustainable agriculture (Andama et al. 2020).
Solanaceae is one of the most important plant families consisting of 90 genera and ~3000-4000 ecologically and economically important species which are found in all habitats ranging from dry deserts to wet tropical rainforest and have growth habits ranging from small ephemeral herbs to large perennial trees (Knapp, 2004). Among all the genera in Solanaceae, the genus Solanum contributes ~75 % of all species (Symon 1981;Knapp 2002). Solanum genus exhibits tremendous variation and diversity of trichomes; for instance, widely studied and domesticated crops such as tomato (Solanum lycoperiscum) and tobacco (Nicotiana tabacum) have different types of glandular trichomes, while wild weeds such as silverleaf nightshade (Solanum elaeagnifolium) and Carolina horsenettle (Solanum carolinense) possess only non-glandular trichomes (Peiffer et al. 2009;Weinhold and Baldwin 2011;Burrows et al. 2013;Kariyat et al. 2018). A general convention in trichome literature is to reduce the diversity of trichome types by constraining them to just types of glandular and non-glandular, while these types are quite diverse and are often complicated to resolve. Moreover, this basic classification also fails to explain the huge variation among subtypes of glandular and non-glandular trichomes, and consequently potential to explore their function. Luckwill (1943) was the first to classify trichomes of Lycopersicon into seven distinct types (four glandular subtypes and three non-glandular subtypes) of trichomes based on their length, number of stalk and base cells, and the presence or absence of gland. Following Luckwill (1943), Uphof (1962) in a summary for different methods of classification of trichomes concluded that the final classification is still subjective. After that, Payne (1978) provided us with a glossary to name various trichomes or structures found in trichomes to improve trichome nomenclature. Later, Channarayappa et al. (1992) revised the trichome classification by Luckwill (1943) and the revised classification is used frequently in trichome-related studies. While these studies have served as a model for trichome morphology assessment, we used scanning electron microscopy, and artistic rendering along with the previous classical classification systems (Roe 1971) and the glossary provided by Payne (1978) to characterize and classify trichomes on both adaxial (upperside) and abaxial (lowerside) leaf surface of 14 representative species from Solanum.

Plant materials
A mixture of wild and domesticated species of genus Solanum (14 species in total) were included in the study. We bought seeds of forest bitterberry (Solanum anguivi; Seeds of all the species were sown in potting mixture (Sunshine professional growing mix: Sun Gro Horticulture Canada Ltd, Agawam, MA, USA; Tayal et al. 2020) filled trays (12.5″ × 7.5″ × 2″) and kept in controlled environmental conditions (26 °C temperature, ~50 % relative humidity and 16:8 light dark cycle). Germinated seedlings were transplanted in plastic pots (5″ × 4″ × 4″) with similar soil media and environmental conditions and were watered regularly. For electron microscopy, plants of 4-6 weeks of age post-transplanting with at least 10-12 fully developed leaves were used. Young and fully expanded leaves from randomly selected individuals (for each species, sample size varied by treatment, details below) were used for microscopy experiments.

Desktop scanning electron microscope
To capture images from both abaxial and adaxial of leaves for trichome morphology (n = 3-11 plants per side per species), dimension measurements (n = 3-11 plants per side per species) and density analysis (n = 3-11 plants per side per species), a desktop scanning electron microscope (DSEM; SNE-4500 Plus Tabletop; Nanoimages LLC, Pleasanton, CA, USA;  was employed. Circular leaf discs (0.63 cm in diameter) of fresh leaf samples (collection method detailed above) were excised from the plants using a hole punch. No chemical treatments (e.g. glutaraldehyde; Kariyat et al. 2017), critical drying or sputter coating, were done to the leaf samples, and fresh leaf discs mounted on the aluminium stubs using double-sided carbon tape were directly inserted in the DSEM for scanning and image processing. For more details on operational procedures and methodology of DSEM, see  and .

Trichome morphology assessment
Fresh leaf samples (n = 3-11 plants per side per species) as described above were used and magnified ranging from 45× to 1000× depending on trichome type and size, to achieve maximum resolution to extract finer details of trichomes. Images of different trichome types from both abaxial and adaxial surface of leaves were captured at different angles in 3D and later used to classify them. Payne (1978; consisting of glossary for different shapes and structures of/in trichomes), Roe (1971; consisting of terminology for commonly found Solanum trichomes) and Werker (2000; glandular trichome characterization based on structure of secretory head) were the major literature used to characterize trichomes post-image acquisition.

Trichome density assessment
To determine the trichome density from both leaf surfaces of all the species (n = 3-11 plants per side per species), sample preparation was done as described above. The images for trichome count were consistently captured at 60× magnification which contains ~5.32 mm 2 leaf area measured using 'Nanoeye' software linked to DSEM. We calculated the trichome density 1 mm 2 as follows : Trichome density (1 mm 2 ) = Number of trichomes in the image taken at 60 × magnif ication/5.32

Trichome dimension measurements
While scanning the leaf samples (n = 3-11 plants per side per species), once the image achieved maximum resolution visually, scanning was paused using 'Nanoeye' software associated with DSEM and dimensions of various trichome types were measured by tracing trichomes by straight line in 'M. tools' in 'Nanoeye' software. For non-glandular trichomes, length of spikes from base to tip was measured, and in case of glandular trichomes, length of trichome from base to tip and diameter of bulb containing glandular secretions were measured using the measurement tool embedded in the software (Fig. 7). Magnification was altered among samples depending on trichome type at the best resolution.

Line art
The SEM images were used to draw the trichomes manually on paper. The paper was scanned to make a digital copy and the trichomes were traced using a size 4 solid circle brush (Wacom Intuos Pro Digital Graphic Drawing Tablet; Adobe Photoshop). Further, custom brushes in Adobe Photoshop were used to create non-uniform surface of few trichome types. The images were saved into transparent PNG files and inserted into tables with the base of trichomes is on the bottom of the image.

Analysis
Our goal was to characterize and document the trichome types in these 14 species, and to define the finer details on individual trichome types, and their dimensions. Using previous publications as a composite reference, we classified the trichome types using images that reflected its most detailed morphological features. The nomenclature of trichomes was decided by following a checklist of features in the order described below: major shape of the trichome; glandular/ non-glandular trichome type (italicized); additional minute morphological specifications. Further, additional features have also been added as a separate column to know more details of each trichome type. For morphological representation of trichome types, line art was created for each individual trichome type, by a graphical artist. Following this, mean trichome density for each trichome type or trichome types as broad groups (glandular; simple nonglandular; stellate non-glandular) was calculated manually from images at 60× magnification. 60× magnification includes 5.32 mm 2 of leaf area, and thus to get density of trichomes in 1 mm 2 of area, the trichomes density at 60× magnification was divide by a factor of 5.32. The mean ± SE of trichome density has been incorporated in Tables 2 and 3. And, to test whether species and trichomes varied across the 14 species of interest, we also ran a generalized regression analysis with Poisson distribution with species and types (total, glandular and non-glandular) and their interaction as factors, and trichome number as the response variables. Tukey's post hoc tests were conducted to examine pairwise comparisons. All analyses were carried out using JMP15 (SAS Inc., Cary, NC, USA) software and the plots were built using GraphPad Prism (La Jolla, CA, USA). And finally mean ± SE total length and mean ± SE of the diameter of glands on glandular trichomes, and mean ± SE spike length in case of non-glandular trichomes were measured.

Trichome morphology assessment
Basic classification of trichomes includes classifying them into glandular and non-glandular types (italicized; Tables 2 and 3). Here, detailed classification of trichomes was carried out using previously published works as a composite reference. The terminology used for nomenclature of trichomes is described below in Table 1.
Using above-mentioned terminology, we classified all the trichomes found in our samples. Although, three trichome types viz. stellate non-glandular, simple non-glandular and glandular type trichomes were the most found in the Solanum species in our study, these types have also been further characterized into a numerous subtypes based on minor morphological differences ( Fig. 1; Tables 2 and 3).

Density measurements
Consistent with morphological diversity and variation among and within each species, density of trichomes also varied across species Figs 2 and 3; Tables 2 and 3). Oddly, in some of the species, while acquiring images for density count at 60×, we did not observe any trichomes, but while zooming in on different leaf samples at a higher magnification, we observed few trichome types although they were quite rare. These include attenuate basilatus glandular hair with small glandular tip on adaxial leaf surface of S. macrocarpon (Table  2; Serial number 19), attenuate basilatus glandular hair with small glandular tip; mamilla non-glandular hair on abaxial leaf surface of S. lycopersicum (Table 3; 20, 21), subulate glandular hair with multicellular jointed stalk and small glandular tip on abaxial leaf surface of S. melanocerasum (Table 3; 24), subulate non-glandular hair with multiseriate base and tall pedestal on abaxial leaf surface of S. grandiflorum (Table 3; 27), hooked subulate non-glandular hair on abaxial leaf surface of S. caripense (Table 3; 40) and falcate non-glandular hair on abaxial leaf surface of S. ovigerum (Table 3; 44).
The trichome types with highest trichome density include glandular hair with large globular head, single stalk cell and no neck cell on adaxial leaf surface of S. asperolanatum ( Fig.  3; Table 2; 41), and porrect-stellate multiradiate non-glandular hair with subulate rays (5-10 in number) and with short central ray on abaxial leaf surface of S. aethiopicum ( Fig. 2; Table 3; 4). Although the variation was huge, we found that some trichome types had higher trichome density than the other types in each species. For example, in S. anguivi (abaxial) porrect-stellate multiradiate non-glandular hair with subulate rays (5-10 in number) with short/long central ray and pedestal (density: 13.00 ± 1.56; Table 3; 7, 8 and 9) had considerably higher trichome density than the other two trichome types (Table 3; 5 and 6) (Fig. 2). This pattern was observed in almost all the species in which one/two trichome types dominated over the other. In addition to this, occasionally it was difficult at 60× to distinguish between some trichome types, so the density of some trichome types has been compiled. For example, manual counting of both glandular hair with large quadricellular globular head and single stalk cell in case of abaxial leaf surface of S. lycopersicum resulted into total density of 0.94 ± 0.19 (Table 3; 15 and 16). Such cases were observed for all three major trichome types (stellate non-glandular, simple non-glandular and glandular trichomes) (Tables  2 and 3). Additionally, we found significant variation in trichome numbers (total, glandular and non-glandular; at 60× magnification) among species (generalized regression; P ≤ 0.0001) and interaction of species with trichome type (total, glandular and non-glandular) (generalized regression; P ≤ 0.0001) (Figs 4-6), but the variation was non-significant between trichome types (generalized regression; P = 0.6971).

Trichome dimensions
Similar to density, dimensions of each trichome type also varied across species and location ( Fig. 7; Tables 2 and 3). Among nonglandular trichome types, subulate non-glandular hair with multiseriate base and tall pedestal on abaxial leaf surface of S. grandiflorum (Table 3; Serial number 27) was the longest in dimensions, and the shortest non-glandular trichome was subulate non-glandular hair with multicellular jointed stalk and multicellular base on abaxial leaf surface of S. ovigerum (Table  3; 43). Longest glandular trichome type was subulate glandular hair with multicellular jointed stalk, multicellular base, distinct subsidiary cells and small glandular tip on adaxial leaf surface of S. melanocerasum (Table 2; 23), and the shortest glandular trichome type was glandular hair with large globular head on abaxial leaf surface of S. melongena (Table 3; 35). The glandular trichome with largest head was glandular hair with large quadricellular globular head and single stalk cell on abaxial leaf surface of S. lycopersicum (Table 3; 16). The glandular trichome with smallest glandular tip was hooked subulate glandular hair with multicellular jointed stalk and small glandular tip on abaxial leaf surface of S. lycopersicum (Table 3; 17). Dimensions of some trichome types seemed comparable and thus dimension data for few trichome types were collected as one type (stellate and bifid trichomes; all globular glandular trichomes; glandular trichomes with small tip; simple trichomes).
Although the stellate trichomes had almost consistent spike length within a species or/and leaf surface, but central ray of stellate trichomes varied (long/short) resulting into subdividing them into stellate trichomes with short or long central ray. Simple trichomes had the most variation resulting them being both the shortest and the longest trichome found among species in the study. Further, in general, glandular trichomes with globular heads were shorter than glandular   trichomes with small tip on the top. Additionally, glandular trichomes with globular heads had greater diameter of their glandular heads than glandular trichomes with small tip on the top. Due to the high species numbers, processing of multiple samples and incredible trichome diversity, we could not acquire dimensions of some trichome types (e.g. osteolate non-glandular hair with multicellular stalk on abaxial leaf surface of S. dulcamara; Table 3; 57).

Discussion
In this study, we examined the trichome characteristics including their nomenclature, density and dimensions on both adaxial and abaxial leaf surface of 14 Solanum species using scanning electron microscopy on fresh leaf samples. We found that Solanum genus consists of numerous trichome types which vary not only among the species, but also within each species and between adaxial and abaxial leaf surfaces. We also found that three trichome types are most common in Solanum: stellate non-glandular, simple non-glandular and glandular trichomes. Broadly, all trichomes have been characterized into glandular and non-glandular trichomes, but it is not fair, because both glandular and non-glandular trichomes can further be classified into various types based on their shape, size, number of cells, basal cells, neck cells, etc. Besides, it was not certain whether ovoid, verrucate and mamilla trichomes are glandular or nonglandular because of their unusual and perplexing structure but we confirm that they can be characterized as non-glandular trichomes because of their lack of distinction of a clear glandular head.
Trichomes, in general, have been proven to be an excellent phenotypic trait for finding evolutionary and taxonomic relationships among species (Cantino 1990;Eyvazadeh Khosroshahi and Salmaki 2019). For example, Phlomis genus has characteristic multi-nodal-branched trichomes and can be considered as synapomorphy for this group, but Phlomoides lacks this feature and thus, this feature most likely represents a plesiomorphy in the genus (Eyvazadeh Khosroshahi and Salmaki 2019); and separation of African and Asian Leucas spp. was made more explicit by the differences of capitate trichomes with definite morphology and absence of non-glandular trichomes with one cell and more than three cells (Mannethody and Purayidathkandy 2018). Our study characterized the finer details of trichome morphology, and it can be of an aid in exploring phylogenetic and taxonomic relationships among the members of genus Solanum, and their relationship with members of other genus of Solanaceae family and among other plant families. Additionally, trichomes act as excellent cell differentiation models (Hülskamp 2004) and provided with the diversity from this study, it can be explored how differentiation in trichome cells results into production of glandular/non-glandular of various shapes and sizes.
Defence against herbivores is one of the major functions of trichomes, and both glandular and non-glandular trichomes have been well documented to deter herbivore movement and feeding (Kariyat et al. 2013(Kariyat et al. , 2017(Kariyat et al. , 2019. Morphology, density and dimensions relationships of subtypes of trichomes can be employed to find correlations between trichome characteristics with herbivore feeding intensity and behaviour . For instance,  found that significantly higher trichome density on abaxial leaf surface than adaxial leaf surface of Solanaceae species resulted into delayed feeding and lower mass gain of tobacco hornworm caterpillars (Manduca sexta; Lepidoptera: Sphingidae); Kariyat et al. (2017) found damage done to peritrophic membrane (gut lining) of M. sexta caterpillars after feeding on stellate trichomes of horsenettle (S. carolinense; Solanaceae). Further, since the damage done by non-glandular trichomes is primarily because of their structure and many herbivores mow the trichomes off the plant surface before feeding (Kariyat et al. 2017(Kariyat et al. , 2019Kaur and Kariyat 2020b), we speculate that trichomes with a greater number of spikes, and spikes with higher dimensions can result into higher negative impacts on herbivore feeding, but warrant closer examination (Medeiros and Moreira 2005;Andama et al. 2020). The variation in trichome types, density, dimensions and their functional consequences ) between abaxial and  adaxial leaf surfaces also warrants detailed exploration, a reason why we examined these differences in detail (Tables 2 and 3). Contrary to non-glandular trichomes, glandular trichomes are the secretory structures and contain various types of chemicals in their head cells and those chemicals have been found to trigger different defence-related pathways against herbivores (Tissier 2012), and some glandular trichome types have been found to play more prominent roles than the others. For example, type VI (as named by Luckwill 1943; glandular trichomes with quadricellular head) trichomes of Lycopersicon genus were found to contain chemicals possessing insecticidal properties against lepidopteran larvae (Lin et al. 1987). Thus, knowledge of glandular trichome density and dimensions such as length and diameter can help us know the plant parts with higher trichome density, amount of chemicals possessed by trichomes and if the trichomes are tall enough to act against herbivores with its structural features along with chemical defence. Additionally, expanding on this study, histochemistry and volatile collection of various glandular secretions can be done, and anti-herbivore chemicals can be identified (Muravnik et al. 2019), and can lead to further functional assessment and classification, an area that we are currently exploring. Trichomes have been of great importance against herbivores as a defence trait, and thus, have been incorporated in integrated pest management of insect pests of various crops of economic importance including potato (Solanum tuberosum; Solanaceae), cotton (Gossypium spp.; Malvaceae), cowpea (Vigna unguiculata; Fabaceae), to name a few. For instance, hooked non-glandular trichomes of Phaseolus vulgaris (Fabaceae) entrap pests such as black bean aphid (Aphis fabae; Hemiptera: Aphididae) and green stink bug (Nezara viridula; Hemiptera: Pentatomidae) (Rebora et al. 2020). Trichomes also play multiple roles in plants including leaf water uptake and protection from UV light, in addition to defence against herbivores (Sack and Buckley 2020). For example, Li et al. (2021) showed that among three trichome types viz. non-glandular trichomes, linear glandular trichomes and glandular trichomes, only non-glandular trichomes of Sunflower (Helianthus annuus) were found to accumulate and translocate zinc, an important micronutrient for plants, a potential area of research to be explored for mineral and nutrient uptake to enhance crop yield through trichomes. Clearly, trichome characteristics have the potential to be explored in relation to plant ecophysiological functions such as water use efficiency (Thitz et al. 2017), UV protection and mineral uptake.
Taken together, this study documented the variation in trichome types, their density and dimensions in representative Solanum species. So far, the major reasons which withheld the detailed classification and nomenclature of trichomes were (i) the conventional approach to classify trichomes as just glandular or non-glandular trichomes failed to document their morphotypes and function, (ii) the requirement of expensive machinery and skilled professionals, required to process samples across species and families and (iii) the need of intensive and detailed workflow for the closer examination of images to extract all the data on morphology, density and dimensions. By overcoming these challenges, we show the variation in trichome traits within a subset of the genus Solanum and encourage more detailed examination across various plant families.

Conclusions
Overall, the study provides unusually fine details and morphological characterization of trichomes of a mixture of wild and domesticated, annual, and perennial, food crops and weed species of genus Solanum which can act as referral source for further studies of most of trichome-related parameters and their relationships with biotic and abiotic stresses (Tian 2012). Since Solanum is the largest genus of one of the major angiosperm family viz. Solanaceae, exploring trichome diversity by considering 14 species from various groups (e.g. S. macrocarpon and S. lycopersicum are cultivated species, while S. anguivi and S. pyracanthos are wild species) of the family provided us with an updated data source of trichome characteristics with such details that never has been done before. Further, future directions in trichome studies can be focused on understanding variability and organ development while studying gene expression simultaneously using trichomes as a model. Moreover, trichomes are also known to play role in multi-trophic interactions in ecosystem (Weinhold and Baldwin 2011) and thus, each trichome type can be explored for its potential in strengthening plants' defences.